Advances in End-Stage Renal Diseases 2001: International Conference on Dialysis Iii, Miami Beach, Fla., January 2001

Advances in End-Stage Renal Diseases 2001

Nathan W. Levin Claudio Ronco

KARGER

Advances in End-Stage Renal Diseases 2001 International Conference on Dialysis III, January 18–19, 2001, Miami Beach, Fla.

Editors

Nathan W. Levin, New York, N.Y. Claudio Ronco, Vicenza

54 figures, 23 tables, 2001

Basel 폷 Freiburg 폷 Paris 폷 London 폷 New York 폷 New Delhi 폷 Bangkok 폷 Singapore 폷 Tokyo 폷 Sydney

S. Karger Medical and Scientific Publishers Basel 폷 Freiburg 폷 Paris 폷 London New York 폷 New Delhi 폷 Bangkok Singapore 폷 Tokyo 폷 Sydney

Fax+ 41 61 306 12 34 E-Mail [email protected] www.karger.com

Drug Dosage The authors and the publisher have exerted every effort to ensure that drug selection and dosage set forth in this text are in accord with current recommendations and practice at the time of publication. However, in view of ongoing research, changes in government regulations, and the constant flow of information relating to drug therapy and drug reactions, the reader is urged to check the package insert for each drug for any change in indications and dosage and for added warnings and precautions. This is particularly important when the recommended agent is a new and/or infrequently employed drug.

All rights reserved. No part of this publication may be translated into other languages, reproduced or utilized in any form or by any means, electronic or mechanical, including photocopying, recording, microcopying, or by any information storage and retrieval system, without permission in writing from the publisher or, in the case of photocopying, direct payment of a specified fee to the Copyright Clearance Center (see ‘General Information’). © Copyright 2001 by S. Karger AG, P.O. Box, CH–4009 Basel (Switzerland) Printed in Switzerland on acid-free paper by Reinhardt Druck, Basel ISBN 3–8055–쎱쎱–쎱

Vol. 19, No. 2, 2001

Contents

137 Preface Levin, N.W. (New York, N.Y.); Ronco, C. (Vicenza) 139 Vascular Disease and Atherosclerosis in Uremia London, G.M. (Fleury-Mérogis/Paris) 143 Malnutrition and Chronic Inflammation as Risk Factors for Cardiovascular

Disease in Chronic Renal Failure Stenvinkel, P. (Davis, Calif./Mather, Calif.) 152 Contextual Issues in Comparing Outcomes and Care Processes for ESRD

Patients around the World Reddan, D.; Szczech, L.A. (Durham, N.C.); Conlon, P.J. (Dublin); Owen, W.F., Jr. (Durham, N.C.) 157 Should the Hematocrit Be Normalized in Dialysis and in Pre-ESRD Patients? Macdougall, I.C. (London) 168 Should the Hematocrit (Hemoglobin) Be Normalized in Pre-ESRD or

Dialysis Patients? Yes! Besarab, A.; Aslam, M.S. (Morgantown, W. Va.) 175 The Role of Hemodialysis and Peritoneal Dialysis for the Early Initiation of

Dialysis Schulman, G. (Nashville, Tenn.) 179 Peritoneal Dialysis Should Be Considered as the First Line of Renal

Replacement Therapy for Most ESRD Patients Burkart, J.M. (Winston-Salem, N.C.) 185 Technology: Tools or Toys, Is It Economically Feasible with Current

Reimbursement? The Case in Favor Tattersall, J. (Sheffield) 189 A Simplified Method for Adequate Hemodialysis Kleophas, W.; Backus, G. (Düsseldorf) 195 Electrolyte Balancing: Modern Techniques and Outcome Locatelli, F.; Manzoni, C.; Di Filippo, S. (Lecco) 200 Manifestations of Oxidant Stress in Uremia Himmelfarb, J.; McMonagle, E. (Portland, Me.) 206 Daily Hemodialysis: Is It a Complex Therapy with Unproven Benefits? Pierratos, A. (Toronto) 211 Daily Hemodialysis Is a Complex Therapy with Unproven Benefits Gotch, F.A. (San Francisco, Calif.)

© 2001 S. Karger AG, Basel Fax+ 41 61 306 12 34 E-Mail [email protected] www.karger.com

Access to full text and tables of contents, including tentative ones for forthcoming issues: www.karger.com/journals/bpu/bpu_bk.htm

217 Interdialytic Weight Gain and Dry Weight Levin, N.W.; Zhu, F. (New York, N.Y.); Keen, M. (Thousand Oaks, Calif.) 222 Acute Dialysis Quality Initiative Ronco, C. (Vicenza); Kellum, J. (Pittsburgh, Pa.); Mehta, R.L. (San Diego, Calif.) 227 Indications for Dialysis in the ICU: Renal Replacement vs. Renal Support Mehta, R.L. (San Diego, Calif.) 233 Design Issues for Clinical Trials in Acute Renal Failure Star, R. (Bethesda, Md.) 238 Prescription of Adequate Renal Replacement in Critically Ill Patients Paganini, E.P.; Kanagasundaram, N.S.; Larive, B.; Greene, T. (Cleveland, Ohio) 245 Validation of the Blood Temperature Monitor for Extracorporeal Thermal

Energy Balance during in vitro Continuous Hemodialysis Rahmati, S.; Ronco, F.; Spittle, M.; Morris, A.T.; Schleper, C.; Rosales, L.; Kaufman, A.; Amerling, R. (New York, N.Y.); Ronco, C. (Vicenza/New York, N.Y.); Levin, N.W. (New York, N.Y.) 251 Two Years’ Experience with the Dialock® and the CLS-Taurolidine Solution Sodemann, K. (Lahr-Ettenheim); Polaschegg, H.-D. (Koestenberg); Feldmer, B. (Lahr-Ettenheim) 255 Sorbent Augmented Dialysis: Minor Addition or Major Advance in Therapy? Winchester, J.F. (New York, N.Y:); Ronco, C. (Vicenza/New York, N.Y.); Brady, J.A.; Golds, E.; Clemmer, J. (New York, N.Y.); Cowgill, L.D. (Sacramento, Calif.); Muller, T.E.; Levin, N.W. (New York, N.Y.) 260 First Clinical Experience with an Adjunctive Hemoperfusion Device

Designed Specifically to Remove ␤2-Microglobulin in Hemodialysis Ronco, C. (Vicenza/New York, N.Y.); Brendolan, A. (Vicenza); Winchester, J.F.; Golds, E.; Clemmer, J. (New York, N.Y.); Polaschegg, H.D. (Koestenberg); Muller, T.E. (New York, N.Y.); La Greca, G. (Vicenza); Levin, N.W. (New York, N.Y.)

264 Author Index Vol. 19, No. 2, 2001 265 Subject Index Vol. 19, No. 2, 2001

136

Blood Purif Vol. 19, No. 2, 2001

Contents

Blood Purif 2001;19:137–138

Preface

The proceedings of the ‘International Conference on Dialysis II’ (held in Tarpon Springs in January of 2000) have been published as a special issue of Blood Purification and reprinted in a hardcover book entitled ‘Advances in ESRD 2000’. Based on the positive acceptance of the scientific community and the encouraging feedback received on that volume, the publication of the ‘International Conference on Dialysis III’ (held in Miami Beach in January of 2001) has been realized in the same format. This year however, a major innovation has been added: thanks to the cooperation of the authors, the volume has been prepared to be ready at the time of the conference and handed out to all participants. The volume is intended to serve both as a syllabus for a better comprehension of the talks during the conference and as a long-lasting reference publication for future consultation. The present publication contains the manuscripts from the speakers of the conference based on the topic assigned in the program. The program has been designed to offer an educational opportunity for young physicians but at the same time it has been conceived to provide an update on recent technology available in the field of hemodialysis. A series of lectures and debates were presented in which clinical aspects as well as epidemiology and technology were discussed. The conference was enhanced by the presence of an outstanding panel of American and international speakers. Professor London from France presented the first keynote address on vascular disease and atherosclerosis in uremia. The program continued with a debate concerning the opportunity of using peritoneal dialysis as a first treatment modality of ESRD. Drs. Burkart and Shulman sustained their opposite views in a debate in

ABC

© 2001 S. Karger AG, Basel 0253–5068/01/0192–0137$17.50/0

Fax + 41 61 306 12 34 E-Mail [email protected] www.karger.com

Accessible online at: www.karger.com/journals/bpu

which Dr. Levin and Dr. Passlick-Deetjen acted as facilitators for a broader discussion. Another controvers of the conference included the possible benefits achievable with daily dialysis (Drs. Gotch and Pierratos) while Dr. Locatelli from Italy discussed in an interesting lecture the most advanced techniques for electrolyte monitoring and profiling in hemodialysis. A complete session was dedicated to the treatment of acute renal failure. The book not only contains the presentations made by Dr. Star from NIHNDDK, Drs. Metha, Chertow, Paganini and Ronco, but it also contains in a specific chapter the description and definition of ADQI, the new acute dialysis quality inititative whose first conference was held in New York in August 2000 under the auspices of the Renal Research Institute. Dr. Owen presented the second keynote address concerning the factors affecting possible differences observed in the outcomes of hemodialysis patients in different regions or countries. The lecture was followed by another controversy on the opportunity to normalize or not hemoglobin levels in pre-ESRD and hemodialysis patients. Dr. Besarab and Dr. Macdougal were the debators. Dr. Stevinkel from Sweden presented a lecture on malnutrition and chronic inflammation as risk factors for cardiovascular disease in chronic renal failure while Dr. Himmelfarb discussed the clinical aspects of oxidative stress. Dr. Winchester described the potential benefits in the long term of a new technology coupling hemoperfusion with hemodialysis utilizing a new sorbent device. Drs. Moran and Sodemann presented an update on the most recent technologies for alternative vascular access in hemodialysis. One of the most interesting debates concerned the clinical use of the most sophisticated technolo-

gy for hemodialysis including on-line monitors and computerized data acquisition systems. The principal issue discussed by Dr. Tattersall and Dr. Kleophas has been the definition of real utility of these technologies in daily clinical practive and whether they are economically affordable with present reimbursement policies. The conference was preceded by a special session entitled ‘How to manage a dialysis center’ which was specifically designed for nephrology fellows. We appreciate the endorsement of the conference by the American Nephrology Nursing Association, The American Society of Nephrology, The National Kidney Foundation and The Renal Physicians Association. We are indebted to the large number of companies in the field of dialysis that exhibited at the conference and represented important sponsors for the scientific program. We also would like to express out gratitude to the University of Minnesota that was the accredited sponsor for the conference and took the responsibility for content,

138

Blood Purif 2001;19:137–138

quality and scientific integrity of this CME activity. A special sponsorship program was created to offer a registration and attendance to one fellow from each nephrology training program as selected by the program director. A total of over 100 fellows were sponsored in this program. We would like to give our sincere thanks to all the members of the scientific committee for designing an upto-date and interesting program. Our appreciation goes to all the speakers who really contributed to the success of the conference by their knowledge and communication skills. Our gratitude is extended to all the staff of the Renal Research Institute who contributed to the organization of the conference. Last but not least, we acknowledge the unrestricted educational grant offered by Fresenius Medical Care. Our final thanks go to Karger Publishers, for the timely publication of this volume and its outstanding quality. Nathan W. Levin Claudio Ronco

Preface

Blood Purif 2001;19:139–142

Vascular Disease and Atherosclerosis in Uremia Gérard M. London Manhes Hospital, Fleury-Mérogis, and Broussais Hospital, Paris, France

Introduction

Cardiovascular disease is a major cause of morbidity and mortality in patients with end-stage renal disease (ESRD) [1]. While the most frequent underlying cause of these complications is atherosclerosis characterized by the presence of plaques and occlusive lesions [2–5], the spectrum of arterial alterations includes structural changes whose alterations concern principally the viscoelastic properties of large arteries.

Arterial Functions

The arteries have two distinct, interrelated functions: (1) to deliver an adequate supply of blood to peripheral tissues – the conduit function, and (2) to smooth out pressure oscillations due to intermittent ventricular ejection – the cushioning function [6]. Conduit Function The conduit function, i.e., the capacity to transfer blood from the left ventricle to peripheral organs (arterial conductance), is related to the width of the arteries and the almost constancy of mean blood pressure along the arterial tree. This function is efficient, since in conditions of increased demand the blood flow can increase 5–8 times over the baseline value. Alterations in conduit func-

ABC

© 2001 S. Karger AG, Basel 0253–5068/01/0192–0139$17.50/0

Fax + 41 61 306 12 34 E-Mail [email protected] www.karger.com

Accessible online at: www.karger.com/journals/bpu

tion can be functional (endothelium-dependent vasodilation is limited in hypertension, cardiac failure, hypercholesterolemia, smoking) or due to structural remodeling. The principal alterations of conduit function occur through narrowing or occlusion of arteries with restriction of blood flow and resulting ischemia or infarction of tissues downstream. Atherosclerosis characterized by the presence of plaques is the most common disease that disturbs conduit function. Atherosclerosis is primarily an intimal disease, focal and patchy in its distribution. Mechanisms of atherogenesis are complex, including lipid disturbances, thrombogenesis, production of vasoactive substances and growth factors and mediators of inflammation. Atherogenesis depends also on mechanical factors such as alterations in shear stress, with predilection of plaques for sites characterized by disturbances of flow pattern and shear stress, like orifices, bifurcations, bending and pronounced arterial tapering. Dampening Function of Arteries The second role of arteries is to dampen the flow pressure oscillations resulting from intermittent ventricular ejection. Arteries can accommodate the volume of blood ejected from the heart, storing part of the volume during systole and draining this volume during diastole, thereby ensuring continuous perfusion of organs and tissues. The efficiency of dampening function depends on viscoelastic properties of arterial walls (expressed in terms of com-

G.M. London Hôpital Manhes 8, grande rue, Fleury-Mérogis F–91712 Ste-Geneviève-des-Bois (France) Tel. +33 1 69 256485, Fax +33 1 69 256525

pliance, distensibility, or incremental elastic modulus) and their ‘geometric’ characteristics including their diameter and length [6]. The principal alteration in cushioning function is due to the stiffening of arterial walls (i.e. decrease in compliance or distensibility, or increase in elastic modulus), with increase in systolic and pulse pressure as the principal consequences [7]. Two mechanisms are involved. The first involves the generation of a higher pressure wave by the left ventricle ejecting into a stiff arterial system, and a higher velocity (PWV – pulse wave velocity – increases with the stiffening) at which is the pressure wave propagated forward (incident wave) to other arteries [6–8]. The second mechanism is indirect via the influence of increased arterial stiffness and PWV on the timing of incident and reflected pressure waves [8]. Indeed, the incident wave is reflected at any points of structural and functional discontinuity of the arterial tree, generating a reflected wave traveling backward towards ascending aorta. Incident and reflected pressure waves interact and are summed up in a measured pressure wave. The amplitude of the measured pressure wave is determined by the timing between the component waves. The desirable timing is disrupted by increased PWV due to arterial stiffening and increased PWV responsible for an early return of reflected wave from the periphery to the aorta. The earlier return means that the reflected wave impacts on the central arteries during systole rather than diastole, amplifying aortic and ventricular pressures during systole and reducing aortic pressure during diastole. An increase of arterial stiffness is disadvantageous to left ventricular function, inducing left ventricular hypertrophy, increased myocardial oxygen consumption and impairs diastolic myocardial function and ventricular ejection. Increased systolic and pulse pressures accelerate arterial damage, increasing the degenerative changes and arterial stiffening feeding a vicious circle [9–12]. Arterial stiffening is altered primarily during the aging process and in conditions associated with increased collagen content (arteriosclerosis), diffuse fibroelastic intima thickening, fibrosis and calcification, and is generalized throughout the thoracic aorta and central arteries, causing dilation, diffuse hypertrophy.

Arterial Remodeling and Function in ESRD

The arterial alterations in ESRD are heterogeneous and associate atherosclerosis (development of plaques) and remodeling associated with aging (arteriosclerosis) and hemodynamic alterations. Atherosclerosis and arteri-

140

Blood Purif 2001;19:139–142

al occlusive lesions are the most frequent causes of cardiovascular morbidity in patients on renal replacement. Occlusive lesions principally involve the medium-sized conduit arteries, and coronary insufficiency, peripheral artery disease, and cerebrovascular events occupy an important place in the mortality of these patients. The high incidence of atherosclerosis-related complications led Lindner et al. [2] to hypothesize that atherogenesis is accelerated in chronic hemodialysis patients. However, it remains a matter of debate whether or not the atherogenesis of dialysis patients is accelerated and whether or not the nature of atherosclerotic plaques is similar in hemodialysis patients and the general population. Ultrasonographic studies have shown a much higher prevalence of calcified plaques in ESRD patients than in age-matched controls in whom soft plaques are more frequent [5], while ESRD produces atherogenic factors essentially unique to uremia, including dyslipidemia, calcium-phosphate alterations, malnutrition and activation of cytokines and inflammatory mediators. Recent studies have shown that the excessive use of calcium phosphate binders could favor the presence of plaque calcification [13]. The factors specific for ESRD are additive to the number of risk factors observed in subjects with preserved renal function, such as age, hypertension, smoking, diabetes, male gender, and insulin resistance. Many hemodialysis patients already have significant vascular lesions before initiating dialysis and, in many patients, especially older ones, the generalized atherosclerosis can be the primary cause of renal failure. Hypertension is a frequent complication in ESRD, and an association between high blood pressure and occlusive arterial lesions was found in chronic hemodialysis patients. The arterial system in patients with ESRD undergoes structural remodeling which is in many aspects similar to aging and is characterized by dilation, hypertrophy, and stiffening of the aorta and major arteries [14, 15]. Although a large part of the arterial alterations are associated with alterations in hemodynamic factors, nonhemodynamic factors more or less specific to ESRD could play an important role. Chronic increase in blood flow induces dilation of the arterial luminal area and wall hypertrophy [15–17]. In ESRD patients, conditions such as anemia, arteriovenous shunts and overhydration induce a state of chronic volume/flow overload associated with increased systemic and regional blood flow and flow velocity, creating conditions for systemic arterial remodeling. This has been illustrated by cross-sectional studies which showed a direct relationship between the diameter of the aorta and of major arteries and blood flow velocity, as well as by

London

studies indicating that arterial enlargement could be limited by adequate fluid removal during dialysis [14]. Even in the absence of blood pressure changes, the increase in arterial radius is responsible for augmentation of tensile stress (Laplace’s law) that induce activation of hypertrophic process. In comparison with blood pressure and age-matched nonuremic patients, the intima-media thickness of major central arteries is increased in ESRD patients [15, 18]. The increased intima-media thickness is associated with decreased arterial distensibility, increased PWV, and early return of wave reflections [15, 19, 20]. In essential hypertensive patients, decreased arterial distensibility is primarily due to higher distending blood pressure rather than to arterial wall thickening and structural modifications [21]. In ESRD patients where arterial distensibility is decreased in comparison to the age- and blood pressurematched nonuremic population, it is proportional to arterial wall hypertrophy. In ESRD patients arterial hypertrophy is accompanied by alterations of the intrinsic elastic properties of arterial walls (increased Einc). This modification affects elastic and muscular type arteries, including arteries free of atherosclerosis, like the radial artery [22]. The observation that the incremental modulus of elasticity was increased in ESRD patients more strongly favors altered intrinsic elastic properties or major architectural abnormalities like those seen in experimental uremia and the arteries of uremic patients. The nature of these qualitative changes remains to be precisely determined, but several alterations, namely fibroelastic intimal thickening, calcification of elastic lamellae and ground substance deposition are classically observed in these patients [23, 24]. The factors associated with these alterations are not precisely identified, but endothelin [25], parathyroid hormone [26] and chronic inflammatory conditions seem to play an important role.

Consequences of Arterial Remodeling

Arterial stiffening results are increased systolic and pulse pressures, and due to early wave reflections, abnormal increase in aortic and left ventricular systolic pressure. The principal consequence of these alterations is left ventricular hypertrophy [4, 7, 15, 27]. Among ESRD patients, significant relations existed between comparable cardiac and vascular parameters [15] and significant correlations were observed between the common carotid artery intima-media thickness and intima-media crosssectional area and LV wall thickness and/or LV mass. The

Vascular Disease and Atherosclerosis in Uremia

second important consequence of arterial stiffness is compromised coronary perfusion. Cardiac ischemia and alterations in subendocardial perfusion are frequently observed in uremic patients despite patent coronary arteries [28]. In the past, the clinical consequences of arterial stiffening on cardiovascular structure and function have been poorly evaluated. Blacher et al. [29] applied logistic regression and the Cox analysis to characterize a cohort of 241 subjects with ESRD and were able to identify an increased aortic PWV as a significant independent predictor of cardiovascular and all-cause mortality. PWV is a complex parameter integrating arterial geometry and intrinsic elastic properties described by Moens-Korteweg equation: PWV 2 = Eh/2rÚ, where E is the elastic modulus (Einc), r is the radius, h is the wall thickness, and Ú is the fluid density. Blacher et al. [29] have shown that the principal factors associated with the aortic PWV as a predictor of cardiovascular and all-cause mortality in ESRD were the elastic modulus and dilation of arteries.

Conclusions

The principal pathophysiological consequence of vascular alterations in ESRD is decreased arterial distensibility and increased PWV with early wave reflections, whose principal clinical consequences are: increased systolic and pulse pressures, LV hypertrophy and altered coronary circulation. In the absence of controlled studies, it is difficult to propose therapeutical interventions aimed at preventing or treating arterial abnormalities. It is only during recent years that a small number of controlled studies have been conducted which were aimed at examining the effect of antihypertensive drugs on the function of large arteries. It has been shown that long-term administration of either calcium channel blocker nitrendipine, or the ACE inhibitor perindopril led to a decrease in pulse wave velocity and arterial wave reflections, indicative of an improvement of vessel wall elasticity. Nevertheless, these studies did not conclude whether the improvement of elastic properties were due only to decrease in blood pressure or to alterations in intrinsic properties of arterial walls.

Acknowledgments This work was supported by GEPIR (Groupe d’Etude de Physiopathologie de l’Insuffisance Rénale) and UMIF (Union des Mutuelles de L’Ile-de-France).

Blood Purif 2001;19:139–142

141

References 1 Raine AEG, Margreiter R, Brunner FP, Ehrich JHH, Geelings W, Landais P, Loirat C, Mallick NP, Selwood NH, Tufveson G, Valderrabano F: Report on management of renal failure in Europe, XXII, 1991. Nephrol Dial Transplant 1992;7(suppl 2):7–35. 2 Lindner A, Charra B, Sherrard D, Scribner BM: Accelerated atherosclerosis in prolonged maintenance hemodialysis. N Engl J Med 1974;290:697–702. 3 Pascazio L, Bianco F, Giorgini A, Galli G, Curri G, Panzetta G: Echo color Doppler imaging of carotid vessels in hemodialysis patients: Evidence of high levels of atherosclerotic lesions. Am J Kidney Dis 1996;28:713–720. 4 London GM, Drüeke TB: Atherosclerosis and arteriosclerosis in chronic renal failure. Kidney Int 1997;51:1678–1695. 5 Savage T, Clarke AL, Giles M, Tomson CRV, Raine AEG: Calcified plaque is common in the carotid and femoral arteries of dialysis patients without clinical vascular disease. Nephrol Dial Transplant 1998;13:2004–2012. 6 Nichols WW, O’Rourke MF: Vascular impedance; in McDonald’s Blood Flow in Arteries: Theoretical, Experimental and Clinical Principles, ed 4. London, Arnold, 1998. 7 London GM, Guérin AP, Pannier B, Marchais SJ, Benetos A, Safar ME: Increased systolic pressure in chronic uremia: Role of arterial wave reflections. Hypertension 1992;20:10– 19. 8 Nichols WW, Avolio AP, Kelly RP, O’Rourke MF: Effects of age and of hypertension on wave travel and reflections; in O’Rourke M, Safar M, Dzau V (eds): Arterial Vasodilatation: Mechanisms and Therapy. London, Arnold, 1993, 23–40. 9 O’Rourke MF, Kelly RP: Wave reflections in systemic circulation and its implications in ventricular function. J Hypertens 1993;11: 327–337. 10 Watanabe H, Ohtsuka S, Kakihana M, Sugishita Y: Coronary circulation in dogs with an experimental decrease in aortic compliance. J Am Coll Cardiol 1993;21:1497–1506.

142

11 Chang KC, Tseng YZ, Kuo TS, Chen HI: Impaired left ventricular relaxation and arterial stiffness in patients with essential hypertension. Clin Sci 1994;87:641–647. 12 Benetos A, Safar M, Rudnichi A, Smulyan H, Richard JL, Ducimetiere P: Pulse pressure: A predictor of long-term cardiovascular mortality in a French male population. Hypertension 1997;30:1410–1415. 13 Guérin AP, London GM, Marchais SJ, Métivier F: Arteriel stiffening and vascular calcifications in end-stage renal disease. Nephrol Dial Transplant 2000;15:1014–1021. 14 Barenbrock M, Spieker C, Laske V, Heidenreich S, Hohage H, Bachmann J, Rahn KH: Studies of the vessel wall properties in hemodialysis patients. Kidney Int 1994;45:1397– 1400. 15 London GM, Guérin AP, Marchais SJ, Pannier B, Safar ME, Day M, Métivier F: Cardiac and arterial interactions in end-stage renal disease. Kidney Int 1996;50:600–608. 16 Kamiya A, Togawa T: Adaptative regulation of wall shear stress to flow change in the canine carotid artery. Am J Physiol 1980;239:H14– H21. 17 Girerd X, London G, Boutouyrie P, Mourad JJ, Laurent S, Safar M: Remodelling of radial artery and chronic increase in shear stress. Hypertension 1996;27:799–803. 18 Kawagishi T, Nishizawa Y, Konishi T, Kawasaki K, Emoto M, Shoji T, Tabata T, Inoue T, Morii H: High-resolution B-mode ultrasonography in evaluation of atherosclerosis in uremia. Kidney Int 1995;48:820–826. 19 London GM, Marchais SJ, Safar ME, Genest AF, Guérin AP, Metivier F, London AM, Chedid K: Aortic and large artery compliance in end-stage renal failure. Kidney Int 1990;37: 137–142. 20 Barenbrock M, Spieker C, Laske V, Baumgart P, Hoeks APG, Zidek W, Rahn KH: Effect of long-term hemodialysis on arterial compliance in end-stage renal failure. Nephron 1993;65: 249–253.

Blood Purif 2001;19:139–142

21 Laurent S, Girerd X, Mourad JJ, Lacolley P, Beck L, Boutouyrie P, Mignot JP, Safar M: Elastic modulus of the radial artery wall material is not increased in patients with essential hypertension. Arterioscler Thromb 1994;14: 1223–1231. 22 Mourad JJ, Girerd X, Boutouyrie P, Laurent S, Safar ME, London GM: Increased stiffness of radial artery wall material in end-stage renal disease. Hypertension 1997;30:1425–1430. 23 Ibels LS, Alfrey AL, Huffer WE, Craswell PW, Anderson JT, Weil R: Arterial calcification and pathology in uremic patients undergoing dialysis. Am J Med 1979;66:790–796. 24 Amann K, Neusüss R, Ritz E, Irzyniec T, Wiest G, Mall G: Changes of vascular architecture independent of blood pressure in experimental uremia. Am J Hypertens 1995;8:409–417. 25 Demuth K, Blacher J, Guérin AP, Benoit MO, Moatti N, Safar ME, London GM: Endothelin and cardiovascular remodelling in end-stage renal disease. Nephrol Dial Transplant 1998;13: 375–83. 26 Barenbrock M, Hausberg M, Kosch M, Kisters K, Hoeks APG, Rahn KH: Effect of hyperparathyroidism on arterial distensibility in renal transplant recipients. Kidney Int 1998;54:210– 215. 27 Marchais SJ, Guérin AP, Pannier BM, Lévy BI, Safar ME, London GM: Wave reflections and cardiac hypertrophy in chronic uremia: Influence of body size. Hypertension 1993;22: 876–883. 28 Rostand RG, Kirk KA, Rutsky EA: Dialysis ischemic heart disease: Insight from coronary angiography. Kidney Int 1984;25:653–659. 29 Blacher J, Guérin AP, Pannier B, Marchais SJ, Safar ME, London GM: Impact of aortic stiffness on survival in end-stage renal disease. Circulation 1999;99:2434–2439. 30 Blacher J, Pannier B, Guérin AP, Marchais SJ, Safar ME, London GM: Carotid arterial stiffness as a predictor of cardiovascular and allcause mortality in end-stage renal disease. Hypertension 1998;32:570–574.

London

Blood Purif 2001;19:143–151

Malnutrition and Chronic Inflammation as Risk Factors for Cardiovascular Disease in Chronic Renal Failure Peter Stenvinkel Division of Nephrology, Department of Medicine, University of California, Davis, Calif., and Department of Veterans Affairs Medical Center, Mather, Calif., USA

Introduction

Despite the recent considerable improvements in dialysis technology, cardiovascular disease (CVD) still remains the main cause of morbidity and mortality in maintenance hemodialysis (HD) patients. It is obvious that ‘traditional’ risk factors, such as hypertension, chronic heart failure, dyslipidemia, tobacco smoking and diabetes mellitus, may account for a large part of the increased cardiovascular mortality rate observed in these patients. However, based on recent research it is evident that also other, ‘nontraditional’, risk factors, such as inflammation, oxidative stress and malnutrition, may contribute to an increased cardiovascular mortality among dialysis patients. Chronic inflammation, as evidenced by increased levels of various acute phase reactants such as C-reactive protein (CRP), fibrinogen, serum amyloid A (SAA), transferrin, serum albumin and prealbumin, is a common feature in dialysis patients. Various pro-inflammatory cytokines are the major mediators of acute phase protein induction and interleukin (IL)-6 is felt to be the principal cytokine influencing CRP changes. Although the association between inflammation and atherosclerotic CVD by now is well established, the mechanism(s) by which inflammation may accelerate atherosclerosis are not well

ABC

© 2001 S. Karger AG, Basel 0253–5068/01/0192–0143$17.50/0

Fax + 41 61 306 12 34 E-Mail [email protected] www.karger.com

Accessible online at: www.karger.com/journals/bpu

understood. It has been proposed that various acute phase reactants promote atherogenesis by directly affecting different parts of the atherosclerotic process. On the other hand, recent evidence suggests that inflammation may rather be a marker of an atherogenic milieu and that the association is merely indirect. Indeed, inflammation has been proven to be associated with endothelial dysfunction, insulin resistance and oxidative stress, all of which may accelerate atherosclerosis.

Traditional Risk Factors Are Inadequate as Predictors of Cardiovascular Mortality in Dialysis Patients

CVD remains the main cause of morbidity and mortality in maintenance HD patients. The annual mortality rate due to CVD is approximately 9%, which is 10- to 20fold higher than the general population, even when adjusted for age, gender, race and diabetes mellitus [1]. It is evident that the atherosclerotic process is accelerated in CRF patients [2, 3] and it has recently been demonstrated that coronary artery calcification is common and progressive also in young HD patients [4]. The causes of atheroclerotic CVD in the general population are multifactorial

Peter Stenvinkel, MD University of California Davis Department of Internal Medicine, Division of Nephrology TB 136 Davis, CA 95616 (USA) Fax +1 530 752 3791

Fig. 1. A Survival by Kaplan-Meier plot of

173 patients followed during dialysis treatment divided according to subjective global assessment (SGA). SGA 1 = Well-nourished, SGA 2 = mild malnutrition and SGA 3 = moderate malnutrition before start of dialysis treatment (¯2 30.1; p ! 0.0001). B Survival by Kaplan-Meier plot of 176 patients followed during dialysis treatment divided according to CRP levels (! 10 mg/l or x 10 mg/ l, respectively) before start of dialysis treatment (¯2 6.7; p ! 0.01).

and ‘traditional’ risk factors, such as dyslipidemia, left ventricular hypertrophy, diabetes mellitus, hypertension, and tobacco smoking, have all been proven to contribute. Intuitively, it appears reasonable to assume that ‘traditional’ risk factors for the general population are also applicable to HD patients. However, in a recent study by Cheung et al. [5] performed in 936 HD patients, it was found that whereas ‘traditional’ risk factors, such as diabetes mellitus and smoking, were strongly associated with CVD, neither serum total cholesterol nor systolic blood pressure was associated with coronary heart disease. In an ongoing prospective study we have made similar preliminary findings and found that whereas ‘traditional’ risk factors such as age and diabetes mellitus do predict survival during dialysis treatment neither 24-hour ambulatory blood pressure nor serum cholesterol does. Indeed, already in 1982, Degoulet et al. [6] obtained a paradoxical result showing that the higher the serum cholesterol, the better the HD patient survival. Their finding was subsequently confirmed by others [7] and it is now widely appreciated that low cholesterol levels, as observed during malnutrition, may account for this paradoxical finding.

CVD observed in these patients. In an ongoing prospective study, we have found that various markers of malnutrition and inflammation are strong independent predictors of mortality in dialysis patients (fig. 1). Thus, it could be speculated that the impact of ‘nontraditional’ risk factors, such as inflammation and malnutrition, on cardiovascular mortality are so strong in dialysis patients that it obscures the impact of common ‘traditional’ risk factors, such as hypertension, tobacco smoking and dyslipidemia. It may seem puzzling that whereas hypoalbuminemia [7, 8] and inflammation [9, 10] have been shown to be important predictors of mortality, complications from malnutrition and inflammation as such are not common causes of mortality in dialysis patients [11]. In fact, malnutrition accounts for less than 5% of deaths in renal patients while atherosclerotic CVD is by far the commonest cause of mortality in the dialysis population [1]. How can this paradox best be explained? One possible explanation may be the strong documented interactions between CVD and inflammatory as well as nutritional parameters in CRF patients [3]. Based on these findings, a syndrome (MIA) consisting of malnutrition, inflammation and atherosclerosis has been suggested [12].

Nontraditional Risk Factors in Chronic Renal Failure

A number of ‘nontraditional’ risk factors for CVD, such as hyperhomocysteinemia, oxidative stress, vascular calcification, malnutrition and inflammation are commonly found in chronic renal failure (CRF) patients. Thus, it could be speculated that they might provide a rationale for the remarkable prevalence of atherosclerotic

144

Blood Purif 2001;19:143–151

Are There Two Types of Malnutrition in Dialysis Patients?

Malnutrition is a common feature in patients with CRF and is commonly associated with decreased body weight, depleted energy (fat tissue) stores, loss of somatic protein (low muscle mass). Moreover, it has been stated

Stenvinkel

Fig. 2. Association between high-sensitivity CRP and A prealbumin (R = –0.45; p ! 0.01) and B serum albumin (R = –0.37; p !

0.0001) levels in predialysis patients.

that low plasma levels of serum albumin, transferrin, prealbumin and other visceral proteins accompany malnutrition. Various studies show signs of malnutrition in 23–76% of HD and 18–50% of peritoneal dialysis (PD) patients [13, 14] and there has been a growing body of evidence linking poor nutritional status in CRF patients with increasing morbidity and mortality. A number of factors directly associated with the dialytic procedure per se, such as bioincompatibility, nutrient losses in the dialysate and, during PD, poor appetite due to abdominal discomfort and uptake of glucose, may be important contributors to malnutrition in dialysis patients. However, as malnutrition is also prevalent in predialysis patients [3, 15] it is obvious that non-dialysis-related factors also contribute. It was recently, in a large number of patients, demonstrated that as glomerular filtration rate declines, !20–25 ml/min signs of nutritional deterioration develop with declining levels of serum albumin [15]. One component of this decline in the nutritional status may be due to a spontaneous reduction in dietary caloric intake [16]. However, since most malnourished CRF patients also have evidence of inflammation and CVD (table 1) it is possible that dialysis-unrelated factors, such as co-morbidity associated with inflammation and elevated serum levels of pro-inflammatory cytokines, also contribute to malnutrition in CRF. Indeed, Kaizu et al. [17] has demonstrated that the nutritional status in HD patients is affected, at least partly, by circulating IL-6 levels. Moreover, a Japanese group has demonstrated that infusion of recombinant IL-6 lowers serum albumin levels and that IL-6 transgenic mice have a muscle-wasting syndrome [18].

Malnutrition, Inflammation and Vascular Disease in CRF

Table 1. Nutritional and inflammatory parameters and prevalence of CVD in predialysis patients grouped as well-nourished (SGA 1) or malnourished (SGA 1 1), respectively

Age, years Weight, kg Lean body mass, kg Serum albumin, g/l Serum creatinine, Ìmol/l Hand-grip strength, kg Prevalence CVD, % hsCRP, mg/l IL-6, pg/ml Serum hyaluronan ng/ml VCAM-1, ng/ml

SGA 1 (n = 95)

SGA 1 1 (n = 63)

48B1 76.3B1.4 51.8B1.2 34.6B0.6 758B24 37.2B1.3 19 8.2B1.2 6.6B0.8 81.3B6.9 1,105B53

56B1 66.9B1.7 45.7B1.2 32.4B0.8 582B22 25.7B1.4 57 25.7B3.6 11.7B1.6 148.8B21.5 1,436B94

Significance p! 0.0001 0.0001 0.001 0.05 0.0001 0.0001 0.0001 0.0001 0.01 0.001 0.01

The synthetic rate of various serum proteins used as markers of malnutrition, such as serum albumin, prealbumin, retinol binding protein and SAA, are vulnerable to the effects of inflammation (fig. 2). Consequently, their use as nutritional markers in dialysis patients may be problematic. In fact, in healthy subjects subjected to semistarvation [19], or in patients with anorexia nervosa [20], serum albumin levels decline only modestly. Moreover, Bistrian et al. [21] have shown that adult marasmus (protein-calorie malnutrition without inflammation) is associated with preserved serum albumin levels. In fact, the only direct dietary cause for severe hypoalbuminemia seems to be adequate or excessive energy intake when the protein intake is severely limited [22]. Indeed, Heimbür-

Blood Purif 2001;19:143–151

145

Table 2. Proposed features of pure malnutrition (type 1) and inflammatory-associated malnutrition (type 2) [adapted from 12]

S-albumin Co-morbidity Presence of inflammation Food intake Resting energy expenditure Oxidative stress Protein catabolism Reversed by dialysis and nutritional support

Type 1

Type 2

Normal/low Uncommon No Low Normal Increased Decreased

Low Common Yes Low/normal Elevated Markedly increased Increased

Yes

No

ger et al. [23] have reported that serum albumin levels did not differ significantly between well-nourished and malnourished predialysis patients whereas the presence of inflammation was associated with much lower serum albumin levels. Based on these findings, we have proposed that at least two types of malnutrition may be present in CRF patients [12]. Whereas type 1 malnutrition is associated with the uremic syndrome per se, the other cytokine-driven type of malnutrition (type 2) is often associated with significant co-morbidity as outlined in table 2. It is obvious that in the clinical setting these two types of malnutrition may often be combined.

Inflammation Is a Common Feature of Chronic Renal Failure

As inflammation may cause the same changes in the concentration of commonly used biochemical markers of malnutrition, as does inadequate nutritional intake, much of the previously reported relations between serum albumin and total and cardiovascular mortality in HD [7] and PD [8] patients may have been caused by an ongoing inflammatory process rather than poor nutritional intake. Recently, it has been recognized that about 30–50% of predialysis [3], HD [9, 10, 13, 24], and PD [25] patients have serologic evidence of an activated inflammatory response as evidenced by elevated CRP levels. It should be emphasized that CRP measurements in these crosssectional studies were made at a single time point, which may complicate interpretation as it is well recognized that CRP may be a ‘moving target’. Longitudinal studies with repetitive CRP measurements are therefore warranted as Kaysen et al. [26] recently showed that the acute phase

146

Blood Purif 2001;19:143–151

Table 3. Potential causes of inflammation in patients with CRF

Chronic renal failure Reduced renal clearance of cytokines Reduced renal clearance of advanced glycation end products (AGEs) Chronic heart failure The atherosclerotic process per se Various inflammatory diseases Unrecognized persistent infections, e.g., C. pneumoniae, H. pylori, dental and/or gingival infections Additional causes in HD Graft and fistula infections Bioincompatibility Exposure to endotoxins and other cytokine-inducing substances from contaminated dialysate, e.g., backfiltration Additional causes in PD Peritonitis Peritoneal access Bioincompatibility Exposure to endotoxins, plasticizers and other cytokine-inducing substances from contaminated dialysate

response is intermittent and varies significantly in time with no change in dialyzer type or treatment. The observation by Kaysen et al., together with the high prevalence of elevated CRP documented in predialysis patients [3], suggest that factors unrelated to dialysis therapy, such as co-morbidity, might be the most important causes of elevated CRP in CRF patients (table 3). However, several lines of evidence suggest that also factors associated with the dialysis procedure per se might contribute to an inflammatory response (table 3). At first, Haubitz et al. [27] have demonstrated that acute phase proteins are induced during HD, probably due to cytokine release, and Schindler et al. [28] suggest that the dialyzer membrane may play a role in the induction of an inflammatory reaction during the dialysis procedure. Moreover, data by Memoli et al. [29] suggest an important role of poor dialysis bioincompatibility of cuprophane on enhancing the inflammatory effects of IL-6. Available evidence also suggests that the quality of water used to prepare the dialysate might contribute to inflammation [30]. Indeed, Schindler et al. [28] could recently demonstrate that optimized HD therapy using ultrapure dialysate and biocompatible dialyzer membranes was able to reduce, but not to normalize, elevated CRP levels in HD patients again suggesting that non-dialysis-related factors contribute to inflammation in CRF.

Stenvinkel

Causes of Inflammation in Chronic Renal Failure

Does Chronic Inflammation Cause Atherosclerotic Cardiovascular Disease?

Serum levels of CRP appear to reflect generation of pro-inflammatory cytokines (IL-1, IL-6 and TNF-·) which have been reported to be markedly elevated in CRF patients [31, 32] and also to predict mortality [31, 33]. The cause(s) of elevated serum levels of pro-inflammatory cytokines in CRF patients are not well understood. However, available evidence suggests that both decreased renal clearance [34] as well as increased synthesis of cytokines might contribute. In this respect it is of interest that reduced renal function may affect both TNF [35] and IL-1 [36] clearance in nephrectomized rats. The importance of the kidney in cytokine handling is further underscored by Hession et al. [37] who demonstrated that the TammHorsfall glycoprotein might function as a unique renal regulatory glycoprotein that regulates the activity of potent cytokines, such as IL-1 and TNF. As not all dialysis patients have elevated CRP levels it has been proposed that a polymorphism in the genes encoding pro-inflammatory cytokines contribute to the observed differences in the prevalence of elevated CRP. However, a recent preliminary study, in a small patient material, concluded that TNF and interferon gene polymorphism do not play a role in determining CRP levels in dialysis patients [38]. The kidney may play an important role in the metabolism of advanced glycation end products (AGEs), which may be another factor that play a role in activating mononuclear cells and stimulate an inflammatory response. Conversely, inflammation may also play a role in the production of AGEs [39]. Other non-dialysis-related causes of elevated CRP might include factors such as chronic heart failure with edema [40] and the atherosclerotic process per se. In fact, by virtue of its acute phase behavior it has been suggested that CRP may be a marker for severity and progression of atherosclerotic processes in the vessels [41]. However, as many patients with stable and unstable angina pectoris have normal levels of acute phase proteins, this implies that coronary atherosclerosis itself does not always induce a full-blown acute phase response. Whereas noninfectious causes may be the most common cause of an inflammatory response in CRF patients, it should be recognized that also various chronic persistent infections such as Chlamydia pneumoniae [42, 43], Helicobacter pylori and dental or gingival infections may contribute. In fact, in a recent preliminary study of 104 HD patients, it was demonstrated that CRP levels correlated with the titer of periodontal pathogens suggesting that poor dental status may contribute to inflammation [44].

There is no doubt that the most significant process that correlates with inflammation is atherosclerotic CVD, although the relationship between these two processes is complex. Recent studies have established that even small increases in the levels of pro-inflammatory cytokines, such as IL-6 [45], or acute phase proteins such as CRP [46], predict CVD in otherwise healthy adults. Moreover, in non-renal patient populations, elevated levels of CRP are associated with ischemic stroke [47] and mortality in nondisabled older people [48]. In a recent updated metaanalysis including 2,557 cases, Danesh et al. [49] reported that the combined risk ratio for coronary heart disease was 1.9 times higher in the patients that had the highest CRP compared to those in the bottom third. Several groups have by now reported that increased CRP is a strong risk factor for death also among HD [9, 10, 50, 51] and PD [52] patients. Moreover, elevated CRP has been shown to predict cardiovascular mortality [9, 10] and hospitalization [53] in dialysis patients. In fact, available data suggest that the association between inflammation and atherosclerosis is particularly strong in dialysis patients. A strong relation between malnutrition, elevated CRP levels and atherosclerosis has also been documented in predialysis patients [3] and serum hyaluronan, another inflammatory marker generated in response to pro-inflammatory cytokines, is a powerful predictor of mortality in CRF patients commencing renal replacement therapy [54]. Although the association between CVD and inflammation in the dialysis patient population is well documented, we do not know if the acute phase response merely reflects an epiphenomenon accompanying established atherosclerotic disease or whether different acute phase reactants are involved in the initiation and/or progression of atherosclerosis (table 4). In fact, several lines of evidence suggest that different acute phase reactants actually may be directly involved as causative factors in atherogenesis. At first, Torzewski et al. [55] have demonstrated that CRP deposit in the arterial wall of early atherosclerotic lesions. Secondly, as CRP has been shown to localize in heart tissue it has been hypothesized that CRP may directly cause tissue damage [56]. Moreover, during inflammation, SAA is incorporated into HDL and the interaction between acute inflammation and lipoprotein structure could provide one possible link between accelerated vascular disease and inflammation, as recently reviewed by Kaysen [57]. Finally, other acute phase reactants, such as Lp(a)

Malnutrition, Inflammation and Vascular Disease in CRF

Blood Purif 2001;19:143–151

147

Table 4. a Possible direct atherogenic mechanisms by which various acute

phase reactants may cause atherosclerotic CVD CRP deposits in the arterial wall CRP causes direct tissue damage SAA affect lipoprotein structure Lp(a) promotes athero- and thrombogenesis Fibrinogen promotes athero- and thrombogenesis and increases plasma, viscosity b Possible indirect mechanisms by which an acute phase reaction may be associated with atherosclerotic CVD

Endothelial dysfunction, e.g., nitric oxide, soluble adhesion molecules Insulin resistance Increased oxidative stress Stimulation of advanced glycation end products (AGEs) Persistent atherogenic infections, e.g., C. pneumoniae, H. pylori

[58] and fibrinogen [59], may have direct atherogenic and/or thrombogenic properties that may accelerate atherogenesis. On the other hand, several lines of evidence suggest that the association between chronic inflammation and CVD is indirect. It is documented that chronic inflammation is associated with features, such as endothelial dysfunction, insulin resistance and increased oxidative stress, all of which may accelerate atherogenesis.

Does Inflammation Cause Endothelial Dysfunction?

Recent evidence suggests that endothelial dysfunction, which is thought to be associated with early atherosclerotic CVD, may be more prevalent in conditions associated with malnutrition and inflammation. As inflammation has been shown to be associated with reduced bioavailability of nitric oxide, this suggests that endothelial dysfunction may be a critical intermediate phenotype in the relationship between inflammation and CVD [60]. In this respect it is of interest that although Kim et al. [61] showed correlations between serum albumin, CRP and serum markers of endothelial function, an infusion of albumin did not normalize endothelial function. Consequently, their findings suggest that the relationship between low serum albumin levels and endothelial dysfunction may be secondary to other factors, such as inflamma-

148

Blood Purif 2001;19:143–151

tion. Indeed, recent observations have shown correlations between inflammation and various circulating markers of endothelial activation in both type 1 diabetic [62] and CRF [63] patients. Bhagat and Vallance [64] have showed that infusion of endotoxins or pro-inflammatory cytokines in healthy volunteers caused a selective impairment of endothelium-dependent relaxation. Moreover, Sinisalo et al. [65] and Cleland et al. [60] recently showed a relationship between low-grade chronic inflammation and endothelial dysfunction. Finally, long-term exposure of vascular endothelium to IL-1ß and TNF-· causes endothelial dysfunction, intimal thickening and coronary vasospasm in pigs [66]. Taken together, these findings suggest that a synergy exists between the presence of chronic inflammation and the development of malnutrition contributing to cardiovascular risk via an endothelial dysfunction. Inflammation may also affect endothelial function by altering the expression of soluble adhesion molecules (sICAM-1, sVCAM-1 and E-selectin) which has been shown to promote monocyte binding to endothelial cells. In this respect it is of interest that elevated levels of ICAM-1 have been shown to be a prognostic risk factor for future cardiovascular events in both men [67] and women [68]. It is therefore of interest that markedly elevated serum levels of soluble adhesion molecules have been documented in both predialysis [69, 70] and dialysis [70] patients. Bonomini et al. [70] suggested that inadequate clearance contributes to elevated serum levels of adhesion molecules in CRF. However, as it has repeatedly been demonstrated that pro-inflammatory cytokines can upregulate the expression of adhesion molecules from endothelial cells [71], increased synthesis may also contribute to elevated serum levels of adhesion molecules in CRF patients. As ICAM-1 has recently been shown to be an independent predictor of death in dialysis patients [63], further studies are needed to investigate if inflammation may cause accelerated atherogenesis via effects on soluble adhesion molecules.

Inflammation and Insulin Resistance

Insulin resistance is a well-documented feature of CRF [72] which is associated with a premature development of atherosclerosis. Several reports in nonrenal patient populations suggest that elevated CRP levels are associated with several different features of the insulin resistance syndrome including increased body mass index [73–75], serum lipids [73, 74], and fasting glucose [73]. Recently,

Stenvinkel

Festa et al. [76] demonstrated in 1,008 nondiabetic patients, of whom one third had impaired glucose tolerance, that there was a linear increase in CRP with an increase in the number of metabolic disorders. Consequently, chronic low-grade inflammation may be part of the insulin resistance syndrome. In this respect it is of interest that clamp studies in normal subjects have shown that insulin exerts selective effects on hepatic protein synthesis with an increase in serum albumin and decrease in fibrinogen synthesis [77]. It is therefore possible that decreased insulin sensitivity may lead to enhanced CRP expression by counteracting the physiological effects of insulin on hepatic acute phase synthesis [78]. However, other mechanisms may also be operative and it is possible that chronic inflammation may simply represent a triggering factor in the origin of insulin resistance as discussed by Festa et al. [76].

Does Inflammation Cause Increased Oxidative Stress?

Oxidative stress, which occurs when there is excessive free radical production or low antioxidant levels, has emerged as an important co-factor for the development of endothelial dysfunction and atherogenesis [79]. Free radicals are involved in the development of atherosclerosis by generating oxidized low-density lipoprotein (LDL), which by various mechanisms, damage the vascular wall and cause atherosclerotic lesions. Recent data support the hypothesis that increased oxidative stress is present in HD patients [80] and it is probable that loss of antioxidants, such as vitamins C and E, occurs in CRF or as a consequence of the dialysis treatment per se. Plasmalogen phospholipids have been demonstrated to play a significant role in the defense of LDL particles against oxidative stress [81]. As we have found lower levels of erythrocyte plasmalogen phospholipids in malnourished compared to well-nourished predialysis patients, one could speculate that an increased oxidative stress might be a contributing factor to the high prevalence of CVD documented in malnourished and inflamed CRF patients [82]. The cause(s) of increased oxidative stress in malnourished patients are not well understood. However, in view of the documented strong relations between malnutrition and inflammation, is it possible that a chronic inflammatory response may be the primary cause of increased oxidative stress in malnourished CRF patients. Indeed, recent results by Memon et al. [83] have demonstrated that LDL isolated from animals treated with bac-

Malnutrition, Inflammation and Vascular Disease in CRF

terial lipopolysacharide (LPS) was significantly more susceptible to ex vivo oxidation with copper than LDL isolated from saline-treated animals. Unfortunately, data on the effect of inflammation on oxidative stress in CRF patients are scarce. However, recently Nguyen-Khoa et al. [84] found that the inflammatory status and duration of dialysis treatment are the most important factors relating to oxidative stress in HD patients.

Conclusions

CRF is characterized by an exceptionally high mortality rate, much of which is the result of CVD. Recent evidence demonstrates that chronic inflammation, as evidenced by increased levels of pro-inflammatory cytokines and CRP, is a common feature in CRF patients. Chronic inflammation may cause malnutrition and progressive atherosclerotic CVD by several pathogenetic mechanisms and could, thus, contribute to the high mortality rate. The cause(s) of inflammation is probably multifactorial and while it may reflect underlying CVD, it may also be a direct cause of vascular injury. As available data suggest that pro-inflammatory cytokines plays a central role in the genesis of both ‘inflammatory-driven’ malnutrition and CVD, it would be of obvious interest to study what effect anticytokine therapy (such as anti-TNF-· antibodies, soluble TNF-· receptors, IL-1 receptor antagonists and thalidomide) have on survival in dialysis patients. The impact of various anti-inflammatory and antibiotic treatment strategies may also be of interest to study in dialysis patients with signs of malnutrition, inflammation and atherosclerosis (MIA syndrome). In this respect it should be pointed out that Ridker et al. [85] recently demonstrated that whereas CRP levels tended to increase over 5 years in those patients that received placebo, randomization to pravastatin (a HMG-CoA reductase inhibitor) resulted in a significant reduction in CRP that was not related to the magnitude of lipid alterations observed.

Acknowledgements The Trone-Holst Foundation, The Swedish Medical Association, Hospal and the Swedish Medical Research Foundation supported the present work.

Blood Purif 2001;19:143–151

149

References 1 Foley RN, Parfrey PS, Sarnak MJ: Clinical epidemiology of cardiovascular disease in chronic renal failure. Am J Kidney Dis 1998;32(suppl 5):S112–S119. 2 Lindner A, Charra B, Sherrard DJ, Schribner BH: Accelerated atherosclerosis in prolonged maintenance haemodialysis. N Engl J Med 1974;290:697–701. 3 Stenvinkel P, Heimbürger O, Paultre F, Diczfalusy U, Wang T, Berglund L, et al: Strong associations between malnutrition, inflammation and atherosclerosis in chronic renal failure. Kidney Int 1999;55:1899–1911. 4 Goodman WG, Goldin J, Kuizon BD, Yoon C, Gales B, Sider D, et al: Coronary-artery calcification in young adults with end-stage renal disease who are undergoing dialysis. N Engl J Med 2000;342:1478–1483. 5 Cheung AK, Sarnak MJ, Yan G, Dwyer JT, Heyka RJ, Rocco MV, et al: Atherosclerotic cardiovascular disease risks in chronic hemodialysis patients. Kidney Int 2000;58:353–362. 6 Degoulet P, Legrain M, Reach I, Aime F, Devres C, Rojas P, et al: Mortality risk factors in patients treated by chronic hemodialysis. Nephron 1982;31:103–110. 7 Lowrie EG, Lew NL: Death risk in hemodialysis patients: The predictive value of commonly measured variables and an evaluation of death rate differences between facilities. Am J Kidney Dis 1990;15:458–482. 8 Avram MM, Fein PA, Bonomini L, Mittman N, Loutoby R, Avram DK, et al: Predictors of survival in continuous ambulatory peritoneal dialysis patients: A five-year prospective study. Perit Dial Int 1996;16(suppl 1):S190–S194. 9 Zimmermann J, Herrlinger S, Pruy A, Metzger T, Wanner C: Inflammation enhances cardiovascular risk and mortality in hemodialysis patients. Kidney Int 1999;55:648–658. 10 Yeun JY, Levine RA, Mantadilok V, Kaysen GA: C-reactive protein predicts all-cause and cardiovascular mortality in hemodialysis patients. Am J Kidney Dis 2000;35:469–476. 11 Bergstrom J, Lindholm B: Malnutrition, cardiac disease and mortality – An integrated point of view. Am J Kidney Dis 1998;32:834– 841. 12 Stenvinkel P, Heimbürger O, Lindholm B, Kaysen GA, Bergstrom J: Are there two types of malnutrition in chronic renal failure? Evidence for relationships between malnutrition, inflammation and atherosclerosis (MIA syndrome). Nephrol Dial Transplant 2000;15: 953–960. 13 Qureshi AR, Alvestrand A, Danielsson A, Divino-Filho JC, Gutierrez A, Lindholm B, et al: Factors influencing malnutrition in hemodialysis patients. A cross-sectional study. Kidney Int 1998;53:773–782. 14 Bergstrom J, Lindholm B: Nutrition and adequacy of dialysis. How do hemodialysis and CAPD compare? Kidney Int 1993;34:S39–S50. 15 Kopple JD, Greene T, Chumlea WC, Hollinger DL, Maroni BJ, Merrill D, et al: Relationship between GFR and nutritional status – Results

150

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

Blood Purif 2001;19:143–151

from the MDRD study. Kidney Int 2000;57: 1688–1703. Ikizler TA, Greene JH, Wingard RL, Parker RA, Hakim RM: Spontaneous dietary intake during progression of chronic renal failure. Kidney Int 1995;6:1386–1391. Kaizu Y, Kimura M, Yoneyama T, Miyaji K, Hibi I, Kumagai H: Interleukin-6 may mediate malnutrition in chronic hemodialysis patients. Am J Kidney Dis 1998;31:93–100. Tsujinaka T, Fujita J, Ebisuri C, Yano M, Kominami E, Suzuki K, et al: Interleukin-6 receptor antibody inhibits muscle atrophy and modulates proteolytic systems in interleukin-6 transgenic mice. J Clin Invest 1996;97:244– 249. Keys A, Brozek J, Henschel A, Mickelsen O, Taylor HL: The Biology of Human Starvation. Minneapolis, The University of Minnesota Press, 1950. Smith G, Robinsson PH, Fleck A: Serum albumin distribution in early treated anorexia nervosa. Nutrition 1996;12:677–684. Bistrian B, Sherman M, Blackburn G, Marshall R, Shaw C: Cellular immunity in adult marasmus. Arch Intern Med 1977;137:1408–1411. Lunn P, Austin S: Dietary manipulation of plasma albumin concentrations. J Nutr 1983; 113:1791–1801. Heimbürger O, Qureshi AR, Blaner WS, Berglund L, Stenvinkel P: Hand-grip muscle strength, lean body mass and plasma proteins as markers of nutritional status in patients with chronic renal failure close to start of dialysis therapy. Am J Kidney Dis 2000, in press. Owen WF, Lowrie EG: C-reactive protein as an outcome predictor for maintenance hemodialysis patients. Kidney Int 1998;54:627–636. Yeun JY, Kaysen GA: Acute phase proteins and peritoneal dialysate albumin loss are the main determinants of serum albumin and peritoneal dialysis patients. Am J Kidney Dis 1997; 30:923–927. Kaysen GA, Dublin JA, Muller HG, Rosales LM, Levin NW: The acute phase response varies with time and predicts serum albumin levels in hemodialysis patients. Kidney Int 2000; 58:346–352. Haubitz M, Schulze M, Koch KM: Increase of C-reactive protein serum values following haemodialysis. Nephrol Dial Transplant 1990;5: 500–503. Schindler R, Boenisch O, Fischer C, Frei U: Effect of the hemodialysis membrane on the inflammatory reaction in vivo. Clin Nephrol 2000;53:452–459. Memoli B, Postiglione L, Cianciaruso B, Bisesti V, Cimmaruta C, Marzano L, et al: Role of different dialysis membranes in the release of interleukin-6 soluble receptor in uremic patients. Kidney Int 2000;58:417–424. Tielemans C, Husson C, Schurmans T, Gastaldello K, Madhoun P, Delville JP, et al: Effects of ultrapure and non-sterile dialysate on the inflammatory response during in vitro hemodialysis. Kidney Int 1996;49:236–243.

31 Kimmel PL, Phillips TM, Simmens SJ, Peterson RA, Weihs KL, Alleyne S, et al: Immunologic function and survival in hemodialysis patients. Kidney Int 1998;54:236–244. 32 Pereira BJG, Shapiro L, King AJ, Falagas ME, Strom JA, Dinarello CA: Plasma levels of IL1ß, TNF-· and their specific inhibitors in undialyzed chronic renal failure, CAPD and hemodialysis patients. Kidney Int 1994;45:890– 896. 33 Bologa RM, Levine DM, Parker TS, Cheigh JS, Seur D, Stenzel KH, et al: Interleukin-6 predicts hypoalbuminemia, hypocholesterolemia, and mortality in hemodialysis patients. Am J Kidney Dis 1998;32:107–114. 34 van Riemsdijk-van Overbeeke IC, Baan CC, Hesse CJ, Loonen EHM, Niesters HGM, Zietse R, et al: TNF-·: mRNA, plasma protein levels and soluble receptors in patients with chronic hemodialysis, on CAPD and with end-stage renal failure. Clin Nephrol 2000;53:115–123. 35 Bemelmans MH, Gouma DJ, Buurman WA: Influence of nephrectomy on tumor necrosis factor clearance in murine model. J Immunol 1993;150:2007–2017. 36 Poole S, Bird TA, Selkirk S, Gaines-Das RE, Choudry Y, Stephenson SL, et al: Fate of injected interleukin-1 in rats: Sequestration and degradation in the kidney. Cytokine 1990; 2:416–422. 37 Hession C, Decker JM, Sherblom AP, Kumar S, Yue CC, Mattaliano RJ, et al: Uromodulin (Tamm-Horsfall glycoprotein): A renal ligand for lymphokines. Science 1987;237:1479– 1484. 38 Varagunam M, McCloskey DJ, Sinott PJ, Raftery MJ, Yaqoob MM: Proinflammatory cytokine gene polymorphism and C-reactive protein in dialysis patients. J Am Soc Nephrol 2000;11:305A. 39 Anderson MM, Requena JR, Crowley JR, Thorpe SR, Heinecke JW: The myeloperoxidase system of human phagocytes generates Nepsilon-(carboxymethyl)lysine on proteins: A mechanism for producing advanced glycation end products as sites of inflammation. J Clin Invest 1999;104:103–113. 40 Niebauer J, Volk HD, Kemp M, Dominguez M, Schumann RR, Rauchhaus M, et al: Endotoxin and immune activation in chronic heart failure: A prospective cohort study. Lancet 1999;353:1838–1842. 41 Heinrich J, Schulte H, Schonfeld R, Kohler E, Assmann G: Association of variables of coagulation, fibrinolysis and acute phase with atherosclerosis in coronary and peripheral arteries and those arteries supplying the brain. Thromb Haemost 1995;73:374–378. 42 Stenvinkel P, Heimbürger O, Jogestrand O, Karnell A, Samuelsson A: Does persistent infection with Chlamydia pneumoniae increase the risk of atherosclerosis in chronic renal failure? Kidney Int 1999;55:2531–2532. 43 Cardiovascular Risk Extended Evaluation in Dialysis Patients (CREED): C-reactive protein and atherosclerosis in dialysis patients. Nephrol Dial Transplant 1998;13:2710–2711.

Stenvinkel

44 Spittle M, Craig R, Adhikarla R, Ronco C, Lewin NW: Relationship between antibodies to peridontal pathogens and C-reactive protein levels in hemodialysis patients. J Am Soc Nephrol 2000;11:299A. 45 Ridker PM, Rifai N, Stampfer MJ, Hennekens CH: Plasma concentration of interleukin-6 and the risk of future myocardial infarction among apparently healthy men. Circulation 2000;101: 1767–1772. 46 Ridker PM, Cushman M, Stampfer MJ, Russell PT, Hennekens CH: Inflammation, aspirin and the risk of cardiovascular disease in apparently healthy men. N Engl J Med 1997;336: 973–979. 47 Muir KW, Weir CJ, Alwan W, Squire IB, Lees KR: C-reactive protein and outcome after ischemic stroke. Stroke 1999;30:981–985. 48 Harris TB, Ferrucci L, Tracy RP, Corti MC, Wacholder S, Ettinger WHJ, et al: Association of elevated interleukin-6 and C-reactive protein levels with mortality in the elderly. Am J Med 1999;106:506–512. 49 Danesh J, Whincup P, Walker M, Lennon L, Thomson A, Appleby P, et al: Low grade inflammation and coronary heart disease: Prospective study and updated meta-analyses. BMJ 2000;321:199–204. 50 Bergstrom J, Heimbürger O, Lindholm B, Qureshi AR: Elevated serum C-reactive protein is a strong predictor of increased mortality and low serum albumin in hemodialysis patients. J Am Soc Nephrol 1995;6:573. 51 Iseki K, Tozawa M, Yoshi S, Fukiyama K: Serum C-reactive and risk of death in chronic dialysis patients. Nephrol Dial Transplant 1999;14:1956–1960. 52 Noh H, Lee SW, Kang SW, Shin SK, Choi KH, Lee HY, et al: Serum C-reactive protein: A predictor of mortality in continuous ambulatory peritoneal dialysis patients. Perit Dial Int 1998;18:387–394. 53 Ikizler TA, Wingard RL, Harvell J, Shyr Y, Hakim RM: Association of morbidity with markers of nutrition and inflammation in chronic hemodialysis patients: A prospective study. Kidney Int 1999;55:1945–1951. 54 Stenvinkel P, Heimbürger O, Wang T, Lindholm B, Bergstrom J, Elinder CG: High serum hyaluronan indicates poor survival in renal replacement therapy. Am J Kidney Dis 1999;34: 1083–1088. 55 Torzewski J, Torzewski M, Bowyer DE, Frohlich M, Koenig W, Waltenberger J, et al: Creactive protein frequently colocalizes with the terminal complement complex in the intima of early atherosclerotic lesions of human coronary arteries. Arterioscler Thromb Vasc Biol 1998; 18:1386–1392. 56 Lagrand WK, Niessen HWM, Wolbink GJ, Jaspars LH, Visser CA, Verheugt FWA, et al: C-reactive protein colocalizes with complement in human hearts during acute myocardial infarction. Circulation 1997;95:97–103. 57 Kaysen GA: The microinflammatory state in uremia – Causes and potential consequences. J Am Soc Nephrol 2000, in press.

Malnutrition, Inflammation and Vascular Disease in CRF

58 Sandkamp M, Funke H, Schulte H, Kohler E, Assmann G: Lipoprotein(a) is an independent risk factor for myocardial infarction at a young age. Clin Chem 1990;36:20–23. 59 Smith EB, Thompson WD: Fibrin as a factor in atherogenesis. Thromb Res 1994;73:1–19. 60 Cleland SJ, Sattar N, Petrie JR, Forouhi NG, Elliott HL, Connell JMC: Endothelial dysfunction as a possible link between C-reactive protein levels and cardiovascular disease. Clin Sci 2000;98:531–535. 61 Kim SB, Chi HS, Park JS, Hong CD, Yang WS: Effect of increasing serum albumin on plasma D-dimer, von Willebrand factor, and platelet aggregation in CAPD patients. Am J Kidney Dis 1999;33:312–317. 62 Schalkwijk CG, Poland DC, van Dijk W, Kok A, Emeis JJ, Drager AM, et al: Plasma concentrations of C-reactive protein is increased in type 1 diabetic patients without clinical macroangiopathy and correlates with markers of endothelial dysfunction: Evidence for chronic inflammation. Diabetologia 1999;42:351–357. 63 Stenvinkel P, Lindholm B, Heimbürger M, Heimbürger O: Elevated serum levels of soluble adhesion molecules predicts death in predialysis patients: Association with malnutrition, inflammation and cardiovascular disease. Nephrol Dial Transplant 2000;15:1624–1630. 64 Bhagat K, Vallance P: Inflammatory cytokines impair endothelium-dependent dilatation in human veins in vivo. Circulation 1997;96: 3042–3047. 65 Sinisalo J, Paronen J, Mattila KJ, Syrjala M, Alfthan G, Palosuo T, et al: Relation of inflammation to vascular function in patients with coronary heart disease. Atherosclerosis 2000; 149:403–411. 66 Shimokawa K, Ito A, Fukumoto Y, Kadokami T, Nakaike R, Sakata M, et al: Chronic treatment with interleukin-1-beta induces coronary intimal lesions and vasospastic responses in pigs in vivo: The role of platelet-derived growth factor. J Clin Invest 1996;97:769–776. 67 Ridker PM, Hennekens CH, Roitman-Johnson B, Stampfer MJ, Allen J: Plasma concentrations of soluble intercellular adhesion molecule 1 and risk of future myocardial infarction in apparently healthy men. Lancet 1998;351:88– 92. 68 Ridker PM, Hennekens CH, Buring JE, Rifai N: C-reactive protein and other markers of inflammation in the prediction of cardiovascular disease in women. N Engl J Med 2000;342: 836–843. 69 Rabb H, Calderon E, Bittle PA, Ramirez G: Alterations in soluble intercellular adhesion molecule-1 and vascular cell adhesion molecule-1 in hemodialysis. Am J Kidney Dis 1996; 27:239–243. 70 Bonomini M, Reale M, Santarelli P, Stuard S, Settefrati N, Albertazzi A: Serum levels of soluble adhesion molecules in chronic renal failure and dialysis patients. Nephron 1998;79:399– 407. 71 Pober JS, Cotran RS: Cytokines and endothelial cell biology. Physiol Rev 1990;70:427– 451.

72 DeFronzo RA, Alvestrand A, Smith D, Hendler R, Hendler E, Wahren J: Insulin resistance in uremia. J Clin Invest 1981;67:563–568. 73 Mendall MA, Patel P, Ballam L, Strachan D, Nortfield TC: C-reactive protein and its relation to cardiovascular risk factors: A population-based cross-sectional study. BMJ 1996; 312:1061–1065. 74 Yudkin JS, Stehouwer CD, Emeis JJ, Coppack SW: C-reactive protein in healthy subjects: Associations with obesity, insulin resistance, and endothelial dysfunction: A potential role for cytokines originating from adipose tissue. Arterioscler Thromb Vasc Biol 1999;19:972– 978. 75 Visser M, Bouter LM, McQuillan GM, Wener MH, Harris TB: Elevated C-reactive protein levels in overweight and obese adults. JAMA 1999;282:2131–2135. 76 Festa A, D’Agostino R, Howard G, Mykkanen L, Tracy RP, Haffner SM: Chronic subclinical inflammation as part of the insulin resistance syndrome. Circulation 2000;102:42–47. 77 De Feo P, Volpi E, Lucidi P, Cruciani G, Reboldi G, Siepi D, et al: Physiological increments in plasma insulin concentrations have selective and different effects on synthesis of hepatic proteins in normal humans. Diabetes 1993;42:995–1002. 78 Campos SP, Baumann H: Insulin is a prominent modulator of the cytokine-stimulated expression of acute phase plasma protein genes. Mol Cell Biol 1992;12:1789–1797. 79 Halliwell B: The role of oxygen radicals in human disease with particular reference to the vascular system. Haemostasis 1993;23(suppl 1):118–126. 80 Maggi E, Bellazzi R, Gazo A, Seccia M, Bellomo G: Autoantibodies against oxidativelymodified LDL in uremic patients undergoing dialysis. Kidney Int 1994;46:869–876. 81 Engelmann B, Brautigam C, Kulschar R, Duhm J, Prenner E, Hermetter A, et al: Reversible reduction of phospholipid bound arachidonic acid after low density lipoprotein apheresis. Evidence for rapid incorporation of plasmalogen phosphatidylethanolamine into the red blood cell membrane. Biochim Biophys Acta 1994;1196:154–164. 82 Stenvinkel P, Holmberg I, Heimbürger O, Diczfalusy U: A study of plasmalogen as an index of oxidative stress in patients with chronic renal failure. Evidence of increased oxidative stress in malnourished patients. Nephrol Dial Transplant 1998;13:2594–2600. 83 Memon RA, Staprans I, Noor M, Holleran WM, Uchida Y, Moser AH, et al: Infection and inflammation induce LDL oxidation in vivo. Arterioscler Thromb Vasc Biol 2000;20:1536– 1542. 84 Nguyen-Khoa T, Massy ZA, De Bandt JP, Kebede M, Salama L, Lambrey G et al: Oxidative stress and haemodialysis: Role of inflammation and duration of dialysis treatment. Nephrol Dial Transplant 2000, in press. 85 Ridker PM, Rifai N, Pfeffer MA, Sacks F, Braunwald E: Long-term effects of pravastatin on plasma concentrations of C-reactive protein. Circulation 1999;100:230–235.

Blood Purif 2001;19:143–151

151

Blood Purif 2001;19:152–156

Contextual Issues in Comparing Outcomes and Care Processes for ESRD Patients around the World Donal Reddan a Lynda Anne Szczech a Peter J. Conlon b William F. Owen, Jr. a a Duke

Institute of Renal Outcomes Research & Health Policy, Duke University Medical Center, Durham, N.C., USA; Hospital, Dublin, Ireland

b Beaumont

Background Considerations

It is important to assess differences in outcomes between end-stage renal disease (ESRD) populations in different geographical regions, because disparities in outcomes may support the identification of processes of care that drive improved overall quality of care and outcomes. Such comparisons enable us to determine what components of case mix are critical determinants of survival and can also serve as gauges of the quality of care for ESRD programs across the globe. Significant differences in mortality outcome have been reported from various registries around the world. It has been consistently reported that mortality rates in the US population are worse than in Europe or Japan [1, 2]. Such reports have led to suggestions that inferior mortality rates may be related to inferior care [1, 3]. These suggestions are however based on the assumption that all such differences in outcomes are a function of care, rather than of biologic differences in populations. Herein we will examine the validity of the former assumption. There are significant differences in the scientific and statistical validity of reported mortality rates and their putative differences between different countries and reg-

ABC

© 2001 S. Karger AG, Basel 0253–5068/01/0192–0152$17.50/0

Fax + 41 61 306 12 34 E-Mail [email protected] www.karger.com

Accessible online at: www.karger.com/journals/bpu

istries. Confounding variables that compromise comparisons include: (1) different database design, (2) varying structure and financial support for renal replacement therapy, (3) demographic and genetic differences in the base populations, and (4) distinctions in healthcare delivery systems. Hence, it is difficult to precisely compare outcomes and processes of care between populations. In some situations it is virtually impossible to ‘control’ for all of the differences in population characteristics. Trying to correlate the impact of differences in processes of care on differences in outcomes is even more difficult as associations that apply in one population may not necessarily apply in another. International comparison of ESRD care is an example of one situation in which such difficulties apply. When comparing ESRD outcomes between populations, it is important that the contrasting populations share certain characteristics in order for comparisons to have validity. Ideally, data should have a similar source whether it is from a registry, billing data, chart extraction, or is prospectively gathered in an observation study or randomized controlled trial. When comparing data from registries, it is critical that reporting procedures are similar. Voluntarily reported data may have different characteris-

William F. Owen, Jr., MD Duke Institute of Renal Outcomes Research & Health Policy Box 3646, Duke University Medical Center Durham, NC 27710 (USA) Tel. +1 919 668 8008, Fax +1 919 668 7057, E-Mail [email protected]

tics too and should not be compared with data collections that are mandated. To minimize reporter bias, percent participation should at least be similar, and even when it is similar, assumptions of equivalence in data quality should be circumspect. Co-morbidity is an important consideration when comparing populations and always needs to be included in analyses. However, statistical strategies to control for co-morbidity in international comparisons can be hampered by differences in definitions of co-morbid states and degrees of severity. More subtly, differing practices of care may lead to differing detection rates for various co-morbidities and therefore further confound how these comorbidities influence analyses. Lastly, it is important that data used in comparisons comes from the same time period. Herein, we offer a critique that should discourage open comparisons between data reported from various registries around the world, because they differ significantly with regard to the characteristics listed in the previous paragraph. Reports from registries around the world include those from Australia and New Zealand [4], France [5], EDTA [6], Canada [7], Japan [8], Poland [9], and others.

Evidence of Differences in Outcomes

Held et al. [2] in 1990 compared the 5-year survival of patients in three registries: The European Dialysis and Transplantation Association (EDTA) Registry, Japanese Society of Dialysis Therapy and US data from the Health Care Financing Administration (HCFA). For all ESRD patients, Held et al. reported a 5-year survival of 59% for EDTA patients, 61% for Japanese patients, and 40% for US patients. When adjusted for age and proportion of patients with diabetic nephropathy in the US population, the European and Japanese 5-year survival figures became 48% and 54% respectively. After adjusting for age and the proportion of patients with diabetic nephropathy (27% in the USA, 19% in Japan and 10% in Europe) overall mortality remained higher in the USA. The relative risk of death for US patients was 1.15 when compared to Europeans and 1.33 when compared to the Japanese. The differences in mortality were different for different age groups however. The !25 age group had a superior mortality profile in the USA, whereas the age cohort 25–55 years old had a 20% higher and the 655-year-olds had a 30–35% higher mortality rate than their European counterparts.

Outcomes and Care Processes for ESRD Patients

This comparison has several substantial limitations that illustrate the difficulty of such analyses. Initially, it is apparent that reporting patterns differed significantly between registries. The EDTA-ERA database has two annual collection instruments, one of which is center-specific and the other patient-specific. Response rates across different countries are extremely variable with an overall response rate of 67% for the center-specific questionnaire [10], and 60% for the patient-specific questionnaire. Among the three European countries with the largest dialysis populations, France, Germany and Italy, 1994 center questionnaire response rates were 84, 57 and 56%, respectively, and patient questionnaire response rates 74, 50 and 49%, respectively [10]. The response rate has improved recently and was 82.2% in 1995 [11]. The Japanese registry for dialysis therapy has been collecting data annually since 1983. Completion rates for annual questionnaires in 1992 were reported as 99% for facility-specific data and 95% for patient-specific data [8]. However, in a paper discussing long-term survival of a 1993 cohort of dialysis patients, it is notable that Shinzato et al. [12] excluded 120% of patients from the analysis secondary to insufficient data. Of the 131,492 patients on whom initial information was acquired, only 56,431 (43%) of the original cohort were actually included. The reported annual mortality was 5.8% This first comparison of international outcomes is also limited by: (1) the absence of terms assessing the difference in co-morbidity or etiology of ESRD between cohorts, (2) inability to account for significant racial and cultural differences between populations which may result in differences in health-related behaviors, and (3) patients selected from each registry were from different years. In 1996, Marcelli et al. [1] compared data from two registries, the Lombardy Dialysis and Transplant Registry (RLDT) and the US Renal Data System (USRDS). 2,900 white patients from the USRDS Case Mix Severity Study [13] and all 1,296 from RLDT who started renal replacement therapy in 1986 and 1987 were studied. Compared to Lombardy patients, those in the USA were significantly older (mean age 59.9 B 16.4 vs. 55.9 B 14.7 years), had a lower proportion of males (53.7% vs. 62.1%), a greater proportion with diabetic nephropathy (29.9% vs. 9.7%) and a significantly greater proportion of patients with reported co-morbid conditions (heart disease, peripheral vascular disease, cirrhosis, cachexia, malignancy). US patients were less frequently treated with peritoneal dialysis (PD) by day 30 of ESRD (21.2% vs. 30.7%). Overall, 48% of the 4,196 patients died during the 48- to 72-month follow-up to December 31, 1991. Per 100 patient-years the

Blood Purif 2001;19:152–156

153

gross death rate for USRDS patients was 28.7 compared to 13.0 for RLDT patients. The unadjusted relative risk of death for RLDT was 0.439 compared to USRDS patients. The observed lower mortality risk in Lombardy was less pronounced when adjusted for demographic and co-morbid factors. Each of the recorded co-morbid conditions and treatment modality was significantly related to survival. The Cox cumulative survival adjusted for all these explanatory co-variates was 84.4% for US patients at 1 year, 67.0% at 2 years and 33.4% at 5 years, and for RLDT patients 88.3% at 1 year, 75.9% at 2 years and 45.9% at 5 years. The relative mortality risk (RR) for the patients treated in Lombardy adjusted for all the reported co-variates was 29% lower than for US patients (RR = 0.71). This comparative risk varied significantly by age and was 65% lower for Lombardy compared to US patients in the age range 25–44 years (RR = 0.35) and about 20% lower for patients over age 65 years (RR = 0.80). This relative risk was mainly related to hemodialysis and was not statistically significant for PD patients. Although Marcelli’s paper is perhaps the most rigorous printed comparison of international registries, it again presents illustrative limitations that compromise comparisons of outcomes across data sets like registries. A 100% response rate is reported for the RLDT [1, 14]. However, validation procedures are not reported in either paper, and although this is quite a comprehensive database, it cannot be presumed to be free of reporting bias. The voluntary nature of this and of all of the EDTA databases predisposes toward reporting bias. The Lombardy patients selected for Marcelli’s study were all incident, white ESRD patients during 1986 and 1987. The compared US population was selected from the case mix severity special study for 1989 [15]. This comparative study was not merely a comparison between countries, but also a comparison in time. Adjusted death rates for the USRDS population have dropped steadily over the years, and this time-dependent improvement in mortality rate is not accounted for in the analysis.

Nonactionable Factors and Patients’ Outcomes

There are significant differences in demographics and other nonactionable outcome determinants between populations. The American ESRD Program has the highest incident treatment rate in the world, with the greatest growth occurring in patients 175 years old. The average rate of increased incidence in the cohort older than 65 was 14% per year for the years 1988–1992 and 8% per year for

154

Blood Purif 2001;19:152–156

the years 1992–1996 [16]. The theory that a selection bias exists for patients with poorer prognoses and greater morbidity is a tenable one. Until there is a consensus rationale explaining the higher US incident treatment rates, and an appropriate surrogate for this difference identified, it will not be possible to accurately control for differences in the compared populations in studies such as these. Older patients arguably have greater co-morbidity and this may not be accounted for in statistical outcome models. Functional impairment and quality of life have both been independently linked to increased mortality. After controlling for other co-variates including age, race, sex, primary cause of renal failure and the presence of co-morbidity, both the Karnofsky performance score (KPS) and the Spitzer Quality of Life Index (SQLI) scores have been found to independently correlate with risk of mortality [17]. Noninclusion of these and other co-variates representative of co-morbidity in the statistical models used for comparisons of outcomes between countries could account for some of the differences. The prevalence of functioning renal transplant as a modality of renal replacement therapy in 1997 was 28.1% in the USA [16]. This has remained relatively stable since 1988. Prevalence of functioning transplants varies considerably amongst and regionally within the registries against which the USA is compared. In 1992 the percentage of Lombardy RRT patients with functioning renal transplants was 21% and in 1989, 18% [14]. Transplantation is usually performed on a selected population with less co-morbidity [18, 19]. The relatively higher transplantation rate in the USA leads to selection of older sicker patients for dialysis. Reported etiology of ESRD differs considerably between countries. The commonest etiology in the USA is diabetes mellitus (33.2%), followed by hypertension (24%) [16]. In Europe, glomerulopathy is listed as the cause in 21% of the patients in Germany and Italy and 22% in France. Diabetes varies from 10% in Germany to 22% in France [6]. In the RLDT/USRDS paper the difference in prevalence of diabetics was striking between the two groups with 30% of US patients diabetics versus 10% of RLDT. When the analysis was modified to only include diabetics, there was no significant difference in outcome between groups. In the corresponding analysis limited to nondiabetics there was a significant difference in outcome. Diabetes may therefore be a surrogate for prevalence of co-morbidity in the USRDS population. Other nonactionable factors that are unaccounted for in this and other comparisons include smoking, noncompliance with the dialysis prescription and other deleterious health-related behaviors, and voluntary withdrawal

Reddan/Szczech/Conlon/Owen

from dialysis. Approximately 1 in 5 dialysis patients in the USA (24% in the 164 age group) withdraw from dialysis prior to death [20]. The link between mortality and poor quality of life is also relevant [17]. Poor quality of life and voluntary withdrawal from dialysis are surrogates of morbidity. Neither has been controlled for in prior international comparisons. Recent data demonstrates that USA dialysis patients are most likely to terminate their dialysis treatments prematurely, and F15% of US ESRD patients routinely skip hemodialysis treatments. Similar findings have been impugned for PD patients in the US, in comparison to other areas of the world. McClellan et al. [21] have demonstrated significant variability in characteristics in a regional US population. CHF prevalence across 213 dialysis centers had an interquartile range of 6.9–26.7%. There was also quite a large distribution of adjusted mortality across centers. Such variability between centers illustrates one of the potential problems in comparing data from a small region with that of a larger area. RLDT in 1989 consisted of 44 dialysis centers; USRDS in 1992 received reports from 2,086 centers (this number has since increased to 3,093 in 1999 [16]. It is highly likely that individual comparisons between different US geographical regions and Lombardy might lead to a spectrum of possible results. The overall health of the general population from which the incident ESRD population is drawn also needs to be accounted for in models. Coronary heart disease (CHD) mortality differs significantly between the general populations of the countries studied. Verschuren et al. [22] in the Seven Countries Study demonstrated that the age-standardized 25-year CHD mortality rate was 20.3% in Northern Europe, 16% in the USA, 4.7% in Southern Europe (Mediterranean) and 3.2% in Japan. As cardiovascular disease is the leading cause of death in the ESRD population, these differences in CHD mortality in the general population may be exaggerated in the ESRD population. This increase in prevalence of cardiovascular mortality is another surrogate of morbidity that is not controlled for in this analysis. Moreover, it is likely that detection and treatment rates for CHD differ substantially in various countries and within populations for a given country. It is noteworthy that in the Lombardy study, USRDS cardiovascular mortality was 5 times greater than that in RLDT patients [1]. Healthcare financing priorities also differ significantly throughout the world. 1996 OECD data reveals that US health expenditure expressed as a percentage of GDP was 13.6% [23]. This has increased from 10.8% in 1986. OECD data from other countries is significantly different

Outcomes and Care Processes for ESRD Patients

to that from the US healthcare expenditures, as a percentage of GDP in Italy was 7.8% in 1996 and 7.0% for 1986. Such differences in expenditure could be a surrogate for different thresholds for renal replacement therapy and therefore may suggest further unaccounted for demographic differences between ESRD populations.

Outcomes and Processes of Clinical Care

A number of actionable candidate factors, including dialysis dose and time, reuse practices, hematocrit, and nutritional support have been proposed that influence outcomes and could therefore influence international comparisons. Though differences in delivered care sometimes appear quite striking, it is significant that in most population-based, multivariate models of mortality outcomes, the actionable measures of patient care, such as dialysis dose, hematocrit, nutritional support, frequency of physician interaction with the patients, etc., account for !20% of the differences in patient mortality. The survival impact of these differences is therefore statistically quite small, and the likelihood is that differences in outcome are more likely to be explained by nonactionable variables. ‘For international studies, the development of common standard data collection instruments are needed in all ESRD Registries, including the recording of comorbid conditions and their severity ... also bear in mind patient characteristics.’ This quote from Locatelli et al. [3] offers the fairest statement of future needs. Towards addressing the scientific needs posed by this issue, an international, multicenter, prospective study is in progress. Entitled the International Dialysis Outcomes & Practices Study (IDOPS), this cohort analysis is examining the relationship between physician and nursing practices in the care of ESRD patients and their outcomes. The study uses uniform data collection tools, which account for differences in patient co-morbidity and health-related behaviors. As for augmenting patient outcomes in the USA, proof of the renal communities response is evident in HCFA’s 1998 ESRD Core Indicators Report [24]. Improvement has been observed in the delivered dose of hemodialysis, anemia correction, and maintenance of adequate iron stores. We foresee continued improvement, if financial and intellectual support for this initiative continues through Congress, HCFA, and the ESRD Networks. In summary, the purported differences in mortality between ESRD patients in the USA and other countries are not yet explained and may be more related to differ-

Blood Purif 2001;19:152–156

155

ences in incident population demographics and co-morbidity than to care differences. The controversy provoked by suggestions of better non-US outcomes has had the beneficial effect of stimulating ongoing and ever more vigilant continuous quality improvement in the USA, and

such improvement is always beneficial. IDOPS will reveal if there are indeed care differences that significantly impact on outcome. The question as to whether superior ESRD care in Europe explains superior clinical outcomes is a difficult question that remains unanswered.

References 1 Marcelli D, Stannard D, Conte F, Held PJ, Locatelli F, Port FK: ESRD patient mortality with adjustment for comorbid conditions in Lombardy (Italy) versus the United States. Kidney Int 1996;50:1013. 2 Held PJ, Brunner F, Odaka M, Garcia JR, Port FK, Gaylin DS: Five-year survival for endstage renal disease patients in the United States, Europe, and Japan, 1982 to 1987. Am J Kidney Dis 1990;15:451. 3 Locatelli F, Marcelli D, Conte F: Editorial: Dialysis patient outcomes in Europe vs the USA. Why do Europeans live longer? Nephrol Dial Transplant 1997;12:1816. 4 Disney AP: Demography and survival of patients receiving treatment for chronic renal failure in Australia and New Zealand: Report on dialysis and renal transplantation treatment from the Australia and New Zealand Dialysis and Transplant Registry. Am J Kidney Dis 1995;25:165. 5 Jacobs C, Selwood NH: Renal replacement therapy for end-stage renal failure in France: Current status and evolutive trends over the last decade. Am J Kidney Dis 1995;25:188. 6 Brunner FP, Selwood NH: Profile of patients on RRT in Europe and death rates due to major causes of death groups. The EDTA Registration Committee. Kidney Int Suppl 1992; 38:S4. 7 Fenton S, Desmeules M, Copleston P, Arbus G, Froment D, Jeffery J, Kjellstrand C: Renal replacement therapy in Canada: A report from the Canadian Organ Replacement Register. Am J Kidney Dis 1995;25:134. 8 Teraoka S, Toma H, Nihei H, Ota K, Babazono T, Ishikawa I, Shinoda A, Maeda K, Koshikawa S, Takahashi T, et al: Current status of renal replacement therapy in Japan. Am J Kidney Dis 1995;25:151.

156

9 Rutkowski B, Puka J, Lao M, Baczyk K, Chrzanowski W, Kokot F, Ksiazek A, Nartowicz E, Poplawski A, Sulowicz W, Szewczyk Z: Renal replacement therapy in an era of socioeconomic changes Report from the Polish Registry. Nephrol Dial Transplant 1997;12:1105. 10 Mallick NP, Jones E, Selwood N: The European (European Dialysis and Transplantation Association-European Renal Association) Registry. Am J Kidney Dis 1995;25:176. 11 Briggs JD, EDTA: Personal communication, 1999. 12 Shinzato T, Nakai S, Akiba T, Yamazaki C, Sasaki R, Kitaoka T, Kubo K, Shinoda T, Kurokawa K, Marumo F, Sato T, Maeda K: Survival in long-term haemodialysis patients: Results from the annual survey of the Japanese Society for Dialysis Therapy. [Corrected and republished article originally printed in Nephrol Dial Transplant 1996;11:2139–42]. Nephrol Dial Transplant 1997;12:884. 13 Blagg CR: The US Renal Data System and the Case-Mix Severity Study. Am J Kidney Dis 1993;21:106. 14 Locatelli F, Marcelli D, Conte F, Limido A, Lonati F, Malberti F, Spotti D: 1983 to 1992: Report on regular dialysis and transplantation in Lombardy. Am J Kidney Dis 1995;25:196. 15 Held PJ, Port FK, Webb RL, Wolfe RA, Bloembergen WE, Turenne MN, Holzman E, Ojo AO, Young EW, Mauger EA, et al: Excerpts from United States Renal Data System 1995 Annual Data Report. Am J Kidney Dis 1995;26:S1.

Blood Purif 2001;19:152–156

16 Excerpts from United States Renal Data System 1999 Annual Data Report. Am J Kidney Dis 1999;34:S1. 17 McClellan WM, Anson C, Birkeli K, Tuttle E: Functional status and quality of life: Predictors of early mortality among patients entering treatment for end-stage renal disease. J Clin Epidemiol 1991;44:83. 18 Gaylin DS, Held PJ, Port FK, Hunsicker LG, Wolfe RA, Kahan BD, Jones CA, Agodoa LY: The impact of comorbid and sociodemographic factors on access to renal transplantation. JAMA 1993;269:603. 19 Port FK, Wolfe RA, Mauger EA, Berling DP, Jiang K: Comparison of survival probabilities for dialysis patients vs. cadaveric renal transplant recipients. JAMA 1993;270:1339. 20 Excerpts from the United States Renal Data System 1998 Annual Data Report. Am J Kidney Dis 1998;32:S1. 21 McClellan WM, Soucie JM, Flanders WD: Mortality in end-stage renal disease is associated with facility-to-facility differences in adequacy of hemodialysis. J Am Soc Nephrol 1998;9:1940. 22 Verschuren WM, Jacobs DR, Bloemberg BP, Kromhout D, Menotti A, Aravanis C, Blackburn H, Buzina R, Dontas AS, Fidanza F, et al: Serum total cholesterol and long-term coronary heart disease mortality in different cultures. Twenty-five-year follow-up of the seven countries study. JAMA 1995;274:131. 23 OECD in Figures, 1999. 24 Administration HCF: 1998 Annual Report, End-Stage Renal Disease. Core Indicators Project. Department of Health and Human Services, Health Care Financing Administration, Office of Clinical Standards and Quality. Baltimore, Md, Dec 1997–1998.

Reddan/Szczech/Conlon/Owen

Blood Purif 2001;19:157–167

Should the Hematocrit Be Normalized in Dialysis and in Pre-ESRD Patients? Iain C. Macdougall Department of Renal Medicine, King’s College Hospital, London, UK

Introduction

Why Do We Aim for Subnormal Hematocrit Levels?

Just over a decade ago, a new therapeutic agent was introduced that could effectively treat the anemia associated with chronic renal failure. This treatment, called recombinant human erythropoietin (epoetin), was rapidly shown to be successful in 90–95% of patients treated. It could increase the hematocrit (or hemoglobin – see below) to whatever level the physician desired, and yet the early treatment studies elected to aim for partial rather than full correction of the anemia. This practice has largely persisted, with few nephrologists currently aiming for normalization of hematocrit in their patients. The first question we have to ask ourselves is why this is the case. Indeed, the appropriate target hematocrit for dialysis patients has arguably been one of the most controversial and debated issues in nephrological practice throughout the last decade [1, 2]. Attempts to provide guidelines for nephrologists have suggested a hematocrit of 33–36% (hemoglobin 11–12 g/dl) in the NKF-DOQI Guidelines [3], and a hemoglobin of 111 g/dl (with no upper limit specified) in the European Best Practice Guidelines [4]. And yet most of us involved in the creation of such guidelines would have to confess that these target levels are somewhat arbitrary, based on the limited amount of scientific evidence available in the literature. What is clear is that the normal (and physiological) levels of hematocrit present in healthy non-uremic individuals are not a common finding in renal patients, even with the ‘technology’ available to achieve this.

ABC

© 2001 S. Karger AG, Basel 0253–5068/01/0192–0157$17.50/0

Fax + 41 61 306 12 34 E-Mail [email protected] www.karger.com

Accessible online at: www.karger.com/journals/bpu

There are several possible reasons for this. The first is a historical one in that the first two pivotal studies conducted on either side of the Atlantic aimed for subnormal correction of the anemia [5, 6]. Second, there has been some concern that full correction of anemia in renal failure patients may expose such individuals to an increased risk of developing adverse events, such as hypertension and vascular access thrombosis, or even (in the early studies) hypertensive encephalopathy or seizures. Third, the largest study ever set up to investigate this issue (the US Normal Hematocrit Study) failed to live up to the expectation that dialysis patients randomized to a normal hematocrit would have a better outcome than those aiming for a conventional (subnormal) hematocrit [7]. Fourthly, in the current climate of evidence-based medicine, the purist nephrologist might argue that there is no direct evidence that targeting a normal hematocrit in patients receiving epoetin results in a significant improvement in morbidity and mortality compared with partial correction of the anemia. Finally, there is an economic issue. At the present time, epoetin therapy remains a fairly costly treatment; many units have large numbers of patients on epoetin, and the treatment is long-term, i.e. usually until the patient dies or receives a renal transplant. As will be described later, the cost of fully correcting anemia in dialysis patients is considerably greater than aiming for partial correction.

Dr. Iain C. Macdougall Renal Unit, King’s College Hospital East Dulwich Grove London SE22 8PT (UK) Tel. +44 207 346 6234, Fax +44 207 346 6472, E-Mail [email protected]

Hematocrit or Hemoglobin?

Before discussing the allocated topic, it is perhaps worth spending a moment or two reflecting on the choice of hematocrit or hemoglobin as a measure of anemia. For good reasons, the European Best Practice Guidelines have recommended the hemoglobin concentration to monitor anemia [4]. Measurement of hematocrit is extremely variable, dependent on the method used, and dependent on the age of the sample when analyzed. The microhematocrit method can yield different results from measurement of hematocrit by commercially-available automated blood count analyzers, and one automated analyzer may give a different result from another. Furthermore, there is no international standard for hematocrit measurement as there is for measurement of hemoglobin, and indeed most hematology laboratories perform daily quality control on their automated blood count analyzers using a known standard. Most units outside the USA use hemoglobin concentration, including their Canadian neighbors, and it is to be hoped that Americans will soon take heed of these facts and do likewise. Unfortunately, this has clearly not happened for the current meeting debate which still has ‘hematocrit’ in the title. To promote the trend, however, I shall use the term ‘hemoglobin’ to describe the degree of anemia throughout the remainder of this article.

Should Hemoglobin Be Normalized in Dialysis and in Pre-ESRD Patients?

This is the question to be addressed in the current debate. The first issue is whether this question is equally appropriate for dialysis and for predialysis patients. There are certainly factors which are different in these two patient populations, and these are sufficiently diverse to suggest that this topic should be discussed separately.

Normalization of Hemoglobin in Dialysis Patients

There are clearly arguments for and against this motion based on the current literature available. My remit is to discuss factors which would argue against normalizing hemoglobin in this patient population. The first point to make is that one has to consider each patient as an individual rather than as one of a population. Thus, it may not be appropriate to target the same hemoglobin for every dialysis patient in the same unit. To illustrate this point, I

158

Blood Purif 2001;19:157–167

would like to present 2 completely contrasting fictitious cases which have been deliberately selected to include factors which may influence the choice of target hemoglobin. Case 1 is a 26-year-old man who recently reached endstage renal failure from chronic glomerulonephritis. He has been on automated peritoneal dialysis for 3 months, and his hemoglobin at the time of starting dialysis was 9.2 g/dl, having been 11 g/dl 4 months earlier. He was started on epoetin therapy, and his hemoglobin was gradually increasing. He had borderline hypertension, easily controlled by a long-acting calcium antagonist, but he was otherwise very well with no other co-morbid conditions. He works as a builder, and until recently was captain of the local rugby team. Unfortunately, his level of fitness had deteriorated to an extent that he could no longer compete at the level required of him. He had undergone a number of cardiac investigations as part of a workup for renal transplantation; he had a normal exercise treadmill test with no signs of myocardial ischemia, and a normal echocardiogram with an ejection fraction of 56% and no evidence of left ventricular hypertrophy. What should his target hemoglobin on epoetin be? Case 2 is a frail 74-year-old lady with diabetic nephropathy who has been on unit-based hemodialysis for 13 years. She was heavily transfusion-dependent when she started dialysis, but this ceased when she commenced epoetin therapy in 1992. She did, however, also suffer from active sero-negative rheumatoid arthritis over the years, and this had resulted in some resistance to her epoetin, with hemoglobin levels ranging from 7.2 to 9.9 g/ dl over the last 8 years. She had suffered two myocardial infarctions in 1994 and 1997, and this had left her with a moderately severe ischemic cardiomyopathy, causing NYHA Grade III heart failure. She had two previous failed vascular access episodes, with a thrombosis of her left radial fistula in 1996, and a thrombosed left brachial fistula in 1999. She was currently dialyzing satisfactorily via a right brachial PTFE graft. Because of her arthritis, she was largely wheelchair-bound and often breathless at rest due to her heart failure. Of late, she was becoming somewhat forgetful and intermittently confused, although she did have a supportive and caring husband. A CT scan of her brain showed changes of diffuse atherosclerotic cerebrovascular disease, but with no discreet lesion. Her family ask you what level of hemoglobin you are aiming for. What do you answer? In brief, Case 1 is a fit young man with no significant co-morbidity who has a physically demanding job and an active lifestyle, whereas Case 2 is a frail elderly lady with

Macdougall

multiple other medical problems, including significant cardiac disease and a limited life expectancy. In real life, one has to select a target hemoglobin for each patient, but should this be the same for both patients or should it be individualized? Having a unit policy on target hemoglobin, or devising clinical guidelines on this issue, tends to consider a total population, whereas this population is made up of many different individuals who may have very different characteristics or clinical features. With reference to the 2 cases described above, one could offer an argument for normalizing hemoglobin in the first patient, whereas the scientific evidence [7] suggests a note of caution in doing the same for Case 2. Before discussing the disadvantages of normalizing hemoglobin, let us consider in turn some of the factors which may influence the choice of target hemoglobin in an individual patient (table 1). Age There are no studies specifically looking at the choice of target hemoglobin in the elderly versus the younger population. We do know that elderly patients benefit from epoetin therapy both with regard to their exercise capacity [8] and cardiac function [9]. There are, however, two reasons why the nephrologist uses epoetin therapy: firstly for a fairly rapid improvement in anemic symptoms, exercise capacity, and quality-of-life, and secondly to improve cardiac function along with (hopefully) long-term survival. Whilst the first of these criteria is applicable to patients of all ages, the potential impact of the latter phenomenon is clearly much greater in younger patients than in the frail elderly patient whose life expectancy is less than two years. More aggressive anemia management may therefore be more appropriate in the younger patient. Gender Again, few studies have differentiated between males and females in terms of target hemoglobin. Interestingly, the Scandinavian Multicentre Study [10, 11] did aim for a higher target hemoglobin in male patients (14.5–16.0 g/ dl) compared to females (12.5–14.0 g/dl), but there was no gender difference in either the US Normal Hematocrit Trial [7] or the Canadian Multicentre Study [12]. In healthy individuals, there is a higher physiological hemoglobin in males compared to females, although this becomes less marked when the latter become postmenopausal. Some might argue that this physiological difference between males and females should be maintained in renal failure patients; others might suggest that most female dialysis patients are post-menopausal, either due

Normalization of Hematocrit in Renal Patients

Table 1. Factors affecting choice of target hemoglobin in an individual patient

Age Gender Occupation Level of physical activity Length of time with chronic renal failure/renal anemia Starting hemoglobin/length of time on epoetin Co-morbid conditions Dialysis modality

to age or to biological factors causing premature menopause in uremic patients. If one were to select a target hemoglobin for a fit and healthy male hemodialysis patient aged 25 years versus a fit and healthy female dialysis patient aged 25 years who is still menstruating, then there may be some rationale in selecting a slightly lower target hemoglobin concentration for the latter patient. Much of this discussion is, however, speculative with little in the way of scientific data to support it. Occupation If a young patient is in full-time employment, and particularly if the job involves fairly strenuous physical activity, then there may be a rationale for maximizing exercise capacity in these patients. There is indeed evidence that physical capacity and maximum oxygen consumption is greater in dialysis patients with a hemoglobin concentration of around 14 g/dl compared with one around 10 g/dl [8]. This is true for both young and old patients. Level of Physical Activity For the same reasons as indicated in the previous section, a patient with a fairly active lifestyle involving much physical activity (including sport) would benefit from a higher rather than a lower hemoglobin. Length of Time with Chronic Renal Failure/Renal Anemia It is well known that physiological homeostatic mechanisms, such as altered diphosphoglycerate levels (causing a shift in the oxygen dissociation curve), come into play during chronic anemia. It is these compensatory mechanisms that allow a patient with homozygous sickle cell disease to function fairly well with a hemoglobin concentration of around 5–6 g/dl between blood transfusions. The length of time a patient has suffered from renal anemia, and indeed the severity of this condition, may influence

Blood Purif 2001;19:157–167

159

the potency and reversibility of these compensatory mechanisms. Thus, if a patient has been exposed to a hemoglobin of between 6 and 8 g/dl for several years, homeostatic mechanisms may have become so established that to bring the hemoglobin concentration up to 14 g/dl in the space of a few months might be deleterious. It may in fact result in ‘relative’ polycythemia for such patients, and this may also explain why severely anemic patients previously developed hypertensive encephalopathy or seizures even with a suboptimal hemoglobin. Conversely, if a patient has been exposed to anemia for a short time (as illustrated in Case 1) then they may be better equipped to deal with a normal physiological hemoglobin level. Starting Hemoglobin/Length of Time on Epoetin For the same reasons as outlined in the previous section, both the starting hemoglobin concentration and the length of time the patient has been on epoetin therapy, may influence how well the patient can cope with normalization of hemoglobin. Thus, if the anemia has been mild and the patient has been on epoetin for several years, then the further increment in increasing their hemoglobin to normal might be achieved relatively easily. Conversely, if the hemoglobin has been increased fairly rapidly from around 7 g/dl up to 11 g/dl with epoetin, then a further increase to 14 g/dl might be more hazardous. Co-Morbid Conditions Conditions that may influence the choice of target hemoglobin might include cardiac disease, cerebrovascular disease, and other arteriopathic conditions, diabetes mellitus, chronic obstructive pulmonary disease, and precarious or precious vascular access (particularly if there has been a previous thrombosis). As will be discussed in more detail below, patients with known ischemic heart disease or cardiac failure may be no better (or even worse off) with normalization of their hemoglobin compared to partial correction of anemia [7]. Although this study may have its limitations [13], one cannot ignore the findings of this large randomized controlled trial, one of the major conclusions of which was that patients with cardiac disease should not have their hemoglobin normalized until any further data become available. There have been studies of epoetin therapy in diabetic patients, although no useful information has appeared on what the optimal target hemoglobin is in such patients. In view of the known microvascular disease occurring in diabetics, however, there is a rheological argument for running such patients with a slightly lower hemoglobin

160

Blood Purif 2001;19:157–167

to reduce whole blood viscosity and improve red cell fluidity. Nonuremic patients with chronic obstructive pulmonary disease often develop a secondary polycythemia to compensate for chronic hypoxia. Such patients often feel less breathless with a higher hemoglobin, and there is no reason why the same rationale should not be applied to renal failure patients. There is indeed a peritoneal dialysis patient in our unit who has chronic bronchiolitis due to rejection in his heart-lung transplant, and who feels symptomatically less breathless at a hemoglobin of 14 g/dl than when it is reduced to 12 g/dl. Finally, there is some evidence from controlled studies that a higher hemoglobin concentration may exacerbate the risk of a vascular access thrombosis [7, 14]. The incidence of access thrombosis was higher in the initial Canadian Multicentre Study [14] in the group of patients randomized to a hemoglobin of 11.5–13.0 g/dl (7 of 38 patients) compared with those randomized to the lower hemoglobin group (9.5–11.0 g/dl; 4 of 40 patients). Similarly in the US Normal Hematocrit Trial there was an excess of vascular access thrombosis in the patients assigned to the normal hematocrit group compared to those remaining on the lower (conventional) hematocrit (39 vs. 29%; p = 0.001) [7]. The implication of these data may seem somewhat obvious, but patients with a history of vascular access complications, or those with a poor quality access, or those in whom limited sites remain for further vascular access, should probably not aim for complete correction of anemia unless there are compelling reasons to do so. Dialysis Modality As already discussed, there may be a case for targeting a different hemoglobin concentration in a pre-dialysis compared to a dialysis patient. This will be discussed further later on. The other issue to consider is whether the same target hemoglobin concentration should be used in hemodialysis compared with peritoneal dialysis patients. There is one important point in this respect, which is that the hemoglobin concentration in peritoneal dialysis patients is generally much more stable than in hemodialysis patients. Data shown below indicate that the hemoglobin concentration can increase by 1–3 g/dl across a dialysis session, depending on how much fluid is removed (fig. 1). Thus there might be less concern in maintaining a peritoneal dialysis patient with a hemoglobin of 14 g/dl compared with a hemodialysis patient whose predialysis hemoglobin was 14 g/dl and whose post-dialysis hemoglobin was 17 g/dl.

Macdougall

Fig. 1. Change in hemoglobin concentration across dialysis (post-dialysis minus pre-dialysis sample) in relation to weight loss in a cohort of hemodialysis patients.

Why Should Hemoglobin Not Be Normalized in Dialysis Patients?

At a simplistic level, one would instinctively like to normalize hemoglobin in all dialysis patients. The reasons why this is not happening in routine clinical practice have already been discussed, but there are also some fairly compelling scientific arguments to support this practice (table 2). These will be discussed in turn: No Compelling Evidence That a Higher Hemoglobin Is Significantly Better There are several studies systematically examining the effect of normalizing hemoglobin in dialysis patients with respect to various outcomes. Some of these are small in number [8, 15], although there are also three large multicenter studies in the USA [7], Scandinavia [10, 11], and Canada [12]. The initial report of normalizing hematocrit in 13 hemodialysis patients by Eschbach et al. [15] suggested that there were modest improvements in qualityof-life, exercise capacity, and a further reduction in left ventricular hypertrophy. There was, however, no control group in this study. Two more recent studies by McMahon et al. [8, 16] on the same cohort of hemodialysis patients have shown modest improvements in exercise capacity, quality-of-life, maximum oxygen consumption, and cardiac parameters at a hemoglobin of 14 g/dl compared to 10 g/dl, but with an enormous additional cost as discussed in the next section. The US Normal Hematocrit Trial [7] set out with a hypothesis that normalizing hemoglobin in high-risk cardiac patients on regular hemodialy-

Normalization of Hematocrit in Renal Patients

Table 2. Arguments for not normalizing hemoglobin in dialysis

patients No compelling evidence that a higher hemoglobin is significantly better Normalization of hemoglobin greatly increases epoetin dose requirements and cost Requirements for IV iron supplementation greater Normalization of hemoglobin increases the risk of vascular access thrombosis Normalization of hemoglobin results in less efficient dialysis Normalization of hemoglobin increases the risk of dialyzer clotting Hemoglobin concentration increases significantly across dialysis Hemoglobin fluctuates significantly in dialysis patients on epoetin

sis would improve outcome in terms of the two primary endpoints (death and first non-fatal myocardial infarction). Although the follow-up was intended to be 3 years, the study was aborted prematurely after 29 months, at which time there was no benefit shown for the higher hemoglobin (and indeed it was at the brink of showing a worse outcome). The Data Monitoring Group made it clear that even if the study were to continue to its natural end it could not show a positive benefit. At best, there would be no difference between the two groups, and at worst, the normalized hemoglobin group could end up with a worse outcome. This must be the strongest piece of evidence yet for not introducing a global normalization of hemoglobin policy across all dialysis units. The Scandinavian [10, 11] and Canadian [12] Multicentre Studies also

Blood Purif 2001;19:157–167

161

Fig. 2. Relative mortality risk in relation to

hematocrit in a US observational study conducted in 1992 and 1993 in 75,283 hemodialysis patients. p values are calculated versus a hematocrit of 30–33%. Data from Ma et al. [18].

Fig. 3. Relationship between hemoglobin

and Sickness Impact Profile score as a measure of quality-of-life in 1,013 dialysis patients. Data from Moreno et al. [20].

did not live up to full expectations. In the Scandinavian Study, while there was no excess mortality seen in the normalized hemoglobin group, the cardiac benefits were less marked than many had expected [10]. This was also true in the Canadian Multicentre Study which examined specifically the changes in left ventricular mass and cavity volume in hemodialysis patients randomized to either a normal hemoglobin group or a subnormal hemoglobin group. Although both groups showed some reduction in left ventricular mass, there was no reduction in left ventricular cavity volume in patients with left ventricular dilatation, even when the hemoglobin was normalized. The major benefit seen was in the patients with a normal left ventricular cavity volume at baseline, in whom full correction of anemia prevented this worsening [12]. In

162

Blood Purif 2001;19:157–167

summary, although there have been some improvements in quality-of-life [17], exercise capacity [8] and cardiac parameters [10, 12] with normalization of hemoglobin, many of these are less marked than would have been anticipated 5 years ago. There are also no controlled studies to suggest that increasing the hemoglobin to normal with epoetin therapy in dialysis patients improves long-term morbidity and mortality. There are, however, some epidemiological data showing correlations between the degree of anemia and mortality [18], length of hospitalizations [19], and quality-of-life [20]. Although not specifically related to epoetin therapy, the data do suggest that dialysis patients with more profound anemia have an increased mortality rate, longer duration of hospital admission, and lower quality-of-life scores. If one examines

Macdougall

Fig. 4. Mean monthly epoetin doses in the US Normal Hematocrit Trial of 1,233 hemodialysis patients. There was a significant difference in the mean doses between the two groups from 1 month onwards (p ! 0.001). Error bars indicate 95% confidence intervals of the mean.

these data closely, however, it can be seen that the effect is beginning to plateau at around a hematocrit of 30–33%, with a marginal additional benefit in the higher hematocrit groups (fig. 2, 3). Normalization of Hemoglobin Greatly Increases Epoetin Dose Requirements and Cost All of the studies examining normalization of hemoglobin have shown quite considerable increases in epoetin dose requirements attempting to drive the patients’ hemoglobin concentrations into the normal range. The first study by Eschbach et al [15] showed a 69% increase in epoetin dose, and the largest study of its kind (the US Normal Hematocrit Trial) showed that patients required a threefold increase in mean epoetin dose to achieve a normal hematocrit (fig. 4) [7]. Moreno et al. [17] found that normalization of hemoglobin required a 51% in-

Normalization of Hematocrit in Renal Patients

crease in epoetin dose requirements, while the latest study to be published, by McMahon et al. [16], showed an 80% increase in epoetin dose requirements in the patients targeting a normal hemoglobin. With the current fairly high costs of epoetin therapy, it is difficult to justify this extra expenditure for what may be a limited benefit. Using this 80% increase in epoetin dose requirements, the health economist would be able to calculate that for every 5 patients treated for normalization of hemoglobin, he can treat 9 patients aiming for partial correction of anemia. If the cost of epoetin were to reduce, then this argument may be less strong, but there is no sign of this at the present time. Requirements for IV Iron Supplementation Greater In addition to the need for an increased epoetin dose, there is also an increased need for iron supplementation

Blood Purif 2001;19:157–167

163

in patients aiming for normalization of hemoglobin. This was again most dramatically illustrated in the US Normal Hematocrit Trial in which 526 of 618 patients in the normalization of hemoglobin group received IV iron dextran compared with 464 of 615 patients in the lower hemoglobin group (p ! 0.001) [7]. This has both economic and safety considerations. Not only does the cost of targeting a normal hemoglobin increase further, but there are continuing concerns about the short- and long-term safety of IV iron [21]. Thus, patients may be exposed to a greater risk of anaphylactoid reactions, side effects from IV iron, oxidative stress, vascular endothelial damage, and bacterial infections. While the undoubted benefits of using IV iron supplementation seem fully justified at the lower levels of hemoglobin, it is not clear whether the benefit:risk ratio would be maintained in patients destined for hemoglobin normalization. Normalization of Hemoglobin Increases the Risk of Vascular Access Thrombosis As described earlier, two of the most scientifically robust randomized controlled trials of epoetin therapy have shown significantly increased incidences of vascular access clotting at a higher hemoglobin level [7, 14]. While it is difficult to be certain whether this is related to the hemoglobin per se, or to the increased epoetin dose, or to the increased IV iron requirements, it is however hard to ignore the scientific data from these two studies. Although not usually life-threatening, thombosis of the vascular access does result in considerable morbidity and inconvenience for the patient, and the use of temporary dialysis lines while awaiting maturation of a new fistula also carries its own risk, particularly for bacterial sepsis. Normalization of Hemoglobin Results in Less Efficient Dialysis This is more relevant for hemodialysis patients, in whom it has been shown that higher hemoglobin levels result in less efficient clearance of various solutes. This phenomenon was first recognized in the initial epoetin study by Eschbach et al. [6] in which several patients were found to develop profound hyperkalemia. More recently, Movilli et al. [22] found an inverse relationship between hematocrit and both dialyzer clearance (p = 0.003) and Kt/V (p = 0.0002); the lowest Kt/V values were in the group with a hematocrit 637%. This has also recently been the subject of a study by Ronco et al. [23] in which it was shown that dialyzer efficiency was reduced in hemodialysis patients following normalization of hemoglobin with epoetin. Although it may be possible to overcome

164

Blood Purif 2001;19:157–167

this problem by using larger dialyzers and increasing the hours on dialysis, each of these maneuvers carries an increased cost and inconvenience for patients and healthcare workers alike. Normalization of Hemoglobin Increases the Risk of Dialyzer Clotting The relationship between hemoglobin and whole blood viscosity is a logarithmic one in that there is an exponential increase in whole blood viscosity for every unit increase in hemoglobin concentration (fig. 5) [24]. This increase in whole blood viscosity with normalization of hemoglobin undoubtedly increases the risk of blood clots clogging up the dialyzer. This in turn exacerbates the reduced dialyzer clearance of solute, and results in an increased requirement for heparin. The latter again carries cost and safety considerations, in that higher heparin doses may result in an increased risk of bleeding postdialysis. Hemoglobin Concentration Increases Significantly across Dialysis Depending on the amount of fluid ultrafiltered during the dialysis session, there is a variable increase in hemoglobin concentration in a predialysis compared with a postdialysis blood sample [25]. We recently studied this in our unit, systematically taking blood samples for hemoglobin measurement pre- and post-dialysis. This was correlated with the change in body weight across dialysis, and the results are shown in figure 1. Some patients have a fairly minimal increase in hemoglobin concentration, while others have an increase of up to 3–4 g/dl. Although this effect is probably short-lived due to extracellular fluid shifts postdialysis, the patient is nevertheless exposed to a higher hematocrit level for a short spell at the end of dialysis. With a target hemoglobin concentration of 11 g/dl there is a built-in ‘cushion’ for this, but if a patient is striving for a hemoglobin of 14 g/dl then this transient increase in hemoglobin concentration could be potentially hazardous. This phenomenon is not relevant in the peritoneal dialysis population. Hemoglobin Fluctuates Significantly in Dialysis Patients on Epoetin Due to a number of factors, including intercurrent illness and clinical events, patients on dialysis do not maintain a completely stable hemoglobin concentration. Even ignoring the fluid shifts across dialysis, the hemoglobin concentration can fluctuate by F1–2 g/dl in patients on epoetin. A number of factors may explain this, some of

Macdougall

Fig. 5. Relationship between hemoglobin

concentration and whole blood viscosity (WBV) measured at shear rates of 3, 30 and 300 s –1 in 10 hemodialysis patients receiving epoetin over a 12-month period. Data from Macdougall et al. [24].

which are still unexplained. Thus, epoetin dose adjustments, either up or down, are frequently required in dialysis patients, and it is really impossible to keep the hemoglobin concentration within tight margins in this population. While this fluctuation in hemoglobin is probably less relevant at a target hemoglobin of around 11 g/dl, there is clearly an increased risk of serious overshoot of hemoglobin if patients are being targeted towards a normal hemoglobin. Thus, a patient who has been reasonably stable with a hemoglobin of around 14 g/dl may suddenly become more responsive to epoetin and end up with a hemoglobin of 17 g/dl. If a patient who has been maintained around 11 g/dl ‘overshoots’ to 14 g/dl then this may be less hazardous.

Normalization of Hemoglobin in Pre-ESRD Patients

Although there are now many studies on the effect of epoetin in improving anemia in pre-ESRD patients, in contrast to dialysis patients there are limited scientific

Normalization of Hematocrit in Renal Patients

data on normalization of hemoglobin in this population [26]. This makes it impossible to draw any firm conclusions either way regarding this issue. As discussed previously, however, there may be a stronger case for normalizing hemoglobin in pre-ESRD patients who have not been exposed to anemia for such a long time and who have not been rendered severely anemic. There is increasing interest in starting epoetin at an earlier stage in the development of renal failure, and indeed there is an increasing scientific rationale for doing so [27]. We are now aware that by the time the patients are started on regular dialysis treatment, they already have a high chance of having cardiac pathology, either with left ventricular hypertrophy [28], left ventricular dilatation [29], or systolic dysfunction. There is one prospective study examining the cardiovascular effects of normalizing hemoglobin with epoetin therapy in pre-ESRD patients. Hayashi et al. [26] assessed left ventricular mass, 24-hour blood pressure monitoring, and change in renal function in 9 predialysis patients after partial correction (target hematocrit 30%) and normalization (target hematocrit 40%) of their anemia. Left ventricular mass progressively

Blood Purif 2001;19:157–167

165

decreased from 140.6 B 12.1 g/m2 at baseline, to 126.9 B 10.0 g/m2 after partial correction of anemia at 4 months, to 111.2 B 8.3 g/m2 after normalization of anemia at 12 months. There was no change in the rate of progression of renal failure over the 12 months of the study. In addition, there are currently ongoing multicenter studies of epoetin in predialysis patients in the UK, Canada and Australia, and a global study of epoetin targeting two different hemoglobin concentrations in such patients has recently been launched. This is the CREATE study (Cardiovascular Reduction Early Anemia Treatment with Epoetin Beta), the design of which was presented at a Roche Technical Forum at the European Renal Association Meeting in Nice (September 2000). In Group 1, patients are started on epoetin when their hemoglobin falls below 12.5 g/dl, and they are then treated to maintain a hemoglobin concentration between 13.0 and 15.0 g/dl. The Group 2 (control) patients will only start epoetin when their hemoglobin falls below 10.5 g/dl, and they are targeted to maintain a hemoglobin concentration between 10.5 and 11.5 g/dl. Although it may be quite some time before any data are available from this study, it is the first one to address specifically the issue of preventing renal anemia developing in the first place, and also maintaining renal failure patients with a normal hemoglobin concentration throughout the progression of their disease. Until such data are available, however, it is difficult to make any judgements either way about target hemoglobin in pre-ESRD patients. In contrast to dialysis patients, however, pre-dialysis patients do not have the problems of fluid shifts affecting the variability in hemoglobin concentration.

Conclusions

The task I was set was to provide a counter-argument to normalization of hemoglobin in renal failure patients. The case I have mounted draws on scientific data obtained from a number of the ‘normalization of hemoglobin’ studies in dialysis patients. The arguments include the fact that there is no evidence that a higher hemoglobin is of significant benefit, the epoetin dose requirements are greater, the cost is significantly greater, the requirements for iron supplementation are greater, there is an increased risk of vascular access clotting, there is less efficient dialyzer clearance of solute, there is an increased risk of clotting in the dialyzer, heparin requirements may increase, there is an exponential increase in whole blood viscosity, and some hemodialysis patients may develop significant hemoglobin overshoot partly due to fluid removal across dialysis. Despite all of this evidence, it does not stop me normalizing hemoglobin in select renal failure patients. The type of patient I would be more liable to maintain at a normal hemoglobin is illustrated in Case 1, whereas I would be much more nervous of attempting this in Case 2. Thus, the overriding issue with normalization of hemoglobin must be to individualize the target hemoglobin depending on the various characteristics and needs of the patient.

Acknowledgement I am grateful to my secretary, Christine Mitchell, for help in the preparation of the manuscript.

References 1 Nissenson AR, Besarab A, Bolton WK: Target haematocrit during erythropoietin therapy. Nephrol Dial Transplant 1997;12:1813–1816. 2 Jacobs C: Normalization of haemoglobin: Why not? Nephrol Dial Transplant 1999;14(suppl 2):75–79. 3 NKF-DOQI Work Group: NKF-DOQI clinical practice guidelines for the treatment of anemia of chronic renal failure. Am J Kidney Dis 1997; 30(suppl 3):S192–S240. 4 European Best Practice Guidelines for the management of anaemia in patients with chronic renal failure. Nephrol Dial Transplant 1999;14(suppl 5):1–50.

166

5 Winearls CG, Oliver DO, Pippard MJ, Reid C, Downing MR, Cotes PM: Effect of human erythropoietin derived from recombinant DNA on the anaemia of patients maintained by chronic haemodialysis. Lancet 1986;i:1175– 1178. 6 Eschbach JW, Egrie JC, Downing MR, Browne JK, Adamson JW: Correction of the anemia of end-stage renal disease with recombinant human erythropoietin. N Engl J Med 1987;316: 73–78. 7 Besarab A, Kline Bolton W, Browne JK, Egrie JC, Nissenson AR, Okamoto DM, Schwab SJ, Goodkin DA: The effects of normal as compared with low hematocrit values in patients with cardiac disease who are receiving hemodialysis and epoetin. N Engl J Med 1998;339: 584–590.

Blood Purif 2001;19:157–167

8 McMahon LP, McKenna MJ, Sangkabutra T, Mason K, Sostaric S, Skinner SL, Burge C, Murphy B, Crankshaw D: Physical performance and associated electrolyte changes after haemoglobin normalization: A comparative study in haemodialysis patients. Nephrol Dial Transplant 1999;14:1182–1187. 9 Martinez-Vea A, Bardaji A, Garcia C, Ridao C, Richart C, Oliver JA: Long-term myocardial effects of correction of anemia with recombinant human erythropoietin in aged patients on hemodialysis. Am J Kidney Dis 1992;19:353– 357. 10 Furuland H, Linde T, Danielson BG: Cardiac function in patients with end-stage renal disease after normalization of hemoglobin with erythropoietin. J Am Soc Nephrol 1998;9: 337A.

Macdougall

11 Furuland H, Linde T, Danielson BG: Physical exercise capacity in patients with end-stage renal disease after normalization of hemoglobin with erythropoietin. J Am Soc Nephrol 1998;9: 337A. 12 Foley RN, Parfrey PS, Morgan J and the Canadian Normalization of Hemoglobin Study Group: A randomized controlled trial of complete vs partial correction of anemia in hemodialysis patients with asymptomatic concentric LV hypertrophy or LV dilatation. J Am Soc Nephrol 1998;9:208A. 13 Macdougall IC, Ritz E: The Normal Haematocrit Trial in dialysis patients with cardiac disease: Are we any the less confused about target haemoglobin? Nephrol Dial Transplant 1998; 13:3030–3033. 14 Canadian Erythropoietin Study Group: Association between recombinant human erythropoietin and quality of life and exercise capacity of patients receiving haemodialysis. BMJ 1990; 300:573–578. 15 Eschbach JW, Glenny R, Robertson T, Guthrie M, Rader B, Evans R, Chandler W, Davidson R, Easterling T, Denney J, Schneider G: Normalizing the hematocrit in hemodialysis patients with EPO improves quality of life and is safe. J Am Soc Nephrol 1993;4:425. 16 McMahon LP, Mason K, Skinner SL, Burge CM, Grigg LE, Becker GJ: Effects of haemoglobin normalization on quality-of-life and cardiovascular parameters in end-stage renal failure. Nephrol Dial Transplant 2000;15:1425–1430.

Normalization of Hematocrit in Renal Patients

17 Moreno F, Sanz-Guajardo D, Lopez-Gomez JM, Jofre R, Valderrabano F: Increasing the hematocrit has a beneficial effect on quality of life and is safe in selected hemodialysis patients. Spanish Cooperative Renal Patients Quality of Life Study Group of the Spanish Society of Nephrology. J Am Soc Nephrol 2000;11:335–342. 18 Ma JZ, Ebben J, Xia H, Collins AJ: Hematocrit level and associated mortality in hemodialysis patients. J Am Soc Nephrol 1999;10:610–619. 19 Xia H, Ebben J, Ma JZ, Collins AJ: Hematocrit levels and hospitalization risks in hemodialysis patients. J Am Soc Nephrol 1999;10:1309– 1316. 20 Moreno F, Lopez-Gomez JM, Sanz-Guajardo D, Jofre R, Valderrabano F. Quality of life in dialysis patients: A Spanish multicentre study. Nephrol Dial Transplant 1996;11(suppl 2): 125–129. 21 Fishbane S, Maesaka JK, Mittal SK: Is there material hazard to treatment with intravenous iron? Nephrol Dial Transplant 1999;14:2595– 2598. 22 Movilli E, Cancarini GC, Mombelloni S, Feller P, Ravelli M, Maiorca R: The role of hematocrit in efficiency of dialysis. Blood Purif 1990; 8:183–189. 23 Ronco C, Ghezzi PM, Metry G, Spittle M, Brendolan A, Rodighiero M, Milan M, Zanella M, Greca G, Levin N: Effect of hematocrit and blood flow distribution on solute clearance in hollow fiber hemodialyzers. Nephron 2000 (in press).

24 Macdougall IC, Davies ME, Hutton RD, Coles GA, Williams JD: Rheological studies during treatment of renal anaemia with recombinant human erythropoietin. Br J Haematol 1991;77: 550–558. 25 Vlassopoulos D, Sonikian M, Dardioti V: Hadjiconstantinou V: Target haematocrit during erythropoietin treatment in dialysis patients. Which value is true-functional haematocrit? Nephrol Dial Transplant 1999;14:1340–1341. 26 Hayashi T, Suzuki A, Shoji T, Togawa M, Okada N, Tsubakihara Y, Imai E, Hori M: Cardiovascular effect of normalizing the hematocrit level during erythropoietin therapy in predialysis patients with chronic renal failure. Am J Kidney Dis 2000;35:250–256. 27 Macdougall IC: Higher target haemoglobin level and early anaemia treatment: Different or complementary concepts? Nephrol Dial Transplant 2000;15(suppl 3):3–7. 28 Levin A, Thompson CR, Ethier J, Carlisle EJF, Tobe S, Mendelssohn D, Burgess E, Jindal K, Barrett B, Singer J, Djurdjev O: Left ventricular mass index increase in early renal disease: Impact of decline in hemoglobin. Am J Kidney Dis 1999;34:125–134. 29 Foley RN, Parfrey PS, Harnett JD, Kent GM, Martin CJ, Murray DC, Barre PE: Clinical and echocardiographic disease in patients starting end-stage renal disease therapy. Kidney Int 1995;47:186–192.

Blood Purif 2001;19:157–167

167

Blood Purif 2001;19:168–174

Should the Hematocrit (Hemoglobin) Be Normalized in Pre-ESRD or Dialysis Patients? Yes! Anatole Besarab Mohammad Asim Aslam Section of Nephrology, Department of Medicine, West Virginia University School of Medicine, Morgantown, W. Va., USA

Introduction

The major issues regarding recombinant human erythropoietin (rHuEpo) use in patients with chronic kidney disease (CKD), particularly those receiving dialysis, have been cost-effectiveness, dosage, route of administration and resistance to therapy [1]. Anemia management has been optimized by using the subcutaneous route [2] and by using intravenous iron in a timely fashion [3, 4]. Optimal treatment of anemia has assumed greater importance with the increasing age of the end-stage renal disease (ESRD) and progressive CKD (pCKD) populations. This fast-growing segment is increasingly composed of diabetic patients and has more frequent and greater degree of ischemic heart and peripheral vascular disease. Those with pCKD all too often present for dialysis with well-established comorbidities of uremia, such as anemia, left ventricular hypertrophy (LVH), congestive heart failure, malnutrition, or debilitation. It is the purpose of this discourse to make the argument that many of these comorbidities, often seen as inevitable components of kidney failure, could be prevented if relatively normal hemoglobin concentration were maintained during the entire period of pCKD. The optimal hemoglobin (Hb)/hematocrit (Hct) during EPO therapy in both ESRD and pCKD patients should be

ABC

© 2001 S. Karger AG, Basel 0253–5068/01/0192–0168$17.50/0

Fax + 41 61 306 12 34 E-Mail [email protected] www.karger.com

Accessible online at: www.karger.com/journals/bpu

one that maximizes cardiovascular function and activities of daily life (ADL) while minimizing risk. Different organs respond differently to correction of anemia, an effect that confounds the search for optimum Hb. Very few studies have directly addressed the negative impact of the severity of anemia in those with pCKD. In such patients, it is difficult to separate the effects of anemia from those due to mild azotemia. This difficulty does not justify the improper treatment of anemia. Using data from the US Renal Data System, Obrador et al. [5] demonstrated that the majority of patients with pCKD who initiate dialysis in the USA receive suboptimal care. Acceptance of a certain degree of anemia as the norm may explain the finding that the mean Hb in pCKD patients beginning dialysis was only F9 g/dl [5]. Many patients with pCKD may not receive optimal therapy as a result of cost concerns or fear of adverse effects of anemia correction. The role of preventive medicine is underappreciated. In the Canadian study on pCKD, patients with creatinine clearances !25 ml/min averaged a Hb of 10.9 g/dl and showed a progressive decline in mean Hb at each level of kidney disease [6]. Development of long-acting erythropoietin (novel erythropoietic stimulating product) [7] and the wide availability of safe iron preparations permits maintenance or correction of Hb at any desired level

Anatole Besarab, MD Division of Nephrology, Department of Medicine Robert C. Byrd Health Sciences Center, HSS 1264 PO Box 9165, Morgantown, WV 26506 (USA) Tel. +1 304 293 2551, Fax +1 304 293 7373, E-Mail [email protected]

including normal. Indeed, several authors have suggested that normalization of Hb in patients with chronic renal failure might produce greater benefits than those currently achieved [8–10]. Four prospective studies of nearly normalizing Hb have been completed, mostly in ESRD patients. All have shown improvements in quality of life (QOL). Only one has shown a risk. Intuitively, normalization should benefit because in all other forms of hormone deficiency, replacement is given to fully correct the deficiency. The definition of normal Hb differs for patients with pCKD and those on peritoneal dialysis from that of hemodialysis patients. The later have recurrent changes in weight between dialysis sessions. Because of hemoconcentration during dialysis, the ‘normal’ Hb measured predialysis in hemodialysis patients is 1–1.5 g lower than in the other two groups.

not allowed to develop, could the severity of LVH in patients with pCKD be reduced? The purpose of the remainder of this discourse is to make an argument for normalization of Hb in almost all patients and to start before the ESRD stage is reached. Studies have documented that oxygen transport to many organs decreases rapidly as Hct falls to less than 40%. These organs function better at higher Hb levels [22]. With progression of chronic renal failure and worsening anemia, the opportunity to maintain organ function is lost. Once lost, subsequent interventions may never restore full functionality (as with LVH). As a result, QOL, exercise performance and rehabilitative potential are adversely affected. Maintenance of normal Hb may prevent the maladaptive compensations.

Cardiovascular Function Historical, Economic, Philosophic, and Research Perspectives

When the protocols for the initial clinical rHuEpo trials were being developed, hematologists argued for full normalization of Hb. In contrast, nephrologists argued for incomplete correction [11]. Initial clinical trials [12, 13] achieved mean Hb of about 12 g/dl (Hct 33–38). Marketplace economics in the USA produced a relatively low target Hb of 9.5–11 g/dl for its ESRD population [1]. It is now possible to maintain a 3-month rolling average of !12.5 g/dl. European nephrologists apparently also prefer a target of 10–12 g/dl [14]. The rationale for these levels is not clear since they are short of the normal ranges of 12.0– 16 g/dl for menstruating women and 13.5–17.5 g/dl for men and postmenopausal women. The increased cost incurred to maintain Hb at normal levels requires that cogent arguments be made for doing so. If anemia treatment is begun predialysis and extends over years, the costs become important considerations. Until recently there have been few studies investigating outcomes and effects of increasing Hb to 13–15 g/dl. We do need more studies that extend correlation of Hb concentrations to outcome, particularly in those with pCKD. In ESRD patients, Hb of 11–12 g/dl produces a survival advantage [15–17] and lower overall hospitalization [18, 19]. Collins and Keane [20] believe that keeping Hb at 11–13 g/dl is cost-effective with reduction of long-term costs for caring for ESRD patients. Will normalization of Hb produce even greater effects? LVH is a negative predictor of survival in dialysis patients [21]. If anemia were

Hematocrit (Hemoglobin) Normalization in Pre-ESRD and Dialysis Patients

Patients with progressive anemia develop a series of compensations. Cardiac output increases by 20% and a rightward shift in the oxygen dissociation curve occurs. The increase in cardiac output is mediated by both an increase in stroke volume and heart rate. There is a reduction in afterload from hypoxic vasodilatation and decreased viscosity, an increase in preload, and an initial increase in contractility from increased sympathetic activity [23]. Eventually, anemia produces both eccentric LV as well as left ventricular (LV) dilatation. Hb concentrations correlate with both LV end-diastolic volume and LV mass. LVH is associated with the de novo development of ischemic heart disease. LVH, LV dilatation and ischemic heart disease produce cardiomyopathy [24]. Anemia decreases cardiac reserve in hearts with LVH in which the maximum vasodilatory reserve of the coronary circulation is already reduced. Coronary reserve in the patient with renal insufficiency is decreased by anemia because compensatory vasodilatation uses up a portion of the reserve. In ‘uremia’ tolerance for ischemia is further reduced because of arteriolar wall thickening and inadequate capillary growth producing a cardiomyocyte/capillary mismatch [25]. Hypertrophic myocytes become at risk for hypoxia because of the increased diffusion distance. The importance of the above observations is most apparent in those who have angina pectoris but normal coronaries [26]. Seemingly, the myocardium of the renal patient increasingly lives on the brink of anoxia as anemia develops. LVH per se also predisposes to ventricular arrhythmia [27] in part through this effect on coronary reserve and in

Blood Purif 2001;19:168–174

169

part through interstitial fibrosis [28]. Ischemia and the high resistance to electrical conduction establish a stage for re-entry pathways and re-entry arrhythmia [29, 30]. Many ‘sudden deaths’ in dialysis patients probably result from arrhythmia. The risk of LV dilatation, heart failure and premature death increases as the anemia of pCKD worsens. Levin et al [6] have demonstrated that the development of LVH is closely related to the decrease in Hb in patients with progressive renal disease. Prevalence of LVH is related to the degree of renal dysfunction; 26% at creatinine clearance of 50–75 ml/min, 33% at 25–50 ml/min, 41% at !25 ml/ min [31]. By logistic regression, each 1 g/dl decrease in Hb increased the risk of LVH by 6%. Hb level and systolic hypertension are the only two variables that are significant predictors of LVH in these nondialysis patients. When followed over 12 months, an increase in LVMI of 20% or 20 g/m2 correlated only with lower Hb and higher blood pressure. Thus cardiovascular consequences begin early in pCKD and are closely linked to changes in Hb over time. These abnormalities worsen as patients approach dialysis. Approximately 70–75% of patients entering dialysis programs have LVH [32]. Once dialysis is initiated, subset analysis by Foley et al. [24] has shown that a further increase in LVH occurs during the first year of dialysis. In view of the above detrimental effects of LVH, every effort must be made to reduce it or reverse it. The strategy should be treatment at an earlier stage before fibrosis sets in. Maintenance of Hb in these patients before the need for replacement therapy may be the key to preventing progressive increase in LV mass. Partial correction of anemia to a Hct of 30–36% produces only incomplete reductions in LV hypertrophy [33, 34], volume [35], or mass [36] and incomplete improvement in exercise-induced ST-segment depression [37]. Since development of anemia during the predialysis period is a key, normalization of Hb should be begun when the opportunity still exists to maintain muscle function as well as prevent cardiac dilatation. Pharmaco-economic treatment choices should focus on prevention of cardiac complications during the development of renal insufficiency. On the basis of the above discussion, it is useful to evaluate the studies that have attempted to normalize Hb. Eschbach et al. [38] increased the Hct from a mean of 32.6 to 42.0% and noted significant improvements in cardiac function and exercise capacity. Specifically cardiac output decreased and LV mass decreased. Major side effects or risks were not observed in this small uncontrolled study. By contrast, a controlled study of dialysis patients

170

Blood Purif 2001;19:168–174

with cardiac disease (Normal Hematocrit Cardiac Trial) arrived at the opposite result. The study was terminated after an interim analysis showed a tendency toward a higher death rate in the group whose Hb had been normalized [39]. Although the difference did not reach a predefined statistical boundary, the study was halted because of safety concern (access clotting) and the extreme unlikelihood of showing any benefit. Difference in outcome could not be explained by 11 prespecified baseline characteristics using Cox regression analysis. Higher Hb levels per se did not appear to be responsible for the death rate findings since post-hoc analysis showed that mortality decreased with higher Hb levels in both the control group as well as the normal Hct group. Overall there was a 30% reduction in mortality for a 10-point increase in Hct. The Canadian Multicenter Trial [40] evaluated LVH and QOL in 159 anemic hemodialysis patients with asymptomatic LV disease. Two sets of patients were randomized; those with LVH and normal LV cavity and those with LV dilatation alone. In contrast to the Normal Hematocrit Cardiac Trial, patients with symptomatic cardiac disease were excluded. Normalization of Hb to 13– 14.0 g/dl as opposed to maintenance at 10 g/dl for 48 weeks prevented progression of LV dilatation in those starting with a normal LV volume but did not reverse regression of pre-existing concentric LVH or LV dilatation. Along with the previous studies of Levin et al. [6, 31], the results suggest that early normalization may be a key strategy. In the Scandinavian Multicentre Trial [41], a total of 416 hemodialysis, peritoneal and pCKD patients with stable Hb levels between 9 and 12 g/dl were randomized into two Hb target groups, to remain at levels !12 g/ dl or to be increased to 13.5–16 g/dl for a total duration of 18 months. This study found no difference in the two groups with regard to safety. In summary, normalization of Hb appears to be safe in patients who have asymptomatic cardiac dysfunction, a finding that differs from that of normalization in patients with symptomatic coronary artery disease and/or congestive heart failure. Reversal of adaptive circulatory adaptations requires considerable time. In patients with symptomatic heart disease it is prudent to control blood pressure while slowly correcting the anemia. Several studies now indicate no or trivial differences in mean day or nocturnal blood pressure at Hb levels of 10 compared to Hb levels of 14 g/dl [42]. In the Normal Hematocrit Cardiac Trial, normalization did not produce an increase in BP as assessed by ambulatory monitoring nor was there a change in the diurnal pattern [43, 44].

Besarab/Aslam

Cerebrovascular

Maximal delivery of O2 occurs with Hb in the 13.5– 15 g/dl range. Several studies have shown improvements in cognitive function on increasing the Hb above the levels currently recommended. Nissenson and co-workers [45] found better brain function at Hcts of 42% than at the lower clinical targets of 30–36%. Continuous performance tasks and P-300 latency in the auditory oddball task by electroencephalogram improved as Hb was normalized suggesting better signal transmission in the central nervous system. Normalization of Hb increases oxygen supply to the brain and oxygen extraction [46].

Quality of Life

QOL is affected by the ability to perform physical ADL. The definition of what constitutes adequate ADL function differs among age groups, particularly when comparing older sedentary to younger more active-working or pediatric patients. In all age groups, CKD impacts QOL negatively [47, 48]. More than 25% of patients with pCKD reported important symptoms relative to general health, energy, fatigability, sleep disturbance, recreational activities, home management, and work [47]. A low Hb level was the only laboratory variable that correlated with poor QOL scores. Schidler [48] studied 50 patients with mild degree of CKD and found that such patients exhibited negative psychological reactions to their illness, associated with lower QOL and higher depression scores. The Normal Hematocrit Cardiac Trial did report significant improvements in QOL parameters in the patients in the normal Hb group at the time the study was halted [39]. The observation by Valderrabano [49] that improvement in the sickness impact profile (SIP) and Karnofsky scores correlated with Hct over the full range of Hcts in 1,013 Spanish hemodialysis patients led to a prospective 6-month study (Spanish Quality of Life Study) to evaluate functional status and QOL resulting from a deliberate increase in Hct [50]. Only stable hemodialysis patients under the age of 65 were entered and those with diabetes, CVA, seizures or severe co-morbidity were specifically excluded. Although the study was designed to increase hematocrit by about 5 points, the mean increase was 7.5 points in the 115 patients who finished the study. Increasing the Hct from 31 to 39% in 117 patients statistically altered the SIP and Karnofsky scores. When Hb was increased from 10.2 to 12.5 g/dl, the overall SIP decreased from 7.8 to 5.7, the physical dimension component from

Hematocrit (Hemoglobin) Normalization in Pre-ESRD and Dialysis Patients

5.9 to 4.1, and the psychological dimension score from 8.9 to 7.3. The number of patients hospitalized and their length of stay decreased during the 6 months of the study when compared to the 6-month period preceding the study. The greatest benefit accrued in those having the lowest baseline QOL. Similar results have been reported from the Scandinavian Erythropoietin study at 1 year of follow-up [51]. Physical symptoms were reduced with less fatigue, depression and frustration in the normal Hb group than in the lower Hb group. Similar trends occurred in the Leicester Uremic Symptoms Scale, the ADL scale, and Self– Image Scales. Of note, the dropout rate in those with pCKD was low. The Canadian Multicenter Trial has also reported on QOL outcomes. At 24 and 48 weeks into the study, improvements in the Kidney Disease Questionnaire and the SF-36 subscales have been noted without a change in the Health Utilities Index. Importantly, normalization had least effect in those with LVH and greatest impact in those with LV dilatation. Changes in depression, fatigue and relationships were the subscales showing the most effect from normalization. Thus the data on QOL from normalization Hb is consistent. Normalization improves QOL. We believe that even greater potential exists for preventing disability by insisting on Hb normalization much earlier in pCKD.

Physical Activity

During exercise, maximum oxygen uptake can increase 20-fold but cardiac output can only increase 6-fold. Increase in oxygen uptake is dependent on age, gender, height and weight, and habitual physical activity. Reduced exercise capacity in patients with pCKD is commonly observed due to (1) reduction in oxygen delivery, (2) concurrent diseases such as peripheral vascular disease, heart failure, (3) increased proteolysis, (4) malnutrition, and (5) hyperparathyroidism. In addition, muscle abnormalities develop. Overall muscle strength of dialysis patients decreases to about half [52]. Lack of muscle strength may be the principal factor limiting exercise capacity of anemic renal failure patients. Reported increases in exercise and work capacity in ESRD patients are proportional to the Hct increases but show wide interindividual variation. The effect of restoring full physical and intellectual performance derives from limited studies to date. Increasing Hb improves exercise-induced lactate concentration [53]. Lim [54] reported that energy and work capacity

Blood Purif 2001;19:168–174

171

improved more in those who were fully corrected (Hct 35–40%) than in those only partially corrected. Barany et al. [55] found that exercise capacity did increase further with normalization of Hb but that the increase remained subnormal compared to rHuEPO-treated healthy subjects. Although VO2 is higher at a Hct of 35–40% than at 30%, exercise ability is not fully corrected [56]. Eschbach et al. [38] showed a 24% increase in maximal oxygen uptake, a 20% increase in bicycle exercise, a 50–60% increase in pulmonary CO2 diffusion, and an increase in quadriceps isometrics with Hb normalization. Finally an increase in Hb from 10 to 14 g/dl produces a significant increase of maximal exercise performance in hemodialysis patients [57]. Although uremic factors may contribute to the pathogenesis of impaired exercise and muscle dysfunction even after full correction of anemia, one has to question whether the abnormalities would be less if anemia and deconditioning were not allowed to develop in the first place. With respect to the later aspect, Stray-Gunderson et al. [57] studied patients at a Hct of 30% and at 42%. At the higher Hct the subjects were examined after 15 weeks of exercise training. O2 uptake increased markedly after the exercise training. Thus poor conditioning of many dialysis patients contributes to the reduced exercise capacity in patients with pCKD.

Morbidity and Risk

Scandinavian Trials showed any risk of normalizing Hb in patients without symptomatic cardiac disease. Normalization of Hb does not appear to have any negative impact on blood pressure stability [39, 43, 44]. In 20% of patients, additional antihypertensives may be necessary. If blood pressure is adequately controlled, rHuEPO does not accelerate progression of chronic renal insufficiency [59]. More recent studies using more accurate measurements of GFR ([125I]iothalamate clearance) over a 1-year period in pre-dialysis patients have shown no evidence for a Hct-induced change in GFR nor was there any difference between the treated (Hct 35–36) and untreated patients in time to initiation of dialysis [60]. In some studies, therapy in pCKD even appeared to retard progression in nondiabetic patients [61]. Could this be an effect to reduce oxidative stress or ischemia within the kidney?

Conclusions

The available evidence indicates that QOL, exercise capacity and cardiac function improve as Hb approaches normal values in patients with CKD. At present there is incomplete information on normalizing Hb in predialysis patients. Randomized studies are needed to identify appropriate endpoints (time to renal replacement therapy, rate of progression, exercise capacity, LV growth and geometry, cardiac events, morbidity and mortality).

Patients on dialysis have an age-adjusted death rate 3.5 times that of the general population. Anemia predicts mortality independently of diabetes, cardiac failure, hypoalbuminemia, or blood pressure [24, 32]. It also independently predicts the development of congestive heart failure [32]. Many of these effects may result from the affect of anemia on LVH. The presence of LVH at the start of dialysis is independently predictive of mortality for the next 2 years with a relative risk of 2.9 for all-cause mortality and 2.7 for cardiac mortality. Silverberg et al. [58] reported that the patient whose LV mass exceeded the normal value had an actuarial survival of 25% compared to 55% in patients with normal LV mass. The morbidity and mortality data by Collins and co-workers [17, 19] and Locatelli et al. [16] indicate that no patient should have a Hct less than 30%. US studies were unable to demonstrate whether increasing Hb to 112 g/dl prolonged survival due to the small number of subjects [15, 17] with such Hb levels due to the reimbursement system. As stated before, the Normal Hematocrit Cardiac Study did not show benefit. However, neither the Canadian nor the

172

Blood Purif 2001;19:168–174

Besarab/Aslam

References 1 Besarab A, McCrea JB: Evolution of recombinant human erythropoietin usage in clinical practice in the United States. ASAIO J 1993; 39:11–18. 2 Kaufman JS, Reda DJ, Fye CL, Goldfarb DS, Henderson WG, Kleinman JG, Vaamonde CA: Subcutaneous compared to intravenous epoetin in patients receiving hemodialysis. Department of Veterans Affairs Cooperative Study Group on Erythropoietin in Hemodialysis Patients. N Engl J Med 1998;339:578–583. 3 Taylor JE, Peat N, Porter C, Morgan AG: Regular, low-dose intravenous iron therapy improves response to erythropoietin in haemodialysis patients. Nephrol Dial Transplant 1996; 11:1079–1083. 4 Besarab A, Kaiser JW, Frinak S: A study of parenteral iron regimens in hemodialysis patients. Am J Kidney Dis 1999;34:21–28. 5 Obrador G, Ruthazer R, Arora P, Kausz A, Pereira B: Prvalence of and factors associated with suboptimal care before initiation of dialysis in the United States. J Am Soc Nephrol 1999;10:1793–1800. 6 Levin A, Singer J, Thompson C, Ross H, Lewis M: Prevalent left ventricular hypertrophy in the predialysis population: Identifying opportunities for intervention. Am J Kidney Dis 1996;27:347–354. 7 Macdougall IC, Gray SJ, Elston O, Breen C, Jenkins B, Browne J, Egrie J: Pharmacokinetics of novel erythropoietic stimulating protein compared to epoetin alfa in dialysis patients. J Am Soc Nephrol 1999;10:2392–2395. 8 Jacobs C: Normalization of haemoglobin: Why not? Nephrol Dial Transplant 1999;14(suppl 2):75–79. 9 Ritz E, Schwenger V: The optimal target hemoglobin. Sem Nephrol 2000;20:382–386. 10 Macdougall IC, Ritz E: The Normal Hematocrit Trial in dialysis with cardiac disease. Are we any the less confused about the target haemoglobin. Nephrol Dial Transplant 1998;13: 3030–3033. 11 Nissenson AR, Besarab A, Bolton WK, Goodkin DA, Schwab SJ: Target hemoglobin/hematocrit during EPO therapy. Nephrol Dial Transplant 1997;14:1813–1816. 12 Eschbach JW, Egrie JC, Downing MR, Browne JK, Adamson JW: Correction of the anemia of end-stage renal disease with recombinant human erythropoietin. Results of combined phase I & II clinical trials. N Engl J Med 1987; 316:73–78. 13 Winearls CG, Oliver DO, Pippard MJ, Reid C, Downing MR, Cotes PM: Effect of human erythropoietin derived from recombinant DNA on the anemia of patients maintained by chronic haemodialysis. Lancet 1986;ii:1175– 1177. 14 Ritz E, Amann K: Optimal haemoglobin during treatment with recombinant human erythropoietin. Nephrol Dial Transplant 1998;13 (suppl 2):16–22.

Hematocrit (Hemoglobin) Normalization in Pre-ESRD and Dialysis Patients

15 Madore F, Lowrie E, Brugnara C, Lew NL, Lazarus JM, Bridges K, Owen WF: Anemia in hemodialysis patients: Variables affecting this outcome predictor. J Am Soc Nephrol 1997;8: 1921–1929. 16 Locatelli F, Conte F, Marcelli D: The impact of haematocrit levels and erythropoietin treatment on overall cardiovascular mortality and morbidity – The experience of the Lombardy Dialysis Registry. Nephrol Dial Transplant 1998;13:1642–1644. 17 Ma JZ, Ebben J, Xia H, Collins AJ: Hematocrit level and associated mortality in hemodialysis patients. J Am Soc Nephrol 1996;10:610–619. 18 Churchill DN, Muirhead N, Goldstein M, Posen G, Fay W, Beecroft ML, Gorman J, Taylor DW: Effect of recombinant human erythropoietin on hospitalization of hemodialysis patients. Clin Nephrol 1995;43:184–188. 19 Xia H, Ebben J, Ma JZ, Collins AJ: Hematocrit levels and hospitalizations risks in hemodialysis patients. J Am Soc Nephrol 1999;10:1309– 1316. 20 Collins AJ, Keane WF: Higher hematocrit levels. Do they improve patient outcomes, and are they cost-effective. Nephrol Dial Transplant 1998;13:1627–1629. 21 Silverberg JS, Barre PE, Pritchard SS, Sniderman AD: Impact of left ventricular hypertrophy on survival in end-stage renal disease. Kidney Int 1989;36:286–290. 22 Erslev AJ, Caro J, Schuster SH: Is there an optimal hemoglobin level? Transfus Med Rev 1989;3:237–242. 23 London GM, Parfrey PS: Cardiac disease in chronic uremia: Pathogenesis. Adv Renal Replace Ther 1997;4:194–211. 24 Foley RN, Parfrey PS, Harnett JD, Kent GM, Murray DC, Barre PE: The impact of anemia on cardiomyopathy, morbidity, and mortality in end-stage renal disease. Am J Kidney Dis 1996;28:53–61. 25 Amann K, Breitbach M, Ritz E, Mall G: Myocyte/capillary mismatch in the heart of uremic patients. J Am Soc Nephrol 1998;9:1018– 1022. 26 Rostand SG, Kirk KA, Rutsky EA: Dialysisassociated ischemic heart disease: Insights from coronary angiography. Kidney Int 1984; 25:653–659. 27 Saragoca MA, Canziani ME, Cassiolato JL, Gil MA, Andrude JL, Draibe SA, Martinez EE: Left ventricular hypertrophy as a risk factor for arrhythmias in hemodialysis patients. J Cardiovasc Pharmacol 1991;17(suppl 2):S136– S138. 28 Mall G, Huther W, Schneider J, Lundin P, Ritz E: Diffuse intermyocardial fibrosis in uremic patients. Nephrol Dial Transplant 1990;5:39– 44. 29 Amann K, Schwartz U, Törnig J, Stein G, Ritz E: Anomalies cardiaques au cours de l’urémie chronique; in Hambuger J (ed): Actualités Néphrologiques Hôpital Necker. Paris, Flammarion, 1997, pp 1–15.

30 Amann K, Ritz E: Reduced cardiac ischaemia tolerance in uraemia – What is the role of structural abnormalities of the heart? Nephrol Dial Transplant 1996;11:1238–1241. 31 Levin A, Thompson CR, Ethier J: Left ventricular mass index increase in early renal disease: Impact of decline in hemoglobin. Am J Kidney Dis 1999;34:125–134. 32 Harnett JD, Kent GM, Foley RN, Parfrey PS: Cardiac function and hematocrit level. Am J Kidney Dis 1995;25:S3–S7. 33 Silverberg JS, Racine N, Barre PE, Sniderman AD: Regression of left ventricular hypertrophy in dialysis patients following correction of anemia with recombinant human erythropoietin. Can J Nephrol 1990;6:26–30. 34 Wizemann V, Schafer R, Kramer W: Followup of cardiac changes induced by anemia compensation in normotensive hemodialysis patients with left-ventricular hypertrophy. Nephron 1993;64:202–206. 35 Cannella G, La Canna G, Sandrini M, Gaggiotti M, Nordio G, Movilli E, Mombelloni S, Visioli O, Maiorca R: Reversal of left ventricular hypertrophy following recombinant human erythropoietin of anemic dialyzed uremic patients. Nephrol Dial Transplant 1991;6:31–37. 36 Rademacher J, Koch KM: Treatment of renal anemia by erythropoietin substitution. Clin Nephrol 1995;44(suppl 1):S56–S60. 37 Wizemann V, Kaufman N, Kramer W: Effect of erythropoietin on ischemic tolerance in anemic hemodialysis patients with confirmed coronary artery disease. Nephron 1992;62: 161–165. 38 Eschbach JW, Glenny R, Robertson T, Guthrie M, Rader B, Evans R, Chandler W, Davidson R, Eaterling T, Denney J, Schneider G: Normalizing the hematocrit in hemodialysis patients with EPO improves quality of life and is safe (abstract). J Am Soc Nephrol 1993;4:445. 39 Besarab A, Bolton WK, Browne JK, Egrie JC, Nissenson AR, Okamaoto DM, Schwab SC, Goodkin DA: The effects of normal versus anemic hematocrit on hemodialysis patients with cardiac disease. N Engl J Med 1998;339: 584–590. 40 Foley RN, Parfrey PS, Morgan J and the Canadian Normalization of Hemoglobin Study Group: A randomized controlled trial of complete versus partial correction of anemia in hemodialysis patients with asymptomatic concentric LV hypertrophy or LV dilatation. J Am Soc Nephrol 1998;9:208A. 41 Stombom U, Ahlmen J, Danielsson B: The Scandinavian Erythropoietin Study: Effects on quality of life of normalizing hemoglobin levels in uremic patients (abstract). J Am Soc Nephrol 1999;10:142. 42 Mason K, McMahon LP: Normalization of haemoglobin in hemodialysis patients: A comparative study. Nephrology 1997;3(suppl 1): S305.

Blood Purif 2001;19:168–174

173

43 Berns JS, Rudnick MR, Cohen RM, Bower JD, Wood BC: Effects of normal hematocrit on ambulatory blood pressure in epoetin-treated hemodialysis patients with cardiac disease. Kidney Int 1999;56:253–260. 44 Conlon PJ, Kovalik E, Schumm D, Minda S, Schwab SJ: Normalization of hematocrit in hemodialysis patients with cardiac disease does not increase blood pressure. Ren Fail 2000;22: 435–444. 45 Pickett JL, Theberge DC, Brown WS, Schweitzer SU, Nissenson AR, et al: Normalizing hematocrit in dialysis patients improves brain function. Am J Kidney Dis 1999;33: 1122–1130. 46 Metry G, Wilkstrom B, Valind S, et al: Effect of normalization of hematocrit on brain circulation and metabolism in hemodialysis patients. J Am Soc Nephrol 1999;10:854–863. 47 Klang B, Bjorvell H, Clyne N: Quality of life in predialytic uremic patients. Qual Life Res 1996;5:109–116. 48 Schidler NR: Quality of life and psychosocial relationships in patients with chronic renal insufficiency. Am J Kidney Dis 1998;32:557– 566. 49 Valderrabano F: Quality of life and hematocrit in hemodialysis patients. A prospective multicentre study. Second European Epoetin Symposium, Crete, Greece, April 17–19, 1998.

174

50 Moreno F, Sanz-Guajardo D, Lopez-Gomez JM, Jofre R, Valderrabano F: Increasing the hematocrit has a beneficial effect on quality of life and is safe in selected hemodialysis patients. J Am Soc Nephrol 2000;11:335–342. 51 Barany P: Treatment of anemia in hemodialysis patients to a normal hemoglobin concentration. Results of an open randomized clinical trial of epoetin beta. Scandinavian Multicentre Trial. J Am Soc Nephrol 1998;9:1234A. 52 Kettner-Melsheimer A, Weiss M, Huber W: Physical work capacity in chronic renal disease. Int J Artif Organs 1987;10:23–30. 53 Davenport A, Will EJ, Khanna SK, Davison AM: Blood lactate is reduced following successful treatment of anaemia in haemodialysis patients with recombinant human erythropoietin both at rest and after maximal exertion. Am J Nephrol 1992;12:357–362. 54 Lim VS: Recombinant human erythropoietin in predialysis patients. Am J Kidney Dis 1991; 18(suppl 1):34–37. 55 Barany P, Svedenhag J, Katzarski K, Divino Filno J, Norman R, Freyshuss U, Bergström J: Physiologic effects of correcting anemia in haemodialysis patients to a normal hemoglobin concentration (abstract). J Am Soc Nephrol 1996;7:1472.

Blood Purif 2001;19:168–174

56 Suzuki T, Tstsui MY, Okoyama A, Hirasawa Y: Normalization of hematocrit with recombinant human erythropoietin in chronic hemodialysis patients does not fully improve their exercise abilities. Artif Organs 1995;19:1258– 1261. 57 Stray-Gunderson J, Sams B, Goodkin D, Holloway D, Wang C, Thompson J: Improvement in functional capacity in dialysis patients with regular exercise and correction of anemia (abstract). J Am Soc Nephrol 1997;8:112A. 58 Silverberg JS, Barre PE, Pritchard SS, Sniderman AD: Impact of left ventricular hypertrophy on survival in end-stage renal disease. Kidney Int 1989;36:286–290. 59 Besarab A, Nasca T, Ross R: Erythropoietin in patients prior to end-stage renal disease. Curr Opin Nephrol Hypertens 1995;4:155–161. 60 Roth D, Smith RD, Schulman G, Steinman TI, Hatch FE, Rudnick MR, Sloand JA, Freedman BI, Williams WW Jr, Shadur CA, et al: Effects of recombinant human erythropoietin on renal function in chronic renal failure predialysis patients. Am J Kidney Dis 1994;23:777–784. 61 Kuriyama S, Tomonari H, Yoshida H, Hashimoto T, Kawagnuchi Y, Sakai O: Reversal of anemia by erythropoietin therapy retards the progression of chronic renal failure, especially in non-diabetic patients. Nephron 1997;77: 176–185.

Besarab/Aslam

Blood Purif 2001;19:175–178

The Role of Hemodialysis and Peritoneal Dialysis for the Early Initiation of Dialysis Gerald Schulman Vanderbilt University Medical Center, Nashville, Tenn., USA

Between the conception and the creation Between the emotion and the response falls the shadow T.S. Eliot

Introduction

Increasing attention to the goal of reducing the comorbid conditions that attend patients who are initiating dialysis for treatment of end-stage renal disease has led to the suggestion that dialytic therapy be started earlier in the course of chronic renal failure. Studies evaluating the impact of late referral of chronic renal failure patients and reports examining the relationship between the level of renal function at the time of initiation and subsequent clinical outcome and nutrition status have served as the impetus for advocating early dialysis [1–11]. There are two components that must be considered as one considers recommending either peritoneal dialysis or hemodialysis for early initiation. The first component is whether sufficient data exists to support the role of any type of dialysis therapy earlier in the course of chronic renal failure; the second component is whether peritoneal dialysis or hemodialysis is the best therapy to employ when one initiates therapy.

ABC

© 2001 S. Karger AG, Basel 0253–5068/01/0192–0175$17.50/0

Fax + 41 61 306 12 34 E-Mail [email protected] www.karger.com

Accessible online at: www.karger.com/journals/bpu

Early Initiation

Multiple lines of evidence support the role of early referral to a nephrologist for management once chronic renal insufficiency has been identified [4–6]. Survival and hospitalizations, once dialysis has been initiated, appear to be lower in the group of patients who are referred early (table 1). However, the data originate from studies that are observational. Although it seems intuitively reasonable, there have been no prospective trials to support the benefits of early referral. The patients referred late may have been noncompliant with their prior medical regimen, have had a greater number of comorbid conditions or differ in socioeconomic circumstances. Thus, further proof that early referral is beneficial would be welcome. The relation between level of residual renal function, nutritional status and outcome is also very compelling, but again this relation suffers from the lack of prospective controlled studies. While patients appear to decrease nutritional intake as renal function declines, no data exists to support the conclusion that early initiation will result in the correction of nutritional deficiency or that it will impact on outcome. Nevertheless, it is reasonable to assume that risk factors such as hypoalbuminemia, that

Gerarld Schulman, MD, Vanderbilt University Medical Center Department of Medicine, Division of Nephrology, S-3223 Medical Center North Nasville, TN 37232-2372 (USA) Tel. +1 615 936 1179, Fax +1 615 936 1308 E-Mail [email protected]

Table 1. Time to referral and impact on

outcome

Referral

n

Predialysis time

Residual renal function

Outcome

Early Late

263 194

13.5–49 months ! 1 month

F7 ml/min F5 ml/min

Increased hospitalization and death in late referral

are well established in the dialysis population, also begin to adversely effect outcome during the period leading up to the initiation of dialysis. To the extent that nutrition is improved by increasing clearance with an earlier start of dialysis, the patients may benefit. It has been cogently argued that since an adequate dose of peritoneal dialysis has been established to be a weekly Kt/V of 12.0, corresponding to a creatinine clearance of 9–14 ml/min, dialysis should be started to augment native clearance when it falls below this level [12]. Despite the lack of clear evidence, this practice may become more prevalent in the near future.

Hemodialysis or Peritoneal Dialysis for Incremental Dialysis?

Once it has been decided to initiate dialysis early in the course of renal insufficiency, should hemodialysis or peritoneal dialysis be the treatment of choice? Several studies have indicated that residual renal function is better preserved in patients undergoing peritoneal dialysis [13–15]. After 18 months, patients receiving hemodialysis had a 63% reduction in residual renal function compared to only an 11% decline in patients undergoing peritoneal dialysis [13]. On first inspection, it may seem difficult to justify the early application of hemodialysis to augment native renal clearance since it may impact adversely on residual renal function. If one decides to initiate dialytic therapy despite the lack of rigorous evidence that it is beneficial, then it is rational to employ gradually increasing peritoneal dialysis in this endeavor. A lack of enthusiasm for the continued use of peritoneal dialysis once residual renal function declines or to ever use it in patients who are referred for dialytic therapy late in the course of renal failure must also be considered as part of the issue. Several lines of evidence suggest that survival and well-being of patients on peritoneal dialysis depend on the presence of residual renal clearance and not on the clearance achieved by peritoneal exchange [16, 17]. This finding along with other problems with delivery of peritoneal dialysis must be part of any discussion of the

176

Blood Purif 2001;19:175–178

role of peritoneal dialysis. In this regard, the relative advantages of hemodialysis or peritoneal dialysis as modalities of treatment for end-stage renal disease should be considered. The most contentious matter regarding hemodialysis and peritoneal dialysis has been whether one of the therapies confers a survival advantage for the patient with endstage renal disease. This question has been addressed in many forums and in several studies. The studies that compare the two modalities suffer from differences in the statistical methods employed to assess outcome. Thus, results may differ depending on whether incident or prevalent patients are included in the analysis and whether treatment history or an intention-to-treat design is followed. In addition, differences in comorbidity between incident hemodialysis patients and peritoneal dialysis patients have been suggested (table 2). Finally, both hemodialysis and peritoneal dialysis therapies are evolving with respect to dose and technical improvements. In the past, the trend has clearly been for elderly, diabetic and female patients to have a lower mortality on hemodialysis [18–20]. However, there have been commendable improvements in the determination adequate levels of therapy in peritoneal dialysis – clearly ‘one size does not fit all’. The most recent analysis of the survival data, that also has attempted to address some of the disparities in previous studies, demonstrated that within the first 2 years of therapy, short-term peritoneal dialysis appears to be associated with superior outcomes compared with hemodialysis [21]. Given the better preservation of residual renal function with peritoneal dialysis mentioned above, along with these recent findings on survival, peritoneal dialysis would appear to be the rational choice for augmenting clearance as chronic renal failure progresses. However, an important exception to this may be for those patients who elect to be treated by APD. Patients on APD have a decline in residual renal function of –0.28 ml/min/month compared to patients on CAPD who have a decline of –0.1 ml/min/ month. This corresponds to a decline that is similar to hemodialysis patients. One might therefore argue that augmentation with peritoneal dialysis must be accom-

Schulman

Table 2. DMMS wave 2 baseline results

Peritoneal dialysis: 25% fewer CVA Peritoneal dialysis: 20% fewer congestive heart failure Peritoneal dialysis: 17% less pulmonary edema Peritoneal dialysis: Better prepared for dialysis

Table 3. Compliance in CAPD

Compliance with CAPD in the Canusa Study 23.5% missed 1 1 exchange/week and/or 1 2 exchanges/month Noncompliance increases with number of exchanges Noncompliance increases with exchange volume

plished by continuous exchanges to preserve residual renal function. A concern about advocating peritoneal dialysis for incremental therapy is that lack of evidence that it will affect outcome. In fact, this evidence may be hard to ever develop. In two large studies, total clearance (peritoneal plus residual renal function) and clearance derived from residual renal function, correlate with mortality but peritoneal clearance does not [16, 17]. These data suggest that residual renal function and peritoneal clearance may not be equivalent. One peritoneal clearance simply does not substitute for the other (residual renal clearance) on a oneto-one basis. One group of investigators has concluded ‘the presumption of equivalence may be appropriate if the concern were solute removal from the body rather than biological effects on the body. Adding the two clearances to determine whether treatment is adequate may be inappropriate – biologically speaking. The nature of the tradeoff points between peritoneal and renal clearance require clarification because they are not likely to be equivalent’ [17]. Moreover, a crossover in survival occurs that favors patients on hemodialysis after a variable period of time, coinciding with the rate of loss of residual renal function, has been shown in the same studies that have reported an initial advantage of peritoneal dialysis. Given these results in patients who have been receiving peritoneal dialysis as chronic maintenance therapy, it may be difficult to demonstrate an improved outcome that results from the incremental addition of one or two exchanges each day. This should be considered in any prospective trials undertaken to prove the benefits of incremental peritoneal dialysis on improved survival of patients with chronic renal failure.

Peritoneal Dialysis and Hemodialysis as Incremental Therapy for CRF

Beyond any controversy stimulated by the survival statistics, several other problems with the delivery of peritoneal dialysis exist. Compliance is a major problem with peritoneal dialysis [11, 22]. It can occur in as many as one third of the patients in a given center and contributes to technique failure and inadequate dialysis. It is also associated with complications such as peritonitis. Although most information comes from patients on full peritoneal dialysis therapy, can it be assumed that compliance will be better in those patients receiving it for augmentation? No data exist in this regard and caution is warranted (table 3). Cardiovascular complications may be greater in peritoneal dialysis patients. Rates of stroke and myocardial infarction have been reported to be greater. This corresponds to higher levels of cholesterol (total and LDL), triglycerides, PAI-1 and fibrinogen found in CAPD patients. Patients receiving peritoneal dialysis can be expected to have episodes of peritonitis with the attendant morbidity. In patients who fail peritoneal dialysis and ultimately transfer to hemodialysis, the 1-year mortality rate is 39%. Hemodialysis for augmentation therapy is likely to be performed at home. In that setting the newer, and as yet prospectively untested, methods of daily and/or nocturnal hemodialysis is likely to be employed rather than in-center based treatment. These treatment modalities incorporate several of the advantageous features that have been attributed to peritoneal dialysis. Hemodialysis is likely to be used in the home, at night, employing biocompatible membranes and of sufficient length to result in hemodynamic stability as well as to make middle molecule clearance feasible if high-flux membranes are utilized. When used for the augmentation of residual renal function, both hemodialysis and peritoneal dialysis are intermittent. Thus, hemodialysis may offer the same features as peritoneal dialysis. Given the dependence of peritoneal dialysis on residual renal function, the ultimate equivalence of the two treatments after a variable period of time and even the eventual superiority of hemodialysis in some groups in terms of survival, augmentation of residual renal function may be as reasonable as peritoneal dialysis.

Conclusions

Rigorous proof that early initiation is beneficial is still lacking, although numerous observational studies provide a rationale for considering augmentation as residual renal function declines. Peritoneal dialysis, because it is can be performed at home, can be easily applied in an incremen-

Blood Purif 2001;19:175–178

177

tal fashion and in the majority of patients has less of a detrimental effect on residual renal function, has been proposed as a superior mode of dialytic therapy for this endeavor. There is insufficient data to support this contention. Both hemodialysis and peritoneal dialysis are evolving and the technical aspects for delivery of optimal care are improving so that argument about the superiority on one or the other may be ‘full of sound and fury’ signifying very little. The home hemodialysis techniques of daily or nocturnal hemodialysis are just beginning to be applied and have the potential to offer the same features as perito-

neal dialysis. Compliance may be superior with hemodialysis. Hemodialysis is not a technique that depends on the presence of residual renal function for its salutary effects. Ultimately, perhaps as residual renal function declines, hemodialysis may be the preferred method of extracorporeal renal replacement therapy and thus patients should be introduced to this therapy from the start. T.S. Eliot’s shadow is still present. As with many aspects of the care of the patient with chronic renal failure, more study is necessary.

References 1 Churchill DN: An evidence-based approach to earlier initiation of dialysis. Am J Kidney Dis 1997;30:899–906. 2 Ratcliffe PJ, Phillips RE, Oliver DO: Late referral for maintenance dialysis. Br Med J 1984;288:441–443. 3 Innes A, Rowe A, Burden RP, Morgan AG: Early deaths on renal replacement therapy. Nephrol Dial Transplant 1992;7:467–471. 4 Jungers P, Zingraff J, Albouze G, Chaveau P, Page B, Hannedouche T, Man NK: Late referral to maintenance dialysis: Detrimental consequences. Nephrol Dial Transplant 1993; 8:1089–1093. 5 Kahn IH, Catto GRD, Edward N, Macleod AM: Death during the first 90 days of dialysis: A case control study. Am J Kidney Dis 1995; 25:276–280. 6 Sesso R, Belasco AG: Late diagnosis of chronic renal failure and mortality on maintenance dialysis. Nephrol Dial Transplant 1996;11:2417– 2420. 7 Modification of Diet in Renal Disease Study Group, prepared by Kopple J, Greene T, Cameroon J, Chumlea W, Hollinger D, Maroni B, Merrill D, Scherch L, Schulman G, Wang S, Zimmer G: Relationship between nutritional status and the glomerular filtration rate: Results from the MDRD Study. Kidney Int 2000; 57:1688–1703.

178

8 Bergstrom J: Nutrition and mortality in hemodialysis. J Am Soc Nephrol 1995;6:1329– 1341. 9 Bonomini V, Albertazzi A, Vangelista A, Bortolotti GC, Stefoni S, Scolari MP: Residual renal function and effective rehabilitation in chronic dialysis. Nephron 1976:16:89–99. 10 Bonomini V, Feletti C, Scolari MP, Stefoni S: Benefits of early initiation of dialysis. Kidney Int 1985;28:557–559. 11 Churchill DN, Taylor DW, Keshaviah PR, the CANUSA Peritoneal Dialysis Study Group: Adequacy of dialysis and nutrition in continuous peritoneal dialysis: Association with clinical outcomes. J Am Soc Nephrol 1996;7:198– 207. 12 Burkart JM, Schreiber M, Korbet SM: Solute clearance approach to adequacy of peritoneal dialysis. Perit Dial Int 1996;16:457–470. 13 Rottembourg J: Residual renal function and recovery of renal function in patients treated by CAPD. Kidney Int Suppl 1993;40:S106–S110. 14 Lysaght M, Vonesh E, Gotch F: The influence of dialysis treatment modality on the decline in remaining renal function. ASAIO Trans 1991; 37:598–601. 15 Moist LM, Port FK, Orzol SM: Predictors of loss of residual renal function among new dialysis patients. J Am Soc Nephrol 2000:11:556– 560. 16 Bargman JM, Thorpe KE, Churchill DN: The importance of residual renal function for survival in patients on peritoneal dialysis (abstract). J Am Soc Nephrol 1997;185A.

Blood Purif 2001;19:175–178

17 Diaz-Buxo JA, Lowrie EG, Lew NL, Hongyuan-Zhang SM, Xiaofei Zhu, Lazarus JM: Associates of mortality among peritoneal dialysis patients with special reference to peritoneal transport rates and solute clearance. Am J Kidney Dis 1999;33:523–534. 18 Bloembergen W, Port F, Mauger E, Wolfe R: A comparison of mortality between patients treated with hemodialysis and peritoneal dialysis. J Am Soc Nephrol 1995;6:177–183. 19 Fenton S, Schaubel D, Desmeules M, Morrison H, Mao Y, Copleston P, Jeffery J, Kjellstrand C: Hemodialysis versus peritoneal dialysis: A comparison of adjusted mortality rates. Am J Kidney Dis 1997;30:334–342. 20 Vonesh EF, Moran J: Mortality in end-stage renal disease: A reassessment of differences between patients treated with hemodialysis and peritoneal dialysis. J Am Soc Nephrol 1999;10: 354–365. 21 Collins AJ, Wenli H, Hong X, Ebben JP, Everson SE, Constantini EG, Ma JZ: Mortality risks of peritoneal dialysis and hemodialysis. Am J Kidney Dis 1999;34:1065–1074. 22 Bernardini J, Piraino B: Compliance in CAPD and CCPD patients as measured by supply inventories during home visits. Am J Kidney Dis 1998;31:101–107.

Schulman

Blood Purif 2001;19:179–184

Peritoneal Dialysis Should Be Considered as the First Line of Renal Replacement Therapy for Most ESRD Patients John M. Burkart Wake Forest University Baptist Center, Winston-Salem, N.C., USA

To PD or not to PD? How do we increase the life expectancy of the average patient with ESRD? Although life expectancies for our ESRD patients have been increasing, in the USA, they are still not that of the average citizen of the same age but without renal failure. It is true that our patients are becoming older and have more comorbid diseases, but are these diseases a result of our therapies? How do we optimize outcomes for dialysis patients while minimizing cost? These are just some of the questions those of us who provide renal replacement therapy should be asking ourselves as we enter this new millennium. We need to be able to offer a 30-year-old patient with ESRD from acute glomerulonephritis the same life expectancy he or she would have had if they did not have renal disease. Today, with our current approach to the treatment of ESRD, we may not be able to do that. Each of the current standard forms of renal replacement therapies, hemodialysis (HD), peritoneal dialysis (PD) and renal transplantation, have certain advantages. We should leverage these advantages at the appropriate time during the patients’ treatment course so that we optimize patient outcomes while minimizing cost. This is an approach often practiced but not well studied. Patients have often started on one therapy and subsequently transfer to another for various medical reasons, not necessarily a course planned from the initiation of therapy. Are there reasons to plan a course of therapy that uses all modalities? I suggest that there are. If so, what would be the best sequence of therapy choice? There is no doubt that if the

ABC

© 2001 S. Karger AG, Basel 0253–5068/01/0192–0179$17.50/0

Fax + 41 61 306 12 34 E-Mail [email protected] www.karger.com

Accessible online at: www.karger.com/journals/bpu

patient is not significantly uremic, a preemptive transplant would be best. More often than not however, that is not the case and some form of renal replacement therapy must be initiated. This article outlines the arguments for starting most patients, even those where transplantation is not an option, on PD first. The patients are then transferred to HD when PD no longer offers a therapeutic advantage over HD, or, at times when there is no survival advantage for one modality over another, when the patient elects to do so for lifestyle considerations. This approach has been coined the ‘integrated care’ approach to providing renal replacement therapies [1]. Reasons for considering starting all patients on PD first are: patient outcome data, the preservation of native renal function, the possibility of an incremental start of renal replacement therapy, access considerations in the long-term patient, benefits in the renal transplant recipient, qualityof-life issues, and cost constraints.

Patient Outcome Issues Mortality

PD has been considered by some to be an inferior therapy with survival rates less than that for similar HD patients [2]. However, there are many anecdotal patients who have survived on PD for 10–15 years [3]. Are these long-term PD patients merely a curiosity, or is this an outcome many patients can be expected to have? Data are

John Burkart, MD Department of Internal Medicine, Section of Nephrology Wake Forest University School of Medicine Medical Center Boulevard, Winston-Salem, NC 27157-1053 (USA) Tel. +1 336 716 3963, Fax +1 1 336 716 4318, E-Mail [email protected]

now emerging from properly done studies with correct statistical methodologies which suggest that outcomes of groups of similar patients are the same or better on PD than on HD. These data most interestingly suggest that over the first 1–2 years after the onset of ESRD that there actually appears to be a slight survival advantage for patients on PD. This advantage is seen in all age groups except elderly white female diabetics who seem to do worse on PD. Lets examine the evidence. Fenton et al. [4] reviewed data from an incident cohort of 11,970 Canadian patients who initiated dialysis during the calendar years 1990–1994 and found that PD patients had a 27% lower adjusted mortality rate (RR = 0.73; 95% confidence interval, 0.68–0.78). Most interestingly, during the first years of ESRD treatment the RR was lower on PD than on HD and only approached being equal after 2 years on therapy. Vonesh and Moran [5] re-analyzed data from the United States Renal Data Systems (USRDS) database using what they argue to be more appropriate statistical methods and found that over the same time period as that studied by Bloembergen et al. [2] that outcomes on PD were equal to or better than that on HD. Specifically, these authors found that the only subgroup that had both a clinically and statistically significant increased RR while on PD was in elderly female white diabetics. (Patients who were younger than 50 years old did better on PD). Collins et al. [6] evaluated outcome data on 99,048 incident HD patients and 18,110 incident CAPD/CCPD patients in the USRDS database, who initiated dialysis in 1994–1996 with follow-up until 6/30/97. A Poisson regression analysis which adjusted for age, gender, race and primary renal disease showed that outcomes on PD were similar to that on HD. Using a Cox regression analysis, they found that the RR of mortality for all nondiabetics were less on PD than if on HD and that this was also true for all young (!55-year-old) diabetics. Only elderly female diabetics had a higher RR on HD. They concluded that within the first 2 years of therapy, that PD seemed to offer superior outcomes when compared with HD. The reasons for this were unclear and they recommended more studies be carried out. Murphy et al. [7] evaluated outcomes in 822 consecutive ESRD patients at 11 Canadian institutions who started dialysis between 3/1993 and 11/1994 with a mean follow-up of 24 months. At baseline, 34% of patients were on PD. The co-morbidity score was higher at baseline, 3 and 6 months for the HD patients. After adjusting for differences the RR for PD and HD was basically the same,

180

Blood Purif 2001;19:179–184

suggesting that in Canada, there is no difference in the expected outcomes for patients on PD or HD. Van Biesen et al. [1] have shown that in their experience, patients that started on PD and then transferred to HD using an ‘integrated’ approach to modality choice had outcomes better than those who were treated with HD alone. These new data on mortality suggest that: (a) there is a possible early survival advantage for those who start on PD; (b) there is a possible increased RR to do PD in elderly white female diabetics (clinical significant or statistically significant?); (c) one should expect equal survival for most patients (all nondiabetics and young diabetics) over the first 2–4 years of treatment on either therapy.

Residual Renal Function Issues

The more residual renal function (RRF) one has at the time they start dialysis, the better their expected longterm outcomes [8]. This observation has significance when considering the timing for the initiation of dialysis and the effect the chosen dialysis modality may have on the further progression of residual renal damage. Historical data would suggest that PD has a beneficial effect on RRF. Lysaght et al. [9] showed that after time on dialysis, patients on PD had significantly more RRF than a subgroup of HD patients despite both groups having the same average RRF at the start of the observation period. It was uncertain if PD slowed the progression of RRF or if HD accelerated the loss. Subsequent studies have confirmed these observations [10, 11]. RRF is likely to have many benefits. Not only does RRF contribute to overall solute clearance and volume removal, but it does so in a continuous manner. Patients with significant urine volumes are more likely to be euvolemic and have better controlled blood pressures. This should have a positive effect on left ventricular function [12], reducing the risk of developing left ventricular hypertrophy (LVH). LVH is very prevalent in both the pre-ESRD and ESRD populations [13]. It has been shown that current dialysis populations have a markedly increased RR of dying from an early cardiovascular event [14]. Although still not proven, attempts to reduce the development of LVH in dialysis patients should reduce this RR [15]. Taken together, these observations on RRF would suggest that attempts to preserve RRF would likely result in improved patient outcomes and PD seems to be the modality most likely to accomplish this.

Burkart

Incremental Start of Dialysis

The NKF-DOQI guidelines recommend that one should consider initiation of some form of renal replacement therapy once residual renal Kt/V is !2.0/week unless certain conditions are met [16]. These certain conditions include lack of uremic symptoms and assurance that the patient does not have any early signs of malnutrition. Although the guidelines do specify when to initiate dialysis, they do not specify how to initiate dialysis. Should one initiate dialysis with a ‘full dose’ of therapy, ignoring the residual renal clearance in solute clearance calculations, or should one gradually increase the dialysisrelated solute clearance as RRF is lost maintaining minimal total solute clearances at all times? There are no outcome data to give an evidenced base answer to this question. Keshaviah et al. [17] first addressed the issue of a ‘timely’, incremental start to the initiation of dialysis predicting that CAPD regimens could be successfully adjusted to achieve a constant total Kt/V for 2–5 years. Clinical data using this approach has been prospectively collected in three centers [18–20] and retrospectively in one center [21]. All found that it was feasible to start dialysis in a timely manner using an ‘incremental’ approach. Interestingly, in patients who had been followed on PD for at least 12 months, 2/7 were still on 2 or less exchanges/day in one study [18] while 9/24 were on 2 or less exchanges in another [19]. In the one study where BP results were available, there was a marked improvement with the initiation of dialysis associated with a reduction in the number of anti-hypertensive medications taken per day. Published data on the clinical use of incremental HD is limited. In one of the studies above, published in abstract form only [18], the authors reported on their initial experience with ‘incremental’ initiation of dialysis with HD. Only 3 patients were studied and all had to go to full dose (3 treatments/week) of dialysis within a relatively short time (by 4 months after initiation) for either adequacy or for volume control issues. Although the numbers are small, this would be consistent with what was predicted by the DOQI guidelines where although there were guidelines for how to initiate the therapy incrementally with HD, it was obvious that once the RRF Kt/V was !2.0/ week, that 2/week HD would be difficult due to the high Kt/V per treatment needed to achieve minimal total solute clearance targets. Although there were PD-related complications in all studies, these were often able to be handled without the need for hospitalization or transfer to

Peritoneal Dialysis as the First Line of Renal Replacement Therapy

HD because of the RRF which made the need for acute dialysis less necessary. As will be discussed below, this approach tended to allow the patients a better quality of life with minimal intrusion and was therefore well accepted. Cost issues are uncertain. Certainly if the patient is on dialysis for a longer time, the cost for dialysis is more. However, as has been shown in earlier studies, if you maintain health and are able to start dialysis as an outpatient, the overall long-term cost for the system will be less. In has been shown that with advanced chronic renal insufficiency, blood pressure control is very difficult unless one can adequately control blood volume [22]. A timely, incremental but continuous approach to dialysis such as PD can offer may be beneficial. Due to the potential beneficial effects on RRF and quality of life issues, if one is practicing an ‘incremental’ approach to the initiation of dialysis, current data would favor doing that with PD.

Access Issues in the Long-Term Patient

The creation and maintenance of a problem-free vascular access for HD remains a major problem [23]. PD has less access related problems and in the patient who will potentially need to do some form of dialysis for 20–30 years, preserves the need for immediate vascular access, augmenting the total lifespan of easy access for the patient. Up to 50% of new ESRD patients in the USA have no permanent vascular access at the time they initiate therapy [24]. This means they are likely to need a temporary access such as a Perm-a-Cath or Vasc-Cath, both of which are associated with high rates of infectious and clotting complications. Access-related complications may be responsible for up to 50% of the hospitalization costs in ESRD patients [25]. With meticulous attention to Tenckhoff catheter placement and exit care with the use of prophylactic topical antibiotics to reduce the risk of Staphylococcus aureus-related infections, the average PD catheter should have a 80%+ 3 year survival rate [26]. One should be able to achieve very low overall infection rates in PD patients, most treated as an outpatients [27].

Issues Related to Transplantation

Arguably a renal transplant is the best option for most patients with ESRD [28]. However, not all patients are candidates and most are not preemptively transplanted, so most transplant recipients must first be on some form

Blood Purif 2001;19:179–184

181

of renal replacement therapy. Does the modality the patient is on pretransplant offer graft or patient survival advantages? Does the pretransplant modality influence morbidity or cost? Our group has reviewed data from the United Network of Organ Sharing (UNOS) in the USA and found that of 9,291 patients who received a cadaveric transplant between 4/1/94 and end 31/12/95 that PD patients were more likely to produce urine in the first 24 h posttransplant and were less likely to need dialysis than HD patients both before and after adjustment for other risks associated with delayed graft function [29]. There were no differences in early rejection episodes. Fontan et al. [30] had previously reported a similar observation. Analysis of 119 clinical records of patients transplanted in Belgium showed that delayed graft function was seen less often in those on PD than on patients who had been on HD (6/40 vs. 27/79, p = 0.03) while acute renal failure was also seen less often in those patients who had been on PD, (14/79 HD vs. 0/40 PD, p = 0.01) [31]. Although it is hard to find transplant-specific data, there are data which associate the use of temporary vascular access for HD with infections such as bacteremia, endocarditis, perispinal abscesses, etc., all rarely seen on PD. These have been anecdotally seen at our center in HD patients using a temporary vascular access prior to transplantation but not in those on PD. Patients with delayed graft function posttransplant have been shown in some studies to have a lower 1-year graft survival rate when compared to patients with immediate function [32, 33]. Delayed graft function may also have a detrimental long-term affect on renal function [34], and it increases the cost of initial hospitalization by about USD 15,000 [35]. The reasons for these observed differences are primarily speculation. Perhaps related to volume shifts with HD, cytokine production with HD, vascular instability, bio-incompatablity [36], etc. One could argue that the peritoneal membrane is certainly more ‘biocompatible’ than most membranes used for HD. To prevent or minimize delayed graft function would likely be beneficial long term. The data suggest that PD can do this.

Quality-of-Life Issues

Our goal in providing renal replacement therapy is not only to prolong life and maintain health but also to sustain the quality of the patients’ life [37]. Mental health scores measured with the SF-36 questionnaire were closer to that

182

Blood Purif 2001;19:179–184

of the general population in patients on PD than those starting on HD, although these differences disappeared after 18 months. Others evaluated 16,755 HD patients and 1,260 PD patients (728 on CAPD, 532 on CCPD) during 1996 [38]. They found that before adjustment for variables, there was no difference in scoring between groups. However, after adjustment, CAPD and CCPD patients scored higher for mental processes. CCPD patients scored worse on physical function, but higher for mental processes than either CAPD or HD patients. These observations were confirmed in a meta-analysis of recent reports [39]. However, these authors caution that some differences may be due to differences in case mix. A survey conducted by the National Kidney Foundation showed that patients on automated PD or CAPD were more satisfied with their therapy than patients on HD were (97 vs. 85% vs. 77%) [40], while patients on PD were more likely to be employed [41].

Cardiovascular Issues

Providing the entire weekly clearance of fluids and metabolites in 12 h (instead of the 148 h used by continuous treatments or natural renal function) results in substantial stress to the cardiovascular system. Some have documented electrocardiographic changes, which included an increase in premature ventricular contractions during treatment [42, 43], increase in QT dispersion [44], and cardiac ischemia manifesting as ST segment depression with electrocardiography [45] during HD treatment. The latter changes have been predictive of cardiac death. In addition, life-threatening hypotension is unfortunately a common occurrence in HD. These changes are exacerbated by the creation of a 7day week, and the 3-day interval that results. Volume overload and hyperkalemia are especially likely to develop after the increased interval. Moreover, the weekend is associated with increased recreational activities and likely increased food consumption. Potentially contributing to this is the fact that sudden death is more common in the general population on Monday [46]. We obtained data on the date of death, cause of death, and dialysis modality for the entire US dialysis population from 1977 through 1997 [47]. In addition, data was obtained from the Case Mix Adequacy Study of 7,096 HD patients. The date of laboratory blood draws was used as a surrogate for the day of dialysis. If blood draws occurred on Monday, patients were designated as Monday, Wednesday, Friday (MWF) patients. Tuesday, Thursday, Saturday (TTS) patients

Burkart

were similarly assigned. The cause of death was ascertained from the notifying physician. Control deaths were considered to be deaths from infection, malignancy, gastrointestinal bleeding, or other conditions unrelated to cardiac death. We further evaluated the time of death for patients from our own unit over a 2-year period. At the midpoint of this study, there were 206 HD patients and 64 PD patients. Families were interviewed as to the time and nature of death. Dialysis records were also reviewed. Data from the USRDS included 326,728 HD patient deaths and 48,754 peritoneal dialysis patient deaths. For hemodialysis patients, there was a markedly increased risk of cardiac death (18.1% vs. 14.3% expected, p ! 0.00001) on Monday. There was also a markedly increased risk of sudden death on Monday (18.1% vs. 14.3% expected, p ! 0.00001). The control death rate was constant throughout the week. In comparison for PD patients, the death rate was quite steady throughout the week, with the Monday cardiac death rate of 14.8% vs. 14.3% expected. For the patients from the Case Mix Adequacy Study, where the day of HD could be identified, the sudden death rate on Monday was 20.8% (vs. 14.3% expected, p = 0.002), and the cardiac death rate was 20.2% vs. 14.3% expected, p = 0.0005. For TTS patients, there was an increased cardiac death rate on Tuesday (18.5% vs. 14.3% expected, p = 0.03) and an increased sudden death rate (18.1% vs. 14.3%, p = 0.10). In dialysis patients studied from our units, there were 29 instances of sudden death, accounting for 40% of all deaths. Forty-two percent of patients died while on dialysis or !10 h after completing dialysis. 47% of deaths occurred 140 h after the last dialysis. Only 3% of deaths occurred between 10 and 40 h after dialysis. The above results demonstrate how the intermittent nature of HD may affect the occurrence of sudden death. With PD one should be able to avoid the complications related to the intermittent nature of HD, and if one is careful about blood pressure and volume control in PD patients as RRF decreases perhaps one could also prevent development of LVH. Gaggiotti et al. [48] have demonstrated in a group of 67 case-controlled ESRD patients (30

CAPD, 37 HD) with similar baseline estimates of left ventricular function were similar (left ventricular mass index and left ventricular end-diastolic volume) and blood pressure, that after 12 months of therapy, there was a statistically significant reduction in systolic blood pressure, diastolic blood pressure, left ventricular mass index and left ventricular end-diastolic volume in PD but not HD patients. These short-term data would support that concept.

Cost Constraints

There is no doubt that our number one obligation in providing renal replacement therapy for our patients with ESRD is to do the right thing and offer the best therapy to achieve optimal outcomes. Although economic pressures are real, we must also be fiscally responsible. Three studies have attempted to compare total yearly costs for PD and HD and all have some flaws. Nevertheless, all three suggest that PD is more cost-effective than HD (USD 41,256 vs. 52,716 [49], USD 33,781 vs. 48,351 [50] and USD 39,040 vs. 63,608 [51]). Cost data suggest that the policy of PD first would save society money when caring for ESRD patients.

Conclusions

It is time for us to renew our approach to the initiation of dialysis. This paradigm shift includes not only a reevaluation of the timing of the initiation of dialysis but also a careful look at what modality we choose for most patients. Recent data show that when educated and given a choice, up to 45–50% of patients choose PD [52, 53]. If there are data which suggest that long term the patients may do better with a more integrated approach to the initiation of dialysis, considering PD first in the majority of our patients, we may be able to further improve outcomes while further reducing overall costs. I feel the above mentioned data suggests we should consider this approach.

References 1 Van Biesen WM, Vanholder RC, Veys N, Dhondt A, Lameire NH: An evaluation of an integrated care approach for end-stage renal disease patients. J Am Soc Nephrol 2000;11: 116–125.

Peritoneal Dialysis as the First Line of Renal Replacement Therapy

2 Bloembergen WE, Port FK, Mauger EA, Wolfe RA: A comparison of mortality between patients treated with hemodialysis and peritoneal dialysis. J Am Soc Nephrol 1995;6:177–183. 3 Burkart JM: Prevention and treatment of peritoneal membrane failure in the long term PD patient. Adv Ren Replace Ther 1998;5:153– 156.

4 Fenton SSA, Schaubel DE, Desmeules M, Morrison HI, Mao Y, Copleston P, Jeffrey JR, Kjelestrand CM: Hemodialysis versus peritoneal dialysis. A comparison of adjusted mortality rates. Am J Kidney Dis 1997;30:334–342.

Blood Purif 2001;19:179–184

183

5 Vonesh EF, Moran J: Mortality in end-stage renal disease: A reassessment of differences between patients treated with hemodialysis and peritoneal dialysis. J Am Soc Nephrol 1999;10: 354–365. 6 Collins AJ, Hao W, Xia H, Ebben JP, Everson SE, Constantini EG, Ma JZ: Mortality risks of peritoneal dialysis and hemodialysis Am J Kidney Dis 1999;34:1065–1074. 7 Murphy SW, Foley RN, Barret BJ, Kent GM, Morgan J, Barre P, Campbell P, Fine A, Goldstein MB, Handa SP, Jindal KK, Levin A, Mandin H, Muirhead N, Richardson RMA, Parfrey PS: Comparative mortality of hemodialysis and peritoneal dialysis in Canada. Kidney Int 2000;57:1720–1726. 8 Canada-USA (Canusa) Peritoneal Dialysis Study Group: Adequacy of dialysis and nutrition in continuous peritoneal dialysis. J Am Soc Nephrol 1996;7:198–207. 9 Lysaght MJ, Vonesh EF, Gotch F: The influence of dialysis treatment modality on the decline of residual renal function. ASAIO J 1991;3:598. 10 Rottembourg J: Residual renal function and recovery of function in patients treated by CAPD. Kidney Int 1993;43(suppl 40):S106– 110. 11 Cancarini GC, Brunori G, Camerini G, Brassa A, Manili L, Miaorca R: Renal function recovery and maintenance of residual diuresis in CAPD and hemodialysis. Perit Dial Bull 1986; 6:77–79. 12 Levin A, Singer J, Thompson CR, Ross H, Lewis M: Prevalent left ventricular hypertophy in the predialysis population. Identifying opportunities for intervention. Am J Kidney Dis 1996;27:347–354. 13 Foley RN, Perfrey PS, Harnett JD, Kent GM, Martin CJ, Murray DC, Barre PE: Clinical and echocardiographic disease in patients starting end-stage renal disease therapy. Kidney Int 1995;47:186–192. 14 Levy AS: Controlling the epidemic of cardiovascular disease in chronic renal disease: Where do we start? Am J Kidney Dis 1998;32 (suppl 3):S5–S13. 15 Coles GA: Have we underestimated the importance of fluid balance for the survival of PD patients. Perit Dial Int 1997;17:321–326. 16 NKF-DOQI clinical practice guidelines for peritoneal dialysis adequacy. Am J Kidney Dis 1997;30(suppl 2):S67–S131. 17 Keshaviah PR, Emerson PF, Nolph KD: Timely initiation of dialysis: A urea kinetic approach. Am J Kidney Dis 1999;33:344–348. 18 Burkart JM, Satko SG: Incremental initiation of dialysis: One center’s experience over a 2year period. Perit Dial Int 2000;20:418–422. 19 Williams PF: Timely initiation of dialysis (letter). Am J Kidney Dis 1999;34:594–595. 20 De Vecchi AF, Scalamogna A, Finazzi S, Colucci P, Ponticelli C: Preliminary evaluation of incremental peritoneal dialysis in 25 patients. Perit Dial Int 2000; 20:412–417.

184

21 Andersson H, Buxbaum M, Meisil FT: Continuous adaptation of the dialysis prescription maintains adequate Kprt/V in CAPD patients. Perit Dial Int 2000;20:423–428. 22 Lazarus JM, Bourgoinie JJ, Buckalew VM, Greene T, Levey AS, Milas NC, Paranandi L, Peterson JC, Porush JG, Rauch S, Soucie JM, Stollar C: Achievement and safety of low blood pressure goal in chronic renal disease. Hypertension 1997;29:641–650. 23 Hakim R, Himmelfarb J: Hemodialysis access failure: A call to action. Kidney Int 1998;54: 1029–1040. 24 Hakim R, Lazarus J: Initiation of dialysis. J Am Soc Nephrol 1995;6:1319–1328. 25 Feldman H, Korbrin S, Wasserstein A: Hemodialysis vascular access morbidity. J Am Soc Nephrol 1996;7:523–535. 26 Gokal R, Alexander S, Ash S, et al: Peritoneal catheters and exit site practices, toward optimal peritoneal access. 1998 update. Perit Dial Int 1998;18:11. 27 Casey MJ, Burkart JM: Topical application of mupirocin at the exit site reduces exit site infections and peritonitis. Perit Dial Int 2000, in press. 28 Port FK, Wolfe RA, Mauger EA, Berling DP, Jiang K: Comparison of survival probabilities for dialysis patients vs. cadaveric reanal transplant recipients. JAMA 1993;270:1339–1343. 29 Bleyer AJ, Burkarrt JM, Russell GB, Adams PL: Dialysis modality and delayed graft function after cadaveric renal transplantation. J Am Soc Nephrol 1999;10:154–159. 30 Fontan MP, Rodriguez-Carmona A, Bouza P, Falcon TG, Adevera M, Valdes F, Oliver J: Delayed graft function after renal transplantation in patients undergoing peritoneal and hemodialysis. Adv Perit Dial 1996;12:101–104. 31 Van Biesen W, Vanholder R, Van Loo A, Van der Vennet M, Lamiere N: Peritoneal dialysis favorably influences early graft function after renal transplantation compared to hemodialysis. Transplantation 2000;69:508–514. 32 Ojo A, Wolfe R, Held P, Port F, Schmouder R: Delayed graft function: Risk factors and implications for renal allograft survival. Transplantation 1997;63:968. 33 Nicholson M, Wheatley T, Horsburg T, Edwards C, Veitch P, Bell P: The relative influence of delayed graft function and acute rejection on renal transplant survival. Transplant Int 1996;9:415. 34 Giral-Classe M, Hourmant M, Cantarvorich D, Dantal J, Blancho G, Daguin P, Ancelet D, Soulollou JP: Delayed graft function of more than six days strongly decreases long-term survival of transplanted kidneys. Kidney Int 1998; 54:972–978. 35 Almond PS, Troppman C, Escobar F, Frey DJ, Matas AJ: Economic impact of delayed graft function. Transplant Proc 1991;23:1304. 36 Hakim RM, Wingard RL, Parker RA: Effect of the dialysis membrane in the treatment of patients with acute renal failure. N Engl J Med 1994;331:1338–1342.

Blood Purif 2001;19:179–184

37 Merkus M, Jager K, Dekker F, de Haan R, Krediet R: Quality of life in patients on chronic dialysis: Self-assessment 3 months after start of treatment. Am J Kidney Dis 1997;29:584– 592. 38 Diaz-Buxo JA, Lowrie EG, Lew NL, Hongyaun Z, Lazarus JM: Quality of life evaluation using short form 36: Comparison in hemodialysis and peritoneal dialysis patients. Am J Kidney Dis 2000;35:293–300. 39 Cameron JI, Hiteside C, Katz J, Devins G: Differences in quality of life across renal replacement therapies: A meta-analytic comparison. Am J Kidney Dis 2000;35:629–637. 40 Wolcott, et al: NKF report 1997. 41 Julius M, Kneisley JD, Carpentier-Alting P, et al: A comparison of employment rates of patients treated with CAPD vs. in-center hemodialysis. Arch Intern Med 1989;149:839–842. 42 Morrison G, Michelson EL, Brown S, Morganroth J: Mechanism and prevention of cardiac arrhythmias in chronic hemodialysis patients. Kidney Int 1980;17:811–819. 43 Redaelli B, Locatelli F, Limido D, Andrulli S, Signorini MG, Sforzini S, Bonoldi L, Vincenti A, Cerutti S, Orlandini G: Effect of a new model of hemodialysis potassium removal on the control of ventricular arrhythmias. Kidney Int 1996;50:609–617. 44 Lorincz I, Matyus J, Zilahi Z, Kun CC, Karanyi Z, Kakuk G: QT dispersion in patients with end-stage renal failure and during hemodialysis. J Am Soc Nephrol 1999;10:1297–1302. 45 Nakamura S, Uzu T, Inenaga T, Kimura G: Prediction of coronary artery disease and cardiac events using electrocardiographic changes during hemodialysis. Am J Kidney Dis 2000; 36:592–599. 46 Willich SN, Lowel H, Lewis M, Hormann A, Arntz HR, Keil U: Weekly variation in the incidence of sudden cardiac death in the Framingham Heart Study population. Am J Cardiol 1987;60:801–806. 47 Bleyer AJ, Russell GB, Satko SG: Sudden and cardiac death rates in hemodialysis patients. Kidney Int 1999;55:1553–1559. 48 Gaggiotti M, Cancarinin GC, Verzeletti F, Maiorca R: Long term cardioprotective effect of CAPD: Perspective study with HD. Blood Purif 1997;15(suppl 2):6–7. 49 Westman J, George S, Scheel PJ Jr, McMurray SD, Pulliam J: Options for dialysis providers in a globally capitated environment. Nephrol New Issues 1996;10:26–31. 50 McMurray S: Impact of capitation on a free standing unit: Can you survive? Am J Kidney Dis 1997;30:542–548. 51 Burns E, Saul M, Seddon P, Zeidel M: Cost of caring for a dialysis patient: Physician fees and modality. J Am Soc Nephrol 1998;8:187A. 52 Pritchard S: Treatment modality selection in 150 consecutive patients starting ESRD therapy. Perit Dial Int 1996;16:69–72. 53 Schreiber, et al: Preliminary results from the National Pre-ESRD Education Program, Abstract NKF Spring Meeting, 2000.

Burkart

Blood Purif 2001;19:185–188

Technology: Tools or Toys, Is It Economically Feasible with Current Reimbursement? The Case in Favor James Tattersall Sheffield Kidney Institute, Sheffield, UK

Definitions

Tool: ‘Item of equipment which is considered necessary to carry out a trade or profession’ Toy: ‘A thing meant rather for amusement than for serious use’

Introduction

For the majority of the world population, any kind of dialysis technology is economically unfeasible [1]. Only in relatively few affluent countries is dialysis freely available to all ESRD patients. In the United Kingdom, for example, dialysis is seen as poor value for money [2] and has a low priority for healthcare spending. Most UK dialysis units are unable to obtain sufficient funding to provide

adequate dialysis to all their ESRD patients [3]. Even in the USA, reimbursement is provided at such a low rate that it is very difficult for units to provide adequate treatment without the bulk purchasing power and central management of a larger provider chain. The ideal of dialysis is to replace the function of the failed kidney as completely as possible with minimal inconvenience to the patient, minimal cost to society and with full rehabilitation of the patient. It is clear that current dialysis techniques, with costs around USD 50,000 per year per patient [4, 5], 5-year survival around 30% [6] and patient quality of life seriously impaired [7] (table 1), fall far short of this ideal. The natural kidney provides its homeostatic and clearance function continuously within the body. Current dialysis techniques provide around 10% of the clearance power of the natural kidney. Three times weekly hemodialysis

Table 1. Results of the British quality of

life survey (% values) Social life seriously affected Leisure life seriously affected Sex life seriously affected

ABC

© 2001 S. Karger AG, Basel 0253–5068/01/0192–0185$17.50/0

Fax + 41 61 306 12 34 E-Mail [email protected] www.karger.com

Accessible online at: www.karger.com/journals/bpu

Hospital HD

Home HD

CAPD

Transplant General population

74.2 71.9 67.9

63.2 58.9 67.9

65.2 68.5 68.7

21.5 25.5 30.8

James Tattersall, MD Mediqal Limited, Suite F, Astonbury Business Park Stevenage S627EG (UK) Tel. +44 1438 880 190, Fax +44 1438 880 188 E-Mail [email protected]

11.9 12.1 7.8

Table 2. Steps in the development of an artificial kidney

Three times weekly hemodialysis

Large, requires extensive support services Labor-intensive Safe in experienced hands Daily home dialysis Self-contained | Simple to use Overnight daily dialysis Foolproof, more biocompatible Wearable continuous dialysis Intelligent, small Implantable artificial kidney | Completely automatic

has particular problems due to its intermittency. The fluctuations in fluid state make it very difficult to control blood pressure and volume over the weekly cycle. Phosphate control is impossible without strict compliance with a phosphate binding or restriction regime during the 2- or 3-day periods between dialysis. It is hard to see how we can justify the term ‘adequate’ in current dialysis.

Dialysis Intermittency

In order to improve patient convenience and reduce cost, dialysis is normally provided in three weekly sessions around 3.5–4.5 h each. The short dialysis is associated with an increased risk of underdialysis, fluid overload and hypertension, all factors associated with increased mortality and morbidity. Conversely, longer dialysis (8– 12 h) has been shown to maintain blood pressure within the normal range without drugs [8] and improve survival [9]. Daily dialysis has been shown to normalize blood phosphate without the need for phosphate binders in addition to improved blood pressure control. Longer or more frequent dialysis strategies have not been generally accepted as they are considered to be impractical in the hospital setting. However, if the dialysis is delivered at home (or nursing home), longer more frequent regimes become practical [10]. At present, home dialysis is not favored because the technology available is either ineffective or unsuitable for this setting, especially for high-risk or elderly patients. Dialysis technology needs to be developed further to make it really practical and convenient for routine home use for the typical dialysis patient. Alternatively, if we are to continue with the short-dialysis policy, we must use technology to solve its shortcomings.

186

Blood Purif 2001;19:185–188

Towards the Implantable Artificial Kidney

The early dialysis pioneers always considered that dialysis technology would evolve to the point that it could be implanted within the body and work continuously; a real artificial kidney [11]. A dialysis machine, which was selfcontained within a suitcase, was described in the 1970s and considered to be a step in development of a wearable artificial kidney [12]. The steps of this development process are outlined in table 2. Example of current technology evolving towards more efficient, more continuous, biocompatible and practical home therapies include automated peritoneal dialysis (APD) and the AkSys home dialysis system [13]. The costbenefit ratio and efficiency of APD may be further enhanced by on-line production of PD fluid at the bedside using technology originally developed for on-line hemodiafiltration [14].

We Need Technology Toys

Many of today’s indispensable dialysis technology tools were initially considered to be toys by the nephrology community. The use of mathematical modeling of hemodialysis, for example, and Kt/V was considered to be a toy for at least 10 years after it was first proposed in the late 70s. Today, no one would argue that urea kinetic modeling is not a valuable, even essential tool. Access recirculation measurement technology was also considered to be a toy until the more traditional measurement methods were discredited by evidence.

Blood Temperature Modeling

Some technology toys may never become tools but nevertheless contribute significantly to higher quality treatment and improved cost-effectiveness. The use of blood temperature modeling is an example of this. This technology allows us to control or measure the thermal energy flux between the patient and dialysis machine. By playing with this technology, we have learned that vascular stability can safely be improved by lowering the dialysate temperature. We do not need the blood temperature monitoring technology to lower the dialysate temperature but its use has resulted in more effective treatments at no additional cost. Some of today’s technology toys may eventually evolve into tomorrow’s tools but even today may help to improve quality and cost-effectiveness of dialysis (fig. 1).

Tattersall

described. None of these have been sufficiently well validated to be universally accepted. However, experience gained from playing with this technology has shown us that the standard, constant ultrafiltration rate will generally cause hypotension at the end of dialysis and will make it difficult to achieve dry weight. Linearly-decreasing ultrafiltration to a low rate at the end of dialysis reduces these complications and could easily be a standard feature of dialysis without additional cost or complexity [16].

On-Line Hemofiltration and Hemodiafiltration

Fig. 1. The life of a technology toy.

Blood Volume Monitoring

The author has observed a dramatic improvement in blood pressure control, reduction in the use of antihypertensive drugs and reduction in the incidence of dialysis hypotensive episodes following introduction of the CritLine blood volume monitor in our dialysis unit. This technology is intended to provide early warning of impending hypotension due to ultrafiltration-induced low blood volume [15]. This use has not been validated by rigorous study, and in my experience simply does not work in this way. Nevertheless, routine use of the Crit-Line monitor increased awareness of fluid and blood volume issues and is a valuable educational tool for clinical staff and patients. Observation of the blood volume curves during dialysis provided a valuable insight into the patient’s response to fluid removal and helps in the determination of dry weight. Eventually, insights gained from playing in this way will allow us to develop a true dry weight computer and ultrafiltration controller. These tools can be used to reduce the mortality associated with intermittent hospital dialysis and to make home dialysis safer and more effective.

Ultrafiltration Profiling

In an attempt to improve tolerance to ultrafiltration, various ultrafiltration profiles have been developed. Various models to assist in choosing the profile and biofeedback systems for controlling the ultrafiltration have been

We Need Technology Tools

Hemofiltration and hemodiafiltration with on-line production of the substitution fluid are also examples of a technology which have a sound theoretical basis but the clinical benefits have yet to be proven conclusively. These modalities were originally developed to increase vascular stability and the clearance of ß2-microglobulin. Today, it is not clear whether these benefits justify the cost. There is now increasing evidence that chronic inflammation due to bioincompatibility and endotoxin transfer from dialysis fluids into the blood is part of the cause of the cardiovascular disease, which causes death in the majority of dialysis patients [17]. The development of on-line hemofiltration systems involved the reliable and safe production of endotoxin-free fluid at low cost. Even if these modalities are never routinely adopted, the means to produce endotoxin-free fluid is already being used to reduce the inflammation of dialysis and promises to significantly reduce dialysis morbidity and mortality.

Conclusion

Today’s reimbursement systems are motivated by the need to increase cost-effectiveness rather than simple cost-containment. The systems are definitely not intended to stifle the technological development, which will continue to improve quality and reduce cost. The reimbursement system can encourage clinical cost-effectiveness but cannot guide the development itself; this is firmly the responsibility of the nephrology community. To continue with today’s dialysis technology with its high cost and poor outcome is not a viable long-term option. The dialysis companies are driven by the market. If there is no market for new and experimental technology then it will not be developed. Nephrologists have an ethical responsibility not to lose interest in developing new technology.

Blood Purif 2001;19:185–188

187

References 1 Rao M, Juneja R, Shirly RB, Jacob CK: Haemodialysis for end-stage renal disease in Southern India – A perspective from a tertiary referral care centre. Nephrol Dial Transplant 1998;13:2494–500. 2 Haycox A, Jones D: The cost effectiveness of renal provision in the UK. J Manag Med 1996; 10:6–15. 3 Stanton J: The cost of living: Kidney dialysis, rationing and health economics in Britain, 1965–1996. Soc Sci Med 1999;49:1169–1182. 4 Bruns FJ, Seddon P, Saul M, Zeidel ML: The cost of caring for end-stage kidney disease patients: An analysis based on hospital financial transaction records. J Am Soc Nephrol 1998;9: 884–890. 5 Rodriguez-Carmona A, Perez FM, Bouza P, Garcia FT, Valdes F: The economic cost of dialysis: A comparison between peritoneal dialysis and in-center hemodialysis in a Spanish unit. Adv Perit Dial 1996;12:93–96. 6 Fenton SS, Schaubel DE, Desmeules M, Morrison HI, Mao Y, Copleston P, Jeffery JR, Kjellstrand CM: Hemodialysis versus peritoneal dialysis: A comparison of adjusted mortality rates. Am J Kidney Dis 1997;30:334–342.

188

7 Gudex CM: Health-related quality of life in end-stage renal failure. Qual Life Res 1995;4: 359–366. 8 Covic A, Goldsmith DJ, Venning MC, Ackrill P: Long-hours home haemodialysis – the best renal replacement therapy method? QJM 1999; 92:251–260. 9 Innes A, Charra B, Burden RP, Morgan AG, Laurent G: The effect of long, slow haemodialysis on patient survival. Nephrol Dial Transplant 1999;14:919–922. 10 Raj DS, Charra B, Pierratos A, Work J: In search of ideal hemodialysis: Is prolonged frequent dialysis the answer? Am J Kidney Dis 1999;34:597–610. 11 Kolff WJ: Exponential growth and future of artificial organs. Artif Organs 1977;1:8–18. 12 Briefel GR, Galonsky RS, Hutchisson JT, Hessert RL, Friedman EA: Field trial of compact travel dialysis system. J Dial 1976;1:57–66.

Blood Purif 2001;19:185–188

13 Kenley RS: Tearing down the barriers to daily home hemodialysis and achieving the highest value renal therapy through holistic product design. Adv Ren Replace Ther 1996;3:137– 146. 14 Brunkhorst R, Fromm S, Wrenger E, Berke A, Petersen R, Riede G, Westphale J, Zamore E, Ledebo I: Automated peritoneal dialysis with ‘on-line’-prepared bicarbonate-buffered dialysate: Technique and first clinical experiences. Nephrol Dial Transplant 1998;13:3189–3192. 15 Steuer RR, Leypoldt JK, Cheung AK, Senekjian HO, Conis JM: Reducing symptoms during hemodialysis by continuously monitoring the hematocrit. Am J Kidney Dis 1996;27: 525–532. 16 Donauer J, Kolblin D, Bek M, Krause A, Bohler J: Ultrafiltration profiling and measurement of relative blood volume as strategies to reduce hemodialysis-related side effects. Am J Kidney Dis 2000;36:115–123. 17 Don BR, Kaysen GA: Assessment of inflammation and nutrition in patients with end-stage renal disease. J Nephrol 2000;13:249–259.

Tattersall

Blood Purif 2001;19:189–194

A Simplified Method for Adequate Hemodialysis Werner Kleophas Gangolf Backus Dialysezentrum Karlstrasse, Düsseldorf, Germany

Introduction

Over the past three decades, several components have been introduced into hemodialysis technology in order to improve the quality of the treatment. The first hemodialysis systems were all batch systems with the known problems like bacterial contamination and the mixing of fresh and spent dialysate. The introduction of single-pass systems tried to overcome these disadvantages and offered safe treatments to a constantly increasing number of dialysis patients. Despite of these well-known advantages, first contributions to technical requirements were made regardless of physiological principles. The introduction of acetate buffer instead of bicarbonate and glucose-free dialysis solution are examples for this development and describe the beginning of an area where the idea of blood purification became more and more prominent. The next generation of hemodialysis technique offered higher flows and clearances and led to a shortening of treatment time. Complex technical solutions became necessary to allow the re-introduction of bicarbonate and glucose. In addition, efforts were made to avoid the inherent risks of high efficacy treatment procedures like frequent hypotensive episodes and disequilibrium. A large number of technical components were added to standard hemodialysis machines allowing continuous monitoring of blood volume

ABC

© 2001 S. Karger AG, Basel 0253–5068/01/0192–0189$17.50/0

Fax + 41 61 306 12 34 E-Mail [email protected] www.karger.com

Accessible online at: www.karger.com/journals/bpu

and pressure, hematocrit, temperature and Kt/V. This led to ever more sophisticated systems. In contrary to this trend the question arises if at least the same results can be achieved with a simplified hemodialysis tool as well. This is a welcome opportunity (1) to discuss the physiological demands upon hemodialysis technique, (2) to describe a simplified system for the routine use in hemodialysis, (3) to speculate on the relative costs of such a system and (4) on its potential role in the future of dialysis technique.

Physiological Demands upon Hemodialysis Technique

The composition of the extracellular fluid is very similar to that of ocean water [1]. This is an evolutionary hint at the first way of waste product elimination. The ancient cell had just to throw away its waste products and due to dilution in the ancient ocean there was no intoxication (fig. 1). Later on the relation of body cell mass to extracellular fluid changed from 1:∞ to 1:0.25 (fig. 2). Homer W. Smith [2] pointed out that only animals capable of regulating solute substances in their body water were able to settle successfully ecological spaces. Already in the middle of the 19th century, Claude Bernard [3] stated that ‘all

Dr. Werner Kleophas Dialysezentrum Karlstrasse 17–19 D–40210 Düsseldorf (Germany) Tel. +49 211 1679791, Fax +49 211 1679797 E-Mail [email protected]

Fig. 1. Ancient cells: relation from cell mass to extracellular vol-

Fig. 2. Human cells: relation from cell mass to extracellular volume.

ume.

Fig. 3. Human cells in hemodialysis: temporary enlargement of the

extracellular volume by the volume of dialysis fluid.

vital mechanisms however varied they may be have only one subject, that of preserving constant the conditions of life in the internal environment.’ ‘La fixeté du milieu interieur est la condition de la vie libre.’ Based on these ideas, Walter Canon [4] formulated the term of ‘homeostasis’ describing the organisms’ way of regulating the extracellular fluid. These principles were

190

Blood Purif 2001;19:189–194

taken by Homer W. Smith [5] to define the kidney as the regulatory organ of the extracellular fluid. ‘It may fairly be said that this regulatory function, so long overlooked, is just as important as the excretion of waste products of metabolism or of foreign substances, which hitherto has received nearly all the emphasis.’ If this is true for the normal kidney it should also be the guideline for the construction of an artificial kidney. Following this, the artificial kidney should be regarded as a regulatory system and not only as an apparatus for waste product elimination. In hemodialysis the patient’s body opens itself via a thin membrane to the dialysis fluid. It exposes its internal environment to that fluid for the duration of the treatment. The relation from body cell mass to extracellular fluid is enlarged to 1:3 (fig. 3). In that sense hemodialysis fluid should be considered as the temporary extension of the extracellular fluid [6]. It is obvious that from this physiological view the hemodialysis technique has to satisfy certain basic requirements: (1) The dialysis fluid is the crucial part of the whole treatment. Its importance has been underlined early [7–9] and it could be regarded as a ‘drug’ [6]. It should be composed in accordance to the need of the extracellular fluid, that means it should be regarded as an individual fluid of high hygienic standard and all kinds of hemodialysis machines must be able to deliver this fluid. (2) Keeping constant the conditions of life in the internal environment means to avoid fast shifts wherever it is

Kleophas/Backus

possible. This is a demand upon treatment time rather than on hemodialysis technique. However, most of modern hemodialysis machines are designed for high-flow treatments to shorten the time. (3) Regulation of the extracellular fluid. That means: excretion of waste products and toxins; supplementation (e.g., bicarbonate, calcium); keeping balance, i.e., no disturbance of equilibrium (e.g., glucose, vitamins, amino acids), and ultrafiltraion.

A Simplified Batch-Type Dialysis Machine for Routine Use in Hemodialysis

This system was developed by the late B. Tersteegen (1939–1995). Based on the principals formulated by Claude Bernard [3] and Homer W. Smith [5] it was originally named ‘milieu intérieur regulator’. Today it is marketed under the trade name of Genius® from Fresenius Medical Care. The principal design is shown in figure 4. Before each treatment, fresh dialysis fluid is prepared according to the physician’s prescription by mixing mostly dry ingredients with preheated ultrapure water. The whole amount of dialysis fluid is filled into the thermally insulated glass tank of the dialysis machine so that no integrated heater is necessary.

Individual Dialysis Fluid

Fig. 4. Schematics of the batch dialysis machine.

Bacteriological Safety

Conventional dialysis systems need complex logistics to allow more than 50 different dialysis compositions. Many machines offer the possibility of sodium or bicarbonate profiling. This is a different approach to ‘a tailored’ dialysis. All these profiles however must be designed in advance [6]. The decision depends on previous clinical observations. Although profiling is done online, it represents a static view of the patient and does not correspond to the real changes in that specific dialysis session. They are not designed for variation of dialysis composition but for better tolerance of ultrafiltration. In the Genius® system the individual composition of the dialysis fluid is done offline, i.e., before the start of the treatment. The variation of dialysis fluid composition is far superior to those of proportioning systems. Due to the offline production of the dialysis fluid, no complex technical equipment is necessary. Quality management is easy and safe since nobody is able to change the composition of the dialysis fluid without interruption of the session once the treatment had started.

Alarmed by the high frequency of pyrogenic reactions during hemodialysis, the Association of the Advancement of Medical Instrumentation (AAMI) formulated guidelines in 1991 to limit acceptable values of bacterial contamination to 200 colony-forming units (CFU)/ml for incoming water and 2,000 CFU/ml for dialysate. Recent discussions about chronic and microinflammatory disease in hemodialysis patients [10–12] stated that ultrapure dialysate is a very important requirement. The Genius® system offers an intrinsic management of hygienic problems without any additional filters. This was achieved by: (1) A special water treatment to deliver ultrapure water. (2) The use of mostly dry concentrates to produce the dialysis fluid. The risk of bacterial contamination is a known hazard of liquid concentrates. (3) A new design of connectors to avoid the well-known problems with standard coupling elements. (4) A special design in all dialysis fluid pathways: the use of (a) an integrated UV radiator to suppress bacterial growth on all fluid-contacted surfaces, (b) inert materials and surfaces in the dial-

Simplified Hemodialysis

Blood Purif 2001;19:189–194

191

Fig. 5. Survival data, Düsseldorf Karlstrasse

(Kaplan-Meier), all patients in comparison to other series.

ysis circuit (only glass and stainless steel), and (c) a special geometry to avoid sharp corners or dead spaces. (5) A closed system. It is not opened during normal operation. Contamination from the environment is impossible. With these hygienic properties the system reaches near zero germ and pyrogen rates documented by Lonnemann et al. [13].

Volume Control

Since the system consists of a completely filled rigid tank (volume = 75 liters) it provides a simple and reliable method of volumetrically balancing. Only fluid that has been generated by ultrafiltration can leave the system under the control of an ultrafiltration monitor. For the same reasons no pressure measurement is necessary on the blood side. Only one pressure monitor on the dialysate side satisfies all safety needs.

Blood Temperature Regulation

Thermal conditions heavily influence cardiovascular parameters in dialysis patients. A dialysate temperature of 37–38 ° C leads to an increase in body temperature. Due to a related vasodilatation hypotensive episodes may occur frequently. In modern hemodialysis systems, blood temperature monitors are included to calculate a thermal balance. As a closed-loop control these systems are able to regulate the dialysis fluid temperature in order to avoid an increase of blood temperature during the treatment. In

192

Blood Purif 2001;19:189–194

the Genius® system the batch is filled with preheated ultrapure dialysis fluid (38 ° C). Due to the thermal insulation there is a temperature drop of 0.7 ° C/h so that at the end of the treatment the blood of the patient is exposed to a cooler dialysis fluid. This is a simple inherent mechanism for the regulation of thermal balance.

Treatment Procedure

Technical designs of the system have been described [14–16]. The Genius® batch machine receives fresh dialysis fluid from a special automated filling station. Once the machine is filled it can be moved to any treatment room since it has a battery pack inside and no further water or energy equipment is necessary for treatment. Fresh dialysis fluid is taken out of the top compartment of the tank, used dialysate is returned to the bottom compartment. Because of a difference in temperature and weight, used and fresh dialysate do not mix so that the system combines the simplicity of the batch system with the efficacy of a single-pass system.

Clinical Experience

In an uncontrolled series with 399 patients, we reported about long-term survival in our center which was observed about 10–20% higher than in other published series [15] (fig. 5). High serum albumin levels (82% of all patients 14.0 g/dl) and an extremely low cumulative inci-

Kleophas/Backus

Table 1. Blood pressure stability

MAP pre-HD MAP post-HD Hypertensive patients2 Sessions with hypotension3

All patients1 (n = 86)

Nondiabetics (n = 79)

Diabetics (n = 17)

95.9B14.7 (47–152) 88.3B15.1 (47–163) 39.5% (n = 34) 5.2% (n = 54)

95.6B14.6 (47–147) 88.3B15.6 (47–163) 40.6% (n = 28) 4.5% (n = 37)

96.7B14.9 (70–152) 88.2B13.1 (53–133) 35.3% (n = 6) 8.2% (n = 17)

1 Patients on regular treatment 61 year. Dialysis sessions n = 1,031 unselected during 1 month. Patients’ weight: 68.4 B 13.4 kg, weight gain: 2.8 B 1.07 kg, dialysis time: 4.4 B 1.27 h. 2 Patients with blood pressure 1 140/90 or with administration of at least one antihypertensive. 3 Sessions with blood pressure ! 100/70 and with intervention of nurse.

dence of a carpaltunnel syndrome (7%, 10 years under risk) are possibly due to the absence of a microinflammatory state by using ultrapure dialysis fluid. 65% of all patients achieved an acceptable blood pressure (predialytic value !140/90 mm Hg) without any additional antihypertensive drugs. A very low incidence of hypotensive episodes (table 1) is possibly due to the inherent temperature modelling of the system.

Cost-Related Benefits of the System

A batch system does not need a water distribution pipe-work. Instead it needs a certain filling station. The patient treatment and the filling and emptying procedure are separated. This enables to shift a higher proportion of work from experienced nurses to helpers resulting in lower staff costs. The absence of complex technology like multiple monitors and biofeedback systems will do this as well since sophisticated technology is asking for ever more specialized nurses. There is no need for additional disposables for ultrapure dialysis fluid. Since the volume of the batch is limited to 75 liters, costs of water, concentrates and energy are lower than in a conventional system. In general, follow-up costs will be lower in a lean system in comparison to a sophisticated system with a high number of components.

the treatment. Nevertheless this development has led to more and more sophisticated systems in hemodialysis treatment. It is of interest that the best survival data are achieved in the Tassin group [17] where over years none of those devices have been introduced. The duration of a single treatment session is always a point of discussion in that context since in the Tassin group patients are treated in long sessions of 8 h with low flows. Routine use of the described batch-type dialysis system is asking for low flows and longer treatments due to the limitation of the dialysis fluid. This is in contrast to conventional proportioning systems which are designed for high-flow treatments in order to save time. If the Genius® system has potential benefit of low infrastructure costs this could be used for financing longer treatment sessions. Treatments with high ultrafiltration rates and fast shifts mean dialyzing at the edge of hypotension and other complications. In spite of many undoubted advantages, many components in high-tech hemodialysis systems seem to be necessary for the prevention of self-made symptoms. Simplified dialysis systems may possibly also play an important role in new therapeutic concepts like daily home hemodialysis.

Conclusion

Over the past years several new components have been introduced into hemodialysis technology. There is no doubt that many of these devices improved the quality of

Simplified Hemodialysis

Blood Purif 2001;19:189–194

193

References 1 PJ: Endocrines and Osmoregulation. Berlin, Springer, 1971. 2 Smith HW: From Fish to Philosopher. Boston, Little & Brown, 1953. 3 Bernard C: Leçons sur les phénomènes de la vie commune aux animaux et aux végétaux. Paris, Baillière, 1885. 4 Cannon WB: Organization for physiological homeostasis. Physiol Rev 1929;9:399. 5 Smith HW: The Kidney: Structure and Function in Health and Disease. New York, Oxford University Press, 1951. 6 Ronco L, Fabris A, Feriani M: Hemodialysis fluid composition; in Jacobs C, Kyllstrand CM, Koch KM, Winchester JF (eds): Replacement of Renal Function by Dialysis, ed 4. Dordrecht, Kluwer, 1996, pp 256–276. 7 Alwall N: On the artificial kidney. I. Apparatus for dialysis of the blood in vivo. Acta Med Scand 1947;128:317.

194

8 Aoyama S, Kolff W: Treatment of renal failure with the disposable kidney. Results of fifty-two patients. Am J Med 1957;23:565. 9 Coleman BK, Merrill JP: The artificial kindey. Am J Nurs 1952;52:327. 10 Zimmermann I, Herrlinger S, Pruy A, Metzger T, Wanner C: Inflammation enhances cardiovascular risk and mortality in hemodialysis patients. Kidney Int 1999;55:648–658. 11 Baz M, Durand C, Ragon A, et al: Using ultrapure water in hemodialysis delays carpal tunnel syndrome. Int J Artif Organs 1991;14:681– 685. 12 Lonnemann G, Krautzig S, Koch KM: Quality of water in hemodialysis. Nephrol Dial Bentley Transplant 1996;11:946–949. 13 Lonnemann G, Dumann H, Schmidt-Gürtler H: Improved dialysate quality is associated with decreased whole blood cytokine production in ESRD patients on the Genius® hemodialysis system. Blood Purif 1997;15:6.

Blood Purif 2001;19:189–194

14 Polaschegg HD, Levin NW: Hemodialysis machines and monitors; in Jacobs C, et al (eds): Replacement of Renal Function by Dialysis, ed 4. Dordrecht, Kluwer, 1996, pp 333–379. 15 Kleophas W, Haastert B, Backus G, Hilgers P, Westhoff A, van Endert G: Long-term experience with an ultrapure individual dialysis fluid with a batch type machine. Nephrol Dial Transplant 1998;13:3118–3125. 16 Fassbinder W: Renaissance of the batch methods? Nephrol Dial Transplant 1998;13:3010– 3012. 17 Charra B, Calemard E, Ruffet M, Charot L, Laurent G: Survival as an index of adequacy of dialysis. Kidney Int 1992;41:1286–1291. 18 Charra B, Calemard E, Chazot C: Dose of dialysis: What index? Blood Purif 1992;10:13–21.

Kleophas/Backus

Blood Purif 2001;19:195–199

Electrolyte Balancing: Modern Techniques and Outcome Francesco Locatelli Celestina Manzoni Salvatore Di Filippo Department of Nephrology and Dialysis, ‘A. Manzoni’ Hospital, Lecco, Italy

Introduction

Electrolyte balancing is a very important function of hemodialysis, besides the removal of interdialytic water load and the clearance of uremic toxins of different sizes. As far as electrolyte balancing is concerned, modern techniques seem to have recently contributed promising results, particularly in the more adequate handling of dialysate sodium and potassium. That is of special relevance because of the possibility for it to significantly affect the clinical outcome of dialysis patients, at present burdened by a still too high morbidity and mortality rate. Cardiovascular diseases account for more than 50% of mortality in hemodialysis patients, who have an incidence of cardiac death that is 5–10 times higher than that of the agematched general population. Hypertension is a common feature in ESRD patients and it is certainly a major cause of cardiomyopathy and cardiac morbidity in these patients. On the other hand a major clinical problem facing us in dialysis management is the high incidence of intradialytic cardiovascular instability, for some extent due to the progressive increase in the age of the patients starting substitutive therapy and in the median dialytic age of existing patients, with a consequent rise in cardiovascular risk factors. The adequate intradialytic sodium removal is an invaluable tool not only in contrasting overhydration and its serious adverse cardiovascular effects (pulmonary oedema, hypertension, myocardial hypertrophy and there-

ABC

© 2001 S. Karger AG, Basel 0253–5068/01/0192–0195$17.50/0

Fax + 41 61 306 12 34 E-Mail [email protected] www.karger.com

Accessible online at: www.karger.com/journals/bpu

after dilation) but also in improving intradialytic cardiovascular stability. The extent of convective sodium removal is inevitably dependent on the ultrafiltration rate, whereas the diffusive flux is strictly related to the dialysate sodium concentration freely chosen by the clinician; it is therefore clear that this choice is fraught with considerable clinical consequences. During hyponatric dialysis, the sodium loss down the concentration gradient may be excessive, and thus lead to intracellular overhydration, an excessive reduction in extracellular volume, and intradialytic discomfort [1]; on the other hand, hypertension and overhydration can occur when too high dialysate sodium concentrations are used without taking the appropriate precautions [2]. It is therefore important to choose the ‘adequate’ dialysate sodium concentration, i.e. the concentration allowing the matching of intradialytic removal and the interdialytic sodium load. Intradialytic hydrosodium removal is defined as being adequate in relation to the interdialytic intake when the hydrosodium balance is zero, i.e. the final dry body weight is reached and the final plasma water sodium concentration is constant. For this purpose, the required water removal can be easily quantified because it corresponds to the interdialytic increase in body weight. On the contrary, given that it is affected by various factors, a mathematical kinetic model must be used at the start of treatment in order to predict final plasma water sodium concentration and calculate the needed dialysate sodium concentration.

Prof. Francesco Locatelli, MD Divisione di Nefrologia e Dialisi, Ospedale ‘A. Manzoni’ Via dell’Eremo, 9/11 I–23900 Lecco (Italy) Tel. +39 0341489850, Fax +39 0341489860, E-Mail [email protected]

A single-pool kinetic model making it possible to calculate the dialysate sodium concentration needed to achieve the desired end-dialysis plasma water sodium concentration was first proposed by Gotch et al. [3] in 1980. By determining the difference between the end-dialysis plasma water sodium concentrations predicted by the model (using flame photometry to determine plasma and dialysate sodium concentrations) and the measured values in 13 hemodialysis sessions (involving 6 patients) it has been shown that the model had an imprecision of B2.9 mEq/l. When using this kinetic model, modified in using blood and dialysate sodium activities determined by means of the more precise direct potentiometry, the model imprecision in reaching a pre-established target of end-dialysis plasma water sodium activity was of !0.84 mEq/l [4]. Unfortunately, although they make it possible to reach the desired intradialytic sodium removal, these models are unsuitable for routine clinical application because the need to know the initial sodium plasma water concentration and sodium dialysance in real time is very demanding for medical and nursing staffs. Now modern techniques could give us a hand in overcoming these difficulties.

dule (COT Hospal) connected to the dialysate line between the dialyzer and the dialysis machine. By means of a single temperature-compensated conductivity probe, which was alternately activated at the dialysate inlet and outlet, the Module measures the difference between inlet and outlet dialysate conductivity values before and after a change in inlet dialysate conductivity of about 1 mS/cm over a short period of about 2 min, thus determining sodium dialysance as ionic dialysance and sodium concentration as plasma water conductivity. Moreover, the Module is capable of automatically controlling inlet dialysate conductivity (according to the single-pool conductivity kinetic model) in order to achieve the prescribed end-dialysis plasma water conductivity. The only data required are the patient’s initial body weight, body weight loss and treatment time. The results of this study show that the accuracy of the conductivity kinetic model is good (a mean difference between observed and predicted values of –0.04 mS/cm) with an imprecision of !0.14 mS/cm, roughly equivalent to !1.4 mEq/l in terms of sodium concentration. The conductivity kinetic model may therefore be used instead of the sodium kinetic model, and the fact that it does not require any blood sampling or laboratory determinations makes it suitable for the routine clinical application.

Sodium Modeling by Conductivity

According to the basic theory developed by Polashegg [5], if dialysate conductivity is measured at the dialyzer inlet and outlet ports at two different inlet values, sodium dialysance can be easily estimated according to equation 1 (see the Appendix for equations 1 and 2). After estimating ionic dialysance, plasma water sodium concentration can be easily derived as plasma water conductivity using equation 2. Given the linear correlation between plasma water conductivity and sodium concentration, the conductivity values can be used instead of sodium concentration values making it possible to apply a kinetic model by using only conductivity values. The sodium kinetic model is therefore changed to the conductivity kinetic model [6] which, without the need for any blood sampling or laboratory determinations, makes it possible to predict final plasma water conductivity when dialysate conductivity is known or to determine the dialysate conductivity needed to obtain a desired final plasma water conductivity. The conductivity kinetic model has been validated in 57 hemodialysis sessions scheduled to obtain end-dialysis plasma conductivity values of between 14.0 and 14.8 mS/ cm [7]. The dialysis monitor (Monitral S Hospal) was equipped with the specially designed Biofeedback Mo-

196

Blood Purif 2001;19:195–199

Sodium Modeling by Conductivity in Paired Filtration Dialysis (PFD)

The sodium single-pool kinetic model developed for hemodialysis cannot be correctly applied to hemodiafiltration techniques because higher ultrafiltration rates may interfere with dialysate sodium concentrations. PFD is a hemodialfiltration technique in which convection and diffusion take place separately by means of a hemofilter and a hemodialyzer combined in a single unit [8]; it can therefore be considered as the combination of postdilution hemofiltration and mainly diffusive hemodialysis. A PFD single-pool sodium kinetic model has been developed by combining the equations for instantaneous sodium fluxes in hemodialysis and postdilution hemofiltration [9], with blood and dialysate sodium concentrations being determined by direct potentiometry. Clinical results have confirmed the validity of this PFD sodium singlepool kinetic model: the mean difference between the expected and measured end PFD plasma water-ionized sodium concentrations was 0.00 B 0.55 mEq/l, which means that the model is very accurate and its imprecision in predicting final plasma water sodium concentration is

Locatelli/Manzoni/Di Filippo

!1.1 mEq/l and nearly equivalent to that of the hemodialysis model. On the basis of the linear relationship between ultrafiltrate conductivity and plasma water sodium concentration values, a PFD single-pool conductivity kinetic model was also developed [9]. The validity of this model as an alternative to the sodium kinetic model in optimizing sodium removal has been confirmed in clinical tests. The mean difference between the predicted and measured end PFD ultrafiltrate conductivity was –0.01 B 0.05 mS/cm, an imprecision of !0.1 mS/cm. The greater accuracy and precision of the PFD model than the hemodialysis model is not surprising if it is considered that plasma water conductivity is measured in the ultrafiltrate, but only estimated from ionic dialysance in the hemodialysis kinetic model. On the basis of the regression of the ionized plasma water sodium concentrations in relation to ultrafiltrate conductivity values, the error in predicting bloodionized sodium concentrations by means of ultrafiltrate conductivity measurements is !2 mEq/l. These results demonstrate that the conductivity kinetic model is a reliable and inexpensive method for matching intradialytic hydrosodium removal with interdialytic load at each session.

Conductivity and Clinical Outcomes

As stated above, unacceptably high cardiovascular mortality is still the leading cause of death among dialyzed patients and, since hypertension is one of the main determinants of left ventricular hypertrophy (and later possibly myocardial dilation), blood pressure control is an essential part of dialysis treatment. Even if the pathogenetic role of greater blood and extracellular volume is currently a matter of debate, it is common experience that anthypertensive drugs are often unsuccessful in dialysis patients, whereas blood pressure can be controlled in the majority of cases by means of salt restriction and adequate ultrafiltration during dialysis. Furthermore, the hypotension that is the most frequent intradialytic complication is strictly related to intradialytic blood volume changes that mainly depend on sodium removal. The importance of finding a method for accurately modulating sodium removal at each dialysis session lies in the fact that it could reduce intradialytic hypotension and the other effects of sodium depletion, while simultaneously preventing overhydration and its possible side effects. The great accuracy of the conductivity kinetic model in estimating intradialytic sodium removal and its simplicity

Modern Techniques and Electrolyte Balancing

and inexpensiveness could make it a precious tool in reducing cardiovascular morbidity and mortality. A multicenter, prospective, randomized cross-over study has been recently performed aimed at testing whether cardiovascular stability could be improved by using the on-line conductivity ultrafiltrate kinetic modeling simply to reduce variability in sodium balance [10]. Forty-nine uremic patients on chronic 3 times weekly hemodialysis treatment, who had been affected by symptomatic hypotension during three or more PFD sessions in the month preceding study entry, were recruited from 16 participating centers. The 16-week study involved two treatments (A = conventional PFD; B = PFD using the kinetic model) and two sequences (1 = ABB and 2 = BAA), with a 4-week run-in period (A) followed by three consecutive 4-week experimental periods. The standard treatment was performed using fixed dialysate conductivity whereas the experimental treatment used the dialysate conductivity derived from the conductivity kinetic model, in order to obtain an end-dialysis ultrafiltrate conductivity at each dialysis session that was equal to the mean value determined in the same patient during the run-in period. In this way, sodium removal should exactly match the interdialytic sodium load at each dialysis, and thus possibly reduce the negative clinical effects of too much or too little sodium removal related to the variability in sodium intake from one session to another. The results of this study showed that the application of the conductivity kinetic model significantly reduced the intradialytic drop in systolic blood pressure in comparison with standard treatment (p = 0.001), without any period or carryover effect; this treatment effect was maximal at the third hour (23% less than during standard treatment), with no effect on the intradialytic diastolic blood pressure profile. Moreover, there was also a steady trend towards a reduction in the frequency of intradialytic symptoms, as well as in asymptomatic or symptomatic hypotension (fig. 1); this is consistent with a trend towards better cardiovascular stability, although at different non-significant p values. There was no difference between the two treatments in terms of mean predialysis and end-dialysis body weight, or in the ultrafiltrate and dialysate conductivity values: the average estimated sodium balance was therefore similar between the two treatments. In accordance with the study design, end-dialysis ultrafiltrate conductivity was the same for the two treatments; only the variability of this value was lower during the experimental treatment, and should be the key factor related to the better cardiovascular stability. The not very large but highly significant effect on cardiovascular stability observed in the study

Blood Purif 2001;19:195–199

197

Fig. 1. Percent of dialysis sessions characterized by hypotension

(symptomatic or asymptomatic) with conventional PFD (A) and PFD with conductivity kinetic model (B) in the two treatment sequences.

was possibly due to the fact that sodium intake did not vary much from one session to another; a greater benefit in terms of intradialytic cardiovascular stability can be expected when the conductivity kinetic model is applied in the case of patients with a more variable interdialytic sodium intake. A further very interesting clinical application of the model could be the better definition of dry body weight. By allowing a known stepped increase or decrease in dialytic sodium removal, the model allows a slowly progressive modification of the sodium pool. A prospective, controlled randomized study is currently being carried out in order to verify whether this improves blood pressure control in hypertensive patients and the tolerance of dialysis in patients with cardiovascular instability.

Potassium Modeling

During dialysis, potassium is removed by diffusion throughout the dialyzer membrane depending on the concentration gradient between the plasma water and dialysate potassium levels. During standard hemodialysis, which has a constant dialysate potassium concentration, this gradient – and consequently potassium removal – is highest at the beginning and decreases rapidly in the first hour of the treatment and slowly in the following hours. Given that diffusive fluxes of potassium are one of the main factors affecting electric potential that is related to

198

Blood Purif 2001;19:195–199

cardiac activity, intradialytic modifications may modify myocardial cell excitability and lead to dangerous arrhythmias, particularly in the first hour of treatment. On the other hand, higher dialysate potassium concentrations may reduce the incidence of arrhythmias, but do not allow adequate potassium removal. In the attempt to clarify whether a new model of hemodialysis potassium removal, based on a decreasing intradialytic potassium concentration and a constant plasma-dialysate potassium gradient is capable of reducing the arrhythmogenic effect of standard hemodialysis, a multicenter, prospective, randomized cross-over study has been performed, using a machine modified for potassium modeling, in 42 chronic patients who showed an increase in premature ventricular complexes (PVC) during dialysis [11]. The results showed that the new method led to a reduction of 36% in PVC/h and of 32% in PVC couplets/h in comparison with the standard treatment. This result was more evident during the first hour of hemodialysis and was obtained without adversely affecting predialysis plasma potassium levels.

Conclusions

Although only future studies will allow a complete exploration of the potential clinical benefits of the more extensive use of conductivity and potassium modeling, the bulk of the results obtained so far demonstrates their clinical relevance in handling the sodium and potassium pools. The same could become true in the future for other electrolytes (as bicarbonate). When also considering how promising the on-line determination of ionic dialysance in routine assessment of delivered dialysis dose is, one can fully understand how modern techniques will be able to radically change everyday clinical practice and hopefully improve dialysis patients’ outcome in the near future.

Appendix Determination of ionic dialysance (D) according to Polashegg: D (ml/min) = –Qd !

(Cdi1 – Cdo1) – (Cdi2 – Cdo2) Cdi1 – Cdi2

(1)

where Qd = inlet dialysate flux (ml/min); Cd = dialysate conductivity (mS/cm); i and o stand for inlet and outlet port of dialyzer; 1 and 2 stand for two different values of inlet dialysate conductivity. Determination of plasma water conductivity (Cpw) when D value is known: Cpw (mS/cm) = Cdi –

Qd ! (Cdi – Cdo) D

Locatelli/Manzoni/Di Filippo

(2)

References 1 Port FK, Johnson WJ, Klass DW: Prevention of dialysis disequilibrium syndrome by use of high sodium concentration in the dialysate. Kidney Int 1972;3:327–333. 2 Locatelli F, Pedrini L, Ponti R, Costanzo R, Di Filippo S, Marai P, Pozzi C, Bonacina GP: ‘Physiological’ and ‘pharmacological’ dialysate sodium concentrations. Int J Artif Organs 1982;5:17–24. 3 Gotch FA, Lam MA, Prowitt M, Keen M: Preliminary clinical results with sodium-volume modeling of hemodialysis therapy. Proc Clin Dial Transplant Forum 1980;10:12–17. 4 Di Filippo S, Corti M, Andrulli S, Manzoni C, Locatelli F: Determining the adequacy of sodium balance in hemodialysis using a kinetic model. Blood Purif 1996;14:431–436.

Modern Techniques and Electrolyte Balancing

5 Polashegg HD: Automatic non-invasive intradialytic clearance measurements. Int J Artif Organs 1993;16:185–191. 6 Petitclerc T, Hamani A, Jacobs C: Optimization of sodium balance during hemodialysis by routine implementation of kinetic modeling: Technical aspects and preliminary clinical study. Blood Purif 1992;10:308–316. 7 Locatelli F, Di Filippo S, Manzoni C, Corti M, Andrulli S, Pontoriero G: Monitoring sodium removal and delivered dialysis by conductivity. Int J Artif Organs 1995;11:716–721. 8 Ghezzi PM, Frigato G, Fantini GF, Dutto A, Mainero S, Cento G, Marazzi F, d’Andria V, Grivet V: Theoretical model and first clinical results of the paired filtration dialysis. Life Support System 1983;1(suppl):271–276.

9 Di Filippo S, Corti M, Andrulli S, Pontoriero G, Manzoni C, Locatelli F: Optimization of sodium removal in paired filtration dialysis by single pool sodium and conductivity kinetic models. Blood Purif 1997;15:34–44. 10 Locatelli F, Andrulli S, Di Filippo S, Redaelli B, Mangano S, Navino C, Ariano R, Tagliaferri M, Fidelio T, Corti M, Civardi S, Tetta C: Effect of on-line conductivity plasma ultrafiltrate kinetic modeling on cardiovascular stability of hemodialysis patients. Kidney Int 1998; 53:1052–1060. 11 Redaelli B, Locatelli F, Limido D, Andrulli S, Signorini MG, Sforzini S, Bonoldi L, Vincenti A, Cerutti S, Orlandini G: Effect of a new model of hemodialysis potassium removal on the control of ventricular arrhythmias. Kidney Int 1996;50:609–617.

Blood Purif 2001;19:195–199

199

Blood Purif 2001;19:200–205

Manifestations of Oxidant Stress in Uremia Jonathan Himmelfarb Ellen McMonagle Division of Nephrology, Maine Medical Center, Portland, Me., USA

Introduction

Oxidant injury is now thought to contribute to the pathogenesis of a wide array of disease states [1, 2]. These include the development of atherosclerosis, tissue injury related to ischemia-reperfusion injury, and cellular changes associated with both the aging process and carcinogenesis. In certain inflammatory disease states, such as the development of acute respiratory disease syndrome (ARDS), acute renal failure and inflammatory bowel disease, tissue injury can be mediated by oxidative reactions. Oxidative stress can result from the variety of biochemical pathways ranging from metal-catalyzed oxidative reactions to the release of reactive oxygen compounds such as superoxide anion, hydrogen peroxide and hypochlorous acid from activated phagocytic cells [3, 4]. Recently, the importance of myeloperoxidase-catalyzed oxidative reactions (with hypochlorous acid as the major oxidant in causing tissue injury by phagocytic cells) have been emphasized [5]. Thus in vivo detection of injury patterns induced by HOCl would be useful in characterizing inflammatory oxidative injury. In many human disease states, the understanding of how oxidative reactions contribute to the disease process is limited because of inaccessibility of important tissue samples during the disease process. For example, in ath-

ABC

© 2001 S. Karger AG, Basel 0253–5068/01/0192–0200$17.50/0

Fax + 41 61 306 12 34 E-Mail [email protected] www.karger.com

Accessible online at: www.karger.com/journals/bpu

erosclerosis, oxidative modification of lipoproteins likely takes place in the subendothelium which in humans is not routinely accessible for laboratory measurements [6]. In other disease processes that affect patients such as acute renal failure or ARDS, sampling of biological fluids and relevant tissue may not be attainable until well after the injury phase has occurred. Because of their ready access to repeated sampling, biomolecules that develop as a result of oxidative stress in the plasma are attractive in understanding how human disease may result from specific pathways of oxidative stress. Plasma proteins constitute an important target for analysis of oxidative modification, both because of their relative stability in plasma and because of the variety of oxidative reactions that can take place [7].

Biomarkers of Phagocyte-Derived Oxidant Stress

With increasing frequency, both neutrophils and monocytes/macrophages are being implicated as mediators of tissue injury in disease states [4]. Upon stimulation, phagocytic cells elaborate a variety of reactive oxygen species, including superoxide anion, hydrogen peroxide, hydroxyl radical and hypochlorous acid. The bulk of superoxide anion generated by activated phagocytic cells dismutates to hydro-

Jonathan Himmelfarb, MD Maine Medical Center, Division of Nephrology 22 Bramhall Street, Portland, ME 04102 (USA) Tel. +1 207 871 2417, Fax +1 207 871 6306 E-Mail [email protected]

gen peroxide, and hydrogen peroxide in turn can be rapidly metabolized by catalase. However, hydrogen peroxide, in combination with the enzyme myeloperoxidase, can oxidize halides to their corresponding hypohalous acids. Of note, myeloperoxidase is one of the most abundant proteins in phagocytes, constituting approximately 5% of neutrophil protein and 1% of monocyte protein, and is rapidly secreted upon stimulation. In the extracellular milieu, the combination of secreted hydrogen peroxide, myeloperoxidase and the abundant presence of chloride anions leads to the formation of hypochlorous acid, a highly reactive oxidant. Indeed, quantitative studies have revealed that from a stoichiometric perspective, the amount of HOCl generated by activated neutrophils accounts for almost all of the generated H2O2 [8]. Thus, HOCl is considered the most important oxidant produced by phagocytic cells in terms of risk of oxidant-induced tissue injury. Because hypochlorous acid is so highly reactive, it is consumed almost immediately in biological systems, thereby producing a variety of products depending on the reaction substrate [5]. In order to detect the in vivo effects of phagocyte-derived hypochlorous acid, it is important to have biomolecules that can serve as markers of hypochlorous acid oxidation. For systemic inflammatory disorders in which activated phagocytes are in the circulation, plasma proteins constitute an important target for oxidative modification. Oxidative protein modification can include fragmentation of the polypeptide chain, oxidation of amino acid side chains, as well as the generation of cross linkages between proteins [9–12]. The oxidation of amino acyl side chains are a particularly attractive biomarker of oxidative reactions because of easy detectability and high specificity of biochemical end products for specific oxidative pathways of injury. A powerful strategy for understanding the underlying in vivo mechanisms of oxidative injury is to identify stable end products of protein oxidation produced by different reaction pathways [5, 13]. For instance, HOCl generates 3-chlorotyrosine by reacting with tyrosine residues of proteins [14, 15]. In vitro studies indicate that 3-chlorotyrosine is a highly sensitive and specific marker for protein oxidation by HOCl [16]. Moreover, it is stable to acid hydrolysis, making it a potentially useful marker of protein oxidation by the myeloperoxidase pathway in vivo. Confirmation of the utility of 3-chlorotyrosine measurements in identifying pathways of oxidative injury has come from studies examining oxidized low density lipoproteins from the arterial wall of atherosclerotic aortic lesions [5].

Another important potential biomarker of phagocytederived oxidant stress are plasma protein-associated free thiol groups. In human plasma, the free thiol group found on albumin quantitatively constitutes the major source of free thiols in the plasma. Several investigators have demonstrated the importance of free thiol on albumin as a scavenger both of carbon centered free radicals and myeloperoxidase-generated oxidants. Measurements of plasma protein free thiol groups are relatively easy, inexpensive and sensitive, and can be important markers of oxidant stress, particularly with myeloperoxidasecatalyzed reactions. Furthermore, since the free thiol group in albumin is an important carrier of nitric oxide (in the form of S-nitrosoalbumin), oxidation of plasma protein thiol groups may have important biological consequences by decreasing nitric oxide availability to tissues. Another important product of hypochlorous acid-mediated plasma protein oxidation are aldehydes, which can be recognized chemically by the formation of carbonyl groups [5, 17]. Heinecke and co-workers [13, 18, 19], in a series of investigations, have demonstrated that myeloperoxidase-catalyzed oxidative reactions can convert virtually all of the common amino acids to reactive aldehydes. For example, the amino acid threonine can be converted to acrolein, known to be a potent cytotoxic compound [18]. Similarly, tyrosine, when exposed to myeloperoxidase, hydrogen peroxide and chloride anion, is converted to the highly reactive aldehyde p-hydroxyphenylacetaldehyde [19]. Because reactive aldehydes are thought to be of central importance in the genesis of atherosclerosis [5], detection of plasma protein carbonyl formation may be of particular importance in evaluating the pathogenesis of cardiovascular disease.

Oxidant Stress

Blood Purif 2001;19:200–205

Inflammation and Phagocyte Activation in Uremia

Many patients in chronic renal failure exist in a chronic inflammatory state. Stenvinkel et al. [20] have recently emphasized a close association between C-reactive protein as a marker of inflammation, corresponding malnutrition, and cardiovascular disease in patients with chronic renal failure. In addition to inflammation, chronic renal failure is also associated with excessive phagocytic cell activation. As an example, neutrophils from patients with chronic renal failure are primed for the respiratory burst when stimulated with FMLP [21].

201

Fig. 1. Thiol response to HOCl dose curve. * p ! 0.05 vs. no HOCl. n = 5. For analysis of variance, p = 3.8 ! 10 –29.

Fig. 3. Amine response to HOCl dose curve. n = 5. For analysis of

variance, p = 0.194.

important source of oxidative stress in patients with uremia and on hemodialysis given the potent reactive intermediates generated by activated neutrophils and monocytes.

Measurement of Plasma Protein Oxidation in Response to Phagocyte-Derived Oxidants in Hypochlorous Acid

Fig. 2. Carbonyl response to HOCldose curve. * p ! 0.05 vs. no

HOCl. n = 5. For analysis of variance, p = 2 ! 10 –9.

Hemodialysis also serves as a potent stimulator of circulating phagocytes. Products of neutrophil degranulation including both myeloperoxidase and elastase are elevated in the plasma of patients on chronic dialysis therapy compared to normal controls [22]. Furthermore, patients on chronic hemodialysis, particularly when undergoing dialysis using unmodified cellulosic membranes, develop activation of the alternative pathway of complement, increased neutrophil reactive oxygen species production, and neutrophil degranulation [23]. These cells could be an

202

Blood Purif 2001;19:200–205

In order to determine patterns of plasma protein oxidation as a response to phagocyte-derived oxidants, we characterized biochemical reactions that occur in human plasma as a result of exposure to both activated phagocytic cells and hypochlorous acid. For this set of experiments, whole blood was drawn from normal human volunteers, incubated with the following complement and neutrophil agonists: (1) PMA at 100 nM; (2) FMLP at 100 nM; (3) cobra venom factor at 10 units/ml. In additional experiments, 10 Ìl hypochlorous acid (HOCl) was added to 1 ml plasma from normal volunteers and incubated for 15 min at 37 ° C. Free sulfhydryl groups were then assayed according to the method of Ellman [24] as modified by Hu et al. [25]. To determine the concentration of oxidized amine groups, a fluorometric method described by Bohlan et al. [26] and adapted by Davies et al. [27] and Arnhold et al. [28] was utilized. Concentrations of carbonyl groups were determined according to the method of Levine et al. [29], as modified by Hu and Tappel [30] and Reznick et al. [31].

Himmelfarb/McMonagle

Fig. 4. Thiol response to agonists in whole blood. * p ! 0.05 vs. control. n = 7.

Fig. 5. Amine response to agonists in whole

Fig. 6. Carbonyl response to agonists in

blood. * p ! 0.05 vs. control. n = 7.

whole blood. n = 7.

Results

HOCl-Induced Plasma Protein Oxidation – Dose-Response Curve We measured the dose response to HOCl of plasma protein thiol oxidation, amino group oxidation and carbonyl formation. Plasma protein thiol levels showed a significant decrease starting at 0.1 mM HOCl and were decreased to 50% of control at a concentration of 0.6 mM incubation with HOCl (fig. 1). In comparison, an increase in plasma protein carbonyl formation was not detectable until incubation with 5 mM HOCl and did not become significant until incubation with 10 mM HOCl (fig. 2). From the dose-response curve, it can be estimated that a 50% increase in carbonyl formation would be obtained when 2.76 mM HOCl is added to plasma. Oxidation of plasma protein amine groups occurred at a higher concentration of hypochlorous acid than that required for thiol group oxidation or carbonyl group formation (fig. 3). There was a significant decrease in plasma protein free amine groups with incubation of 20 mM HOCl and the estimated HOCl concentration required to oxidize 50% of free amine groups was 38 mM HOCl (data not shown).

significantly oxidized with all three agonists. The thiol concentration in untreated samples was 374 B 49 ÌM compared to 267 B 33 ÌM after incubation with PMA (p = 0.01), 242 B 28 ÌM after incubation with FMLP (p = 0.03) and was 260 B 32 ÌM after incubation with CVF (p = 0.01). There were also significant changes in free amino groups after the addition of phagocytic cell activators to whole blood (fig. 5). The untreated control had free amino group concentrations of 1,982 B 70 FI, which decreased to 1,667 B 40 FI after incubation with PMA (p = 0.01), 1,699 B 29 FI after incubation with FMLP (p = 0.01), and 1,750 B 20 FI after incubation with CVF (p = 0.01). In contrast, there were no significant differences in plasma protein carbonyl concentrations between untreated samples and samples after incubation with PMA, FMLP or CVF (fig. 6). The untreated samples averaged 2.45 B 0.62 nmol carbonyl/mg protein versus 3.56 B 1.30 nmol carbonyl/mg protein after incubation with PMA (p = 0.40), 2.36 B 0.56 nmol carbonyl/mg protein after incubation with FMLP (p = 0.9), and 1.77 B 0.52 nmol carbonyl/mg protein after incubation with CVF (p = 0.26).

In vitro Plasma Protein Oxidation – Effects of Leukocyte Agonists To further examine the patterns of plasma protein oxidation associated with phagocytic cell-derived oxidants, we incubated whole blood with PMA, CVF and FMLP. Figure 4 demonstrates that plasma thiol groups became

We have examined the oxidation of amino acyl thiol groups, amino groups and carbonyl chemistry in response to hypochlorous acid and phagocyte activation in human plasma. The results of this study demonstrate that both

Oxidant Stress

Blood Purif 2001;19:200–205

Summary of in vitro Experiments

203

Fig. 7. Distribution of thiols in a hemodialysis population. Ranges are 10 ÌM thiol. n = 101.

plasma protein thiol oxidation and carbonyl formation can be used as biomarkers of hypochlorous acid-mediated oxidative reactions. In contrast, amino group oxidation is relatively insensitive to the effects of hypochlorous acid, but is somewhat more sensitive to phagocytic cell activation ex vivo.

plasma protein thiol groups to normal levels by having a minimal effect on plasma protein carbonyl expression. Figure 7 demonstrates the distribution of plasma protein free thiol groups in over 100 hemodialysis patients.

Summary and Conclusions Plasma Protein Thiol Oxidation and Carbonyl Formation in Patients with Chronic Renal Failure

After examining patterns of plasma protein oxidation in response to phagocyte-derived oxidants in vitro, we then examined patterns of plasma protein oxidation in vivo in patients with chronic renal failure and in patients on chronic maintenance hemodialysis therapy in comparison to normal volunteers [32]. The results of these experiments demonstrated: (1) There are significant differences in plasma free thiol groups between normal volunteers and both patients with chronic renal failure and chronic hemodialysis patients. (2) There are also significant differences in plasma protein carbonyl groups between normal volunteers, patients with chronic renal failure and patients on chronic hemodialysis therapy. (3) There were no significant differences in amine group oxidation between the three groups. (4) Hemodialysis with both biocompatible and bioincompatible membranes restored

204

Blood Purif 2001;19:200–205

Patients with chronic renal failure and patients on chronic hemodialysis therapy develop increased levels of plasma protein oxidation, likely as a consequence of oxidant stress. The oxidation of plasma protein-associated thiol groups, located predominantly on albumin, quantitatively is the major manifestation of protein oxidation in these patient populations. Hemodialysis improves ‘thiol stress’ in chronic hemodialysis patients, whether cellulosic or synthetic membranes are used. An understanding of the role of uremic oxidant stress, particularly in contributing to cardiovascular disease, may lead to new therapeutic modalities.

Himmelfarb/McMonagle

References 1 Halliwell B, Gutteridge JMC, Cross CE: Free radicals, antioxidants, and human disease: Where are we now? J Lab Clin Med 1992;119: 598–620. 2 Cross CE, Halliwell B, Borish ET, Pryor WA, Ames BN, Saul RL, McCord JM, Harman D: Oxygen radicals and human disease. Ann Intern Med 1987;107:526–545. 3 Weiss SJ: Tissue destruction by neutrophils. N Engl J Med 1989;320:365–376. 4 Miller RA, Britigan BE: The formation and biologic significance of phagocyte-derived oxidants. J Invest Med 1995;43:39–49. 5 Heinecke JW: Mechanisms of oxidative damage by myeloperoxidase in atherosclerosis and other inflammatory disorders. J Lab Clin Med 1999;133:321–325. 6 Steinberg D: Role of oxidized LDL and antioxidants in atherosclerosis. Adv Exp Med Biol 1995;369:39–48. 7 Davies MJ, Fu S, Wang H, Dean RT: Stable markers of oxidant damage to proteins and their application in the study of human disease. Free Radic Biol Med 1999;27:1151–1163. 8 Tetta C, Biasioli S, Schiavon R, Inguaggiato P, David S, Panichi V, Wratten ML: An overview of haemodialysis and oxidative stress. Blood Purif 1999;17:118–126. 9 Stadtman ER: Metal ion-catalyzed oxidation of proteins: Biochemical mechanism and biological consequences. Free Radic Biol Med 1990;9: 315–325. 10 Lee Y, Shacter E: Role of carbohydrates in oxidative modification of fibrinogen and other plasma proteins. Arch Biochem Biophys 1995; 321:175–181. 11 Stadtman ER: Role of oxidized amino acids in protein breakdown and stability. Methods Enzymol 1995;258:379–393. 12 Dean RT, Fu S, Stocker R, Davies MJ: Biochemistry and pathology of radical-mediated protein oxidation. Biochem J 1997;324:1–18.

Oxidant Stress

13 Heinecke JW: Mass spectrometric quantification of amino acid oxidation products in protein: Insights into pathways that promote LDL oxidation in the human artery wall. FASEB J 1999;13:1113–1120. 14 Domigan NM, Charlton TS, Duncan MW, Winterbourn CC, Kettle AJ: Chlorination of tyrosyl residues in peptides by myeloperoxidase and human neutrophils. J Biol Chem 1995;270:16542–16548. 15 Hazen SL, Hsu FF, Mueller DM, Crowley JR, Heinecke JW: Human neutrophils employ chlorine gas as an oxidant during phagocytosis. J Clin Invest 1996;98:1283–1289. 16 Hazen SL, Heinecke JW: 3-Chlorotyrosine, a specific marker of myeloperoxidase-catalyzed oxidation, is markedly elevated in low density lipoprotein isolated from human atherosclerotic intima. J Clin Invest 1997;99:2075–2081. 17 O’Neill CA, Halliwell B, van der Vliet A, Davis PA, Packer L, Tritschler H, Strohman WJ, Rieland T, Cross CE, Reznick AZ: Aldehydeinduced protein modifications in human plasma: Protection by glutathione and dihydrolipoic acid. J Lab Clin Med 1994;124:359–370. 18 Anderson MM, Hazen SL, Hsu FF, Heinecke JW: Human neutrophils employ the myeloperoxidase-hydrogen peroxide-chloride system to convert hydroxy-amino acids into glycolaldehyde, 2-hydroxypropanal and acrolein. J Clin Invest 1997;99:424–432. 19 Hazen SL, Gaut JP, Hsu FF, Crowley JR, d’Avignon A, Heinecke JW: p-Hydroxyphenylacetaldehyde, the major product of L-tyrosine oxidation by the myeloperoxidase-H2O2-chloride system of phagocytes, covalently modifies Â-amino groups of protein lysine residues. J Biol Chem 1997;272:16990–16996. 20 Stenvinkel P, Heimburger O, Paultre F, Diczfalusy U, Wang T, Berglund L, Jogestrand T: Strong association between malnutrition, inflammation, and atherosclerosis in chronic renal failure. Kidney Int 1999;51:1899–1911. 21 Ward RA, McLeish KR: Polymorphonuclear leukocyte oxidative burst is enhanced in patients with chronic renal insufficiency. J Am Soc Nephrol 1995;5:1697–1702.

22 Ward RA: Phagocytic cell function as an index of biocompatibility. Nephrol Dial Transplant 1994;9:46–56. 23 Cheung AK: Biocompatibility of dialysis membranes. J Am Soc Nephrol 1990;1:150–161. 24 Ellman GL: Tissue sulfhydryl groups. Arch Biochem Biophys 1959;82:70–77. 25 Hu ML, Louie S, Cross CE, Motchnik P, Halliwell B: Antioxidant protection against hypochlorous acid in human plasma. J Lab Clin Med 1993;121:257–262. 26 Bohlen P, Stein S, Dairman W, Udenfriend S: Fluorometric assay of proteins in the nanogram range. Arch Biochem Biophys 1973;155:213– 220. 27 Davies KJA, Delsignore ME, Lin SW: Protein damage and degradation by oxygen radicals. II. Modification of amino acids. J Biol Chem 1987;262:9902–9907. 28 Arnhold J, Hammerschmidt S, Wagner M, Meuller S, Arnold K, Grimm E: On the action of hypochlorite in human serum albumin. Biomed Biochim Acta 1990;49:991–997. 29 Levine RL, Garland D, Oliver CN, Amici A, Climent I, Anke GL, Bong-Whan A, Shaltiel S, Stadtman ER: Determination of carbonyl content in oxidatively modified proteins. Methods Enzymol 1990;189:464–478. 30 Hu ML, Tappel A: Potentiation of oxidative damage to proteins by ultraviolet-A and protection by antioxidants. Photochem Photobiol 1992;56:357–363. 31 Reznick AZ, Cross CE, Hu ML, Suzuki YJ, Khwarja S, Safadi A, Motchnik PA, Packer L, Halliwell B: Modification of plasma proteins by cigarette smoke as measured by protein carbonyl formation. Biochem J 1992;286:607– 611. 32 Himmelfarb J, McMonagle E, McMenamin E: Plasma protein thiol oxidation and carbonyl formation in chronic renal failure. Kidney Int (accepted).

Blood Purif 2001;19:200–205

205

Blood Purif 2001;19:206–210

Daily Hemodialysis: Is It a Complex Therapy with Unproven Benefits? Andreas Pierratos Humber River Regional Hospital, Toronto and University of Toronto, Ont., Canada

Introduction

Daily hemodialysis in the form of both short daily hemodialysis and nocturnal hemodialysis is used more frequently as an alternative to conventional hemodialysis and the interest in these modalities has intensified. The argument for their wider use will be defended.

Definitions and History

The typical prescription of daily short hemodialysis is 2 h daily, 6 days a week, using high blood and dialysate flows to achieve the highest clearance possible. It has been used either at the dialysis unit or at home. Nocturnal hemodialysis is performed nightly at home, for an average of 8 h 6–7 nights a week, at variable blood and dialysate flows. Peripheral accesses as well as central venous catheters have been used for both methods. The buttonhole cannulation technique [1] has been used by many centers for the cannulation of AV fistulas. The use of short daily hemodialysis was described first by DePalma et al. [2] in 1969, but the Italian experience of more than 18 years has generated the best known publications, mainly by Buoncristiani’s group in Perugia [3]. Several reports from The Netherlands [4], France [5], USA [6, 7] and elsewhere have been published and recently Woods et al. [8] published retrospective data from several centers. Nightly (nocturnal) hemodialysis was started by Uldall [9, 10] in Toronto in 1994. Several centers in North America [11,

ABC

© 2001 S. Karger AG, Basel 0253–5068/01/0192–0206$17.50/0

Fax + 41 61 306 12 34 E-Mail [email protected] www.karger.com

Accessible online at: www.karger.com/journals/bpu

12] and elsewhere [13] have experience with nocturnal hemodialysis with more than 100 patients currently on the treatment. The history of daily hemodialysis has been published by Kjellstrand and Ing [14].

Results

Short Daily Hemodialysis Quality of life (QOL): Most of the published studies are controlled, comparing the QOL of the same patients prior to and after the conversion to short daily hemodialysis. All studies were consistent in describing improvement of uremic as well as dialysis-related symptoms and remarkable hemodynamic stability. Although some of the older studies did not quantitate the improvement or used their own instruments for the quantitation [15], the more recent articles used previously validated instruments [4, 6]. The improvement in QOL and decrease in symptoms in unstable and symptomatic patients for whom the method was used as ‘salvage’, has been impressive. Blood pressure control: Improvement in blood pressure control has been uniform in all studies with an average decrease in antihypertensives by 50% [8]. Buoncristiani et al. [16] as well as Traeger et al. [17] reported decrease in left ventricular hypertrophy after conversion to short daily hemodialysis. Anemia control and EPO dose: A decrease in EPO dose has been described by most studies on average by about 50% [8, 18]. Kooistra et al. [4] did not detect significant

Andreas Pierratos, MD 112 Joicey Blvd Toronto, Ont M5M 2T6 (Canada) Tel./Fax +1 416 657 2669, E-Mail [email protected]

difference but in most studies the results are confounded by the lack of control for adequate iron administration. Nutrition: Improvement in nutritional parameters in the form of increase in serum albumin and weight gain has been reported by most of the studies [8]. Hospitalization rates: Decrease in hospitalization rates after conversion to short daily hemodialysis has been reported [7] but prospective studies have not yet been published. Vascular access: Despite the use of peripheral accesses daily, a decrease in the complication rate has been reported by Woods et al. [8] in their retrospective study. The complication rate decreased from 0.28 to 0.005/ patient/year after the conversion to short daily hemodialysis. Similar results were reported by Ting [19].

Anemia control improved with decrease in EPO dose by about 40%. The confounding effect of intravenous iron infusion has not been dissected out [24]. Nutritional studies done by our group showed increase in the serum amino acid levels [29] and increase in the total body nitrogen in 75% of the patients [30]. Weight gains have varied with some patients gaining more than 10 kg after the conversion to nocturnal hemodialysis.

Why Is Daily Hemodialysis Better than Conventional Hemodialysis?

Nocturnal Hemodialysis Small molecule clearance is higher than all the other hemodialysis regimens with a Kt/V of 1–2 nightly [10, 20]. Quality of life: significant improvement in QOL was documented by using several validated instruments including SF-36 [21, 22], Beck Depression Index and Sickness Impact Profile [21]. Personal patient testimonies have been very positive, especially by patients with previous co-morbid conditions. Hemodynamic stability of the patients is very high, as a result of the long and frequent dialysis as well as the recumbent position during the treatment. Therefore, dialysis partners have not been necessary. Blood pressure control: Improvement in blood pressure control has been impressive. In most patients all antihypertensives were discontinued [10]. Data on LVH are not yet available. Phosphate control has been excellent without the need for phosphate binders. Despite a high phosphate diet, addition of phosphate in the dialysate is required in 75% of the patients [23, 24]. The calcium phosphate product normalized in all patients. Middle molecule removal four times higher than conventional hemodialysis has been reported, exemplified by ß2-microglobulin [25]. Bone disease is easier to control with effective PTH suppression with phosphate control and the use of high dialysate calcium. Extraosseous tumorous calcifications have dissolved [26]. Sleep apnea is prevalent in ESRD and is an independent predictor of mortality [27]. Nocturnal hemodialysis has been found to correct sleep apnea in the affected individuals [28].

Kinetics The three times weekly hemodialysis regimen is characterized by wide fluctuations of biochemical parameters as well as intravascular fluid volume, described as ‘unphysiology’ by Kjellstrand et al. [31]. Daily dialysis is smoother, closer to the native kidney function. Since continuous ambulatory peritoneal dialysis (CAPD), the ultimate example of a physiological dialysis regimen, offers comparable patient survival with hemodialysis despite the lower weekly Kt/V, it was deduced that lower weekly Kt/V and therefore less total time on dialysis is adequate for daily hemodialysis. This is consistent with the good clinical results described by Buoncristiani [3] in the early 1980s in patients dialyzed with daily Kt/V of 0.23. Gotch [32] proposed the concept of the standard Kt/V (stdKt/V) to allow comparison of the different dialysis modalities. An stdKt/V of 2 per week (current level of adequacy) corresponds to single pool Kt/V (spKt/V) of 1.2 (or eKt/V of 1.05) delivered three times a week (3.5 h each) or a spKt/ V of 0.53–0.56 (or eKt/V of 0.38) delivered six times a week in the form of short dialysis. A review of this and related concepts was published by Gotch [32, 33]. This concept assumes that all dialysis regimens having the same predialysis BUN have the same outcomes. Casino and Lopez [34] described the concept of the equivalent renal urea clearance and used it for the comparison of the different modalities as well as native kidney function. The assumption was made that regimens with similar time average concentration (TAC) have similar outcomes. Following this concept, the amount of dialysis necessary for daily dialysis is higher than Gotch’s model predicts. Clark et al. [35] using computer simulation found that relative to a standard three times weekly hemodialysis regimen, a daily/short-time regimen results in a 3–6% increase in effective small and middle molecule removal. They also found that a daily low-flow/long-time regimen

Daily Hemodialysis

Blood Purif 2001;19:206–210

207

(low-flow nocturnal hemodialysis) substantially increases the effective removal of all solutes. The existence of the different approaches shows the difficulty when comparing the different regimens based on urea kinetics alone. Beyond Urea Kinetics The improvement in the condition of most of the patients after the conversion to daily hemodialysis is often impressive, more than one would predict from the increase in the dialysis dose, especially when it involves patients with small body size and already high Kt/V. Furthermore, there is doubt if outcomes improve with increasing Kt/V 11.4 [36, 37]. Another explanation for the improved results on daily hemodialysis is that the increased dialysis dose may improve the outcomes on daily hemodialysis more than on conventional hemodialysis or that the improvement is not related to the increase in the Kt/V but to other factors. Effect of Dialysis Time The importance of time on dialysis has been hotly debated. The difficulty lies in the fact that increased dialysis time increases the delivered Kt/V and therefore the effect of the two parameters is difficult to separate. The impressive results from the long dialysis practiced in Tassin [38] were ascribed to the long dialysis regimen. Interestingly, Kt/V did not correlate with patient survival in this group [39]. If time is the most important parameter, what can explain the improvement of patients on short daily hemodialysis? Early studies by Twardowski [40] suggested that frequency of dialysis is more important than time. Are then both time and frequency important? Do these methods have anything else in common?

Importance of ‘Unphysiology’: A Unifying Hypothesis?

The so-called ‘unphysiology’ of dialysis could be a significant cause of morbidity and often mortality. Frequently patients are rendered unconscious during hemodialysis and still these events are not taken into account in the evaluation of the dialysis modality. Despite some attempts to quantitate ‘unphysiology’, no correlation with outcomes exists. The best known parameter is the time average deviation (TAD) [41]. Unfortunately this value does not include time on dialysis and therefore the rate of change of the different parameters. The Tassin regimen, short daily hemodialysis and nocturnal hemodialysis, have as a common characteristic the decreased rate

208

Blood Purif 2001;19:206–210

of change in biochemical and hemodynamic parameters during dialysis. We hypothesize that the ‘unphysiology’ of dialysis could be as important as the dialysis dose, especially in the smaller patients with co-morbidities. Improvement of the outcomes could be achieved by increase in dialysis dose or decrease in ‘unphysiology’. In the case of CAPD, improvement of outcomes will be achieved by providing higher dose of dialysis. This could be possibly achieved by the continuous flow peritoneal dialysis [42– 44]. Hemodialysis outcomes can improve mainly by decreasing the ‘unphysiology’ of dialysis, while increasing the dialysis dose of the three times a week regimen may not be as effective. The larger patients are known to be better outcomes [45]. It is conceivable that the better outcome of this group is related to the lower level of ‘unphysiology’ relative to their size. Conversely, the poor outcomes of the small patients could be partially related to the higher level of ‘unphysiology’.

Is Daily Hemodialysis a Complex Therapy?

When it comes to the notion that daily hemodialysis is complex, one can ask the obvious question: In whose eyes? Despite the perceived complexity, none of the 50 patients that we have trained in nocturnal hemodialysis in our center requested to switch to conventional hemodialysis. Some of the patients insist on receiving dialysis 7 nights a week and a few requested that their transplantation be put on hold. Similar is our experience with 13 patients on incenter short daily hemodialysis over 2 years. Home daily/ nocturnal hemodialysis is not unduly complex since it is done happily by the patients at home, who visit the hospital only infrequently. In-center short daily hemodialysis is indeed more complex but the only result of the complexity is the increased expense. Is it then justified to fund daily hemodialysis with the current level of evidence? Finances: The direct cost of daily dialysis is obviously higher than that of conventional hemodialysis when done in the center, but in the case of home hemodialysis, labor cost is decreased. In a multicenter retrospective analysis, Mohr et al. [46, 47] calculated yearly savings of about USD 6,400 by the in-center short daily hemodialysis and about USD 9,500 for home short daily and nocturnal hemodialysis. The main reason for the financial benefits of daily hemodialysis was the decreased hospitalization rates [7] as well as the lower cost of medications, mainly EPO, antihypertensives and phosphate binders in the case of nocturnal hemodialysis. The data need to be documented in a prospective fashion.

Pierratos

Are the Benefits Unproven?

Although there is little doubt about the significant improvement in the QOL of the patients on daily/nocturnal hemodialysis shown by both data and patient preference, we do not have sufficient data on patient mortality to this point. Woods et al. [8] have reported a very high survival rate of 93% over 2 years on short daily hemodialysis. Similarly, Mastrangelo et al. [48] have described high survival of 55% over 10 years on a four times a week regimen. It is important that prospective controlled studies be done to document the benefits of daily hemodialysis in the areas of patient survival, hospitalization rates and cost. On the other hand, the improvement in QOL and the rest of the parameters outlined above, as well the

strong endorsement by the patients would dictate that daily hemodialysis should be used as an alternative dialysis regimen. In the case of in-center hemodialysis, short daily dialysis should be offered. Preference should be given to the smaller size patients as well as patients with dialysis-related or cardiovascular symptoms. At home, both short daily hemodialysis and nocturnal hemodialysis should be offered with preference to nocturnal hemodialysis in view of its several added advantages, including phosphate control, free diet and better hemodynamic profile. I will end on a philosophical note. Although not using critical scientific thinking in the practice of medicine is a mistake, ignoring the very obvious because of the lack of well-validated studies could be an even bigger mistake.

References 1 Twardowski Z, Kubara H: Different sites versus constant sites of needle insertion into arteriovenous fistulas for treatment by repeated dialysis. Dial Transplant 1979;89:78–80. 2 DePlama JR, Pecker EA, Maxwell MH: A new automatic coil dialyser system for ‘daily’ dialysis. Proc EDTA 1969;6:26–34. 3 Bouncristiani U: Fifteen years of clinical experience with daily haemodialysis. Nephrol Dial Transplant 1998;13(suppl 6):148–151. 4 Kostra MP, Vos J, Koomans HA, Vos PF: Daily home haemodialysis in The Netherlands: Effects on metabolic control, haemodynamics, and quality of life. Nephrol Dial Transplant 1998;13:2853–2860. 5 Traeger J, Sibai-Galland R, Delawari E, Arkouche W: Daily versus standard hemodialysis: One year experience. Artif Organs 1998;22: 558–563. 6 Ting G, Freitas T, Carrie B, Saum N, Kjellstrand CM, Zarghamee S: Short daily hemodialysis – Clinical outcomes and quality of life (abstract). J Am Soc Nephrol 1999;9:228A. 7 Ting G, Carrie B, Freitas T, Zarghamee S: Global ESRD costs associated with a short daily hemodialysis program in the United States. Home Hemodial Int 1999;34:1–4. 8 Woods JD, Port FK, Orzol S, Buoncristiani U, Young E, Wolfe RA, Held PJ: Clinical and biochemical correlates of starting ‘daily’ hemodialysis. Kidney Int 1999;55:2467–2476. 9 Uldall PR, Francoeur R, Ouwendyk M: Simplified nocturnal home hemodialysis. A new approach to renal replacement therapy (abstract). J Am Soc Nephrol 1994;5:428. 10 Pierratos A, Ouwendyk M, Francoeur R, Vas S, Raj DS, Ecclestone AM, Langos V, Uldall R: Nocturnal hemodialysis: Three-year experience. J Am Soc Nephrol 1998;9:859–868.

Daily Hemodialysis

11 Lockridge RSJ, Albert J, Andrerson H, Barger T, Coffey L, Craft V, Jennings FM, McPhatter LL, Spencer M, Swafford A: Nightly home hemodialysis: Fifteen months of experience in Lynchburg, Virginia. Home Hemodial Int 1999;3:23–28. 12 O’Sullivan DA, McCarthy JT, Kumar R, Williams AW: Improved biochemical variables, nutrient intake, and hormonal factors in slow nocturnal hemodialysis: A pilot study. Mayo Clin Proc 1998;73:1035–1045. 13 Simonsen O: Slow nocturnal dialysis as a rescue treatment for children and young patients with end-stage renal failure (abstract). J Am Soc Nephrol 2000;11:327A. 14 Kjellstrand CM, Ing T: Daily hemodialysis: History and revival of a superior dialysis method. ASAIO J 1998;44:117–122. 15 Buoncristiani U, Cairo G, Giombini L, Quintaliani G, Bonforte G: Dramatic improvement of clinical-metabolic parameters and quality of life with daily dialysis. Int J Artif Organs 1989; 121:33–36. 16 Buoncristiani U, Fagugli R, Ciao G, Ciucci A, Carobi C, Quintaliani G, Pasini P: Left ventricular hypertrophy in daily dialysis. Miner Electrolyte Metab 1999;25:90–94. 17 Galland R, Traeger J, Delawari E, et al: Control of hypertension and regression of left ventricular hypertrophy by daily hemodialysis (abstract). J Am Soc Nephrol 1999;10:279A. 18 Buoncristiani U, Fagugli R, Pinciaroli MR, et al: Control of anemia by daily hemodialysis (abstract). J Am Soc Nephrol 1997;8:216A. 19 Ting G: Blood access outcomes associated with short daily hemodialysis. Hemodial Int 2000; 44:2–4. 20 Williams AW, O’Sullivan DA, McCarthy JT: Slow nocturnal and short daily hemodialysis: A comparison. Semin Dial 1999;12:431–439.

21 Brissenden JE, Pierratos A, Ouwendyk M, et al: Improvements in quality of life with nocturnal hemodialysis (abstract). J Am Soc Nephrol 1998;9:168A. 22 McPhatter LL, Lockridge RSJ, Albert J, Anderson H, Craft V, Jennings FM, Spencer M, Swafford A, Barger T, Coffey L: Nightly home hemodialysis: Improvement in nutrition and quality of life. Adv Ren Replace Ther 1999;6: 358–365. 23 Mucsi I, Hercz G, Uldall R, Ouwendyk M, Francoeur R, Pierratos A: Control of serum phosphate without any phosphate binders in patients treated with nocturnal hemodialysis. Kidney Int 1998;53:1399–1404. 24 Pierratos A, Ouwendyk M: Nocturnal hemodialysis: Five years later. Semin Dial 1999;124: 19–23. 25 Raj DS, Ouwendyk M, Francoeur R, Pierratos A: Beta-2-microglobulin kinetics in nocturnal haemodialysis. Nephrol Dial Transplant 2000; 15:58–64. 26 Szabo T, Ouwendyk M, Pierratos A: Resolution of soft tissue calcification and improvement of bone density on nocturnal hemodialysis: A case report (abstract). Perit Dial Int 2000; 20(suppl 1):S109. 27 Benz RL, Pressman MR, Hovick ET, Peterson DD: Potential novel predictors of mortality in end-stage renal disease patients with sleep disorders. Am J Kidney Dis 2000;35:1052–1060. 28 Hanly P, Pierratos A: Nocturnal hemodialysis corrects sleep apnea in patients with chronic renal failure. N Engl J Med, in press. 29 Raj DS, Ouwendyk M, Francoeur R, Pierratos A: Plasma amino acid profile on nocturnal hemodialysis. Blood Purif 2000;18:97–102.

Blood Purif 2001;19:206–210

209

30 Pierratos A, Ouwendyk M, Rassi M: Total body nitrogen increases on nocturnal hemodialysis (abstract). J Am Soc Nephrol 1999;10: 299A. 31 Kjellstrand CM, Evans RL, Petersen RJ, Shideman JR, von Hartitzsch B, Buselmeier TJ: The ‘unphysiology’ of dialysis: A major cause of dialysis side effects? Kidney Int Suppl 1975; Jan:30–34. 32 Gotch FA: The current place of urea kinetic modelling with respect to different dialysis modalities. Nephrol Dial Transplant 1998;13 (suppl 6):10–14. 33 Gotch FA: Modeling the dose of home dialysis. Home Hemodial Int 1998;2:37–40. 34 Casino FG, Lopez T: The equivalent renal urea clearance: A new parameter to assess dialysis dose. Nephrol Dial Transplant 1996;11:1574– 1581. 35 Clark WR, Leypoldt JK, Henderson LW, Mueller BA, Scott MK, Vonesh EF: Quantifying the effect of changes in the hemodialysis prescription on effective solute removal with a mathematical model. J Am Soc Nephrol 1999; 10:601–609.

210

36 Collins AJ, Ma JZ, Umen A, Keshaviah P: Urea index and other predictors of hemodialysis patient survival [published erratum appears in Am J Kidney Dis 1994;24:157]. Am J Kidney Dis 1994;23:272–282. 37 Held PJ, Port FK, Wolfe RA, Stannard DC, Carroll CE, Daugirdas JT, Bloembergen WE, Greer JW, Hakim RM: The dose of hemodialysis and patient mortality. Kidney Int 1996;50: 550–556. 38 Charra B, Calemard E, Ruffet M, Chazot C, Terrat JC, Vanel T, Laurent G: Survival as an index of adequacy of dialysis. Kidney Int 1992; 41:1286–1291. 39 Laurent G, Charra B: The results of an 8-hour thrice weekly haemodialysis schedule. Nephrol Dial Transplant 1998;13(suppl 6):125–131. 40 Twardowski ZJ: Effect of long-term increase in the frequency and/or prolongation of dialysis duration on certain clinical manifestations and results of laboratory investigations in patients with chronic renal failure. Acta Med Pol 1975; 16:31–44. 41 Lopot F, Valek A: Quantification of dialysis unphysiology. Nephrol Dial Transplant 1998; 13(suppl 6):74–78.

Blood Purif 2001;19:206–210

42 Amerling R, Ronco C, Levin NW: Continuousflow peritoneal dialysis. Perit Dial Int 2000; 20(suppl):S172–S177. 43 Raj DS, Self M, Work J: Hybrid dialysis: Recirculation peritoneal dialysis revisited. Am J Kidney Dis 2000;36:58–67. 44 Diaz-Buxo JA, Cruz C, Gotch FA: Continuousflow peritoneal dialysis. Preliminary results. Blood Purif 2000;18:361–365. 45 Wolfe RA, Ashby VB, Daugirdas JT, Agodoa LY, Jones CA, Port FK: Body size, dose of hemodialysis, and mortality. Am J Kidney Dis 2000;35:80–88. 46 Mohr PE, Lockridge RS, Ting G: The Quality of Life and Economic Implications of Daily Dialysis. Bethesda, The Project HOPE Center for Health Affairs, 1999, p 8. 47 Mohr PE, Neuman PJ, Lockridge R, et al: The economic implications of daily hemodialysis (abstract). Perit Dial Int 2000;20(suppl 1): S105. 48 Mastrangelo F, Alfonso L, Napoli M, DeBlasi V, Russo F, Patruno: Dialysis with increased frequency of sessions (Lecce dialysis). Nephrol Dial Transplant 1998;13(suppl 6):139–147.

Pierratos

Blood Purif 2001;19:211–216

Daily Hemodialysis Is a Complex Therapy with Unproven Benefits Frank A. Gotch University of California, San Francisco, Calif., USA

Over the past 18 years, approximately 150 patients have been treated with ‘daily’ (5–6 days/week) hemodialysis worldwide [1]. The number of patients in any one center is very small, usually in the range of 7–10. The dose of dialysis is typically described as a weekly summed Kt/ V, usually estimated by approximation equations rather than formal urea kinetic modeling. The treatment times, blood and dialysate flows and dialyzers used have varied widely. The upper bound on therapy parameters is that reported by Pierratos et al. [2] with 6 dialyses per week, treatment time of about 7 h, Qb 200, Qd 300 with F40– F80 polysulfone dialyzers. The dose is not regularly monitored kinetically but can be estimated to be in the range of Kt/V about 1.2 each dialysis. The lower bound on treatment parameters is 6 dialyses per week with treatment time 2.0 h and Kt/V levels of 0.53 per treatment approximated from the Daugirdas equation [3]. Although there are no hard outcome data, i.e., mortality rate, these small observational studies often report better control of blood pressure and improved quality of life as the major benefits of more frequent dialysis. Improved appetite is variably reported but there are no reported systematic kinetic studies of protein catabolic rate to quantitatively assess change in protein intake. It will be argued here that the current clinical database does not permit valid assessment of the clinical benefits of ‘daily’ dialysis which will require prospective studies with well-defined and randomized dialysis doses.

ABC

© 2001 S. Karger AG, Basel 0253–5068/01/0192–0211$17.50/0

Fax + 41 61 306 12 34 E-Mail [email protected] www.karger.com

Accessible online at: www.karger.com/journals/bpu

An Overview of the Kinetics of Variable Frequency Dialysis Therapy: The Rationale for Increased Dialysis Frequency

The effectiveness of dialysis therapy is dependent on normalization of the concentrations of solutes which are normally excreted by the kidney and hence accumulate in body water with renal failure. The mechanism of solute removal rate (J, mass units per min) is first order, i.e., directly proportional to the solute concentration. Thus we can define the removal rate as the product of a clearance constant (K) and concentration C in accordance with J = KC

(1)

During an individual dialysis, C falls as the content of solute in body water falls and consequently the rate of solute removal J falls during dialysis. We can write a simplified, single pool, fixed distribution volume (V) rate equation without generation of solute to describe C as a function of treatment time (t), V(dC/dt) = –KC

(2)

Integration of equation 2 over an individual treatment results in Ct = Co[exp(–Kt/V)]

(3)

where Co and Ct represent beginning and end dialysis concentrations. Rearrangement of equation 3 gives Ct/Co = exp(–Kt/V)

Frank A. Gotch, MD 144 Belgrave Ave San Francisco, CA 94117 (USA) Tel. +1 415 661 6191, Fax +1 415 731 7876 E-Mail [email protected]

(4)

Fig. 1. A–C Illustration of doubling the total

weekly clearance by either doubling clearance (B) or doubling the number of dialyses per week from 3 to 6 (C). The weekly eKt/V doubles in both cases but does the dose of dialysis double in both cases?

It is apparent in simplified equation 4 that Kt/V defines the magnitude of drop in solute concentration over an individual dialysis. Consequently, the level of C and J both fall as Kt/V increases resulting in progressive reduction of the magnitude of solute removal or efficiency of dialysis as the dose or Kt/V increases for individual dialyses. It is for this reason that it is not rational to simply sum the individual Kt/Vs to compare variable frequency dialysis schedules. This can be illustrated as in figure 1 where the bottom panel A depicts a patient receiving an equilibrated eKt/V 1.1 for each of 3 dialyses per week and hence a weekly eKt/V 3.3. In panel B the clearance time product has been doubled with frequency held constant at 3 resulting in doubling of weekly eKt/V to 6.6. In contrast, in panel C the clearance time product is held constant but frequency doubled to 6 which also results in a weekly eKt/ V of 6.6. The first-order nature of solute removal mandates that the therapy option in panel B will result in greatly decreased efficiency of solute removal over the course of each treatment resulting from the high eKt/V and resultant low solute concentrations. In panel C, efficiency will be much better maintained during each dialysis and results in much greater increase in the effective dose of dialysis even though both options result in weekly summed eKt/V of 6.6. The difference in efficiency can also be expressed quantitatively as discussed below.

212

Blood Purif 2001;19:211–216

Kinetics of Intermittent and Continuous Dialysis: Definitions of Equivalency between Intermittent and Continuous Clearance

In the steady state of either continuous dialysis or renal clearance we can write equation 1 as Gss = KssCss

(5)

where Gss is solute generation rate which must equal the steady-state removal rate defined as the product of steadystate clearance and concentration. Solution of equation 5 for Kss is Kss = Gss/Css

(6)

The relationships in equation 6 are very useful to develop definitions of Kss values which are mathematically considered to be equivalent to the clearance provided by intermittent dialysis therapies of variable frequency and intensity. This approach to normalization of intermittent clearance requires empirical selection of some concentration value on the intermittent concentration profile (Css) to define Kss. Three empirical Css definitions have been reported as illustrated in figure 2. The weekly BUN profile shown in figure 2 was calculated for a patient with normalized protein catabolic rate (NPCR) 1.0 and a treatment schedule that delivered eKt/V 1.2 during each of 6 consecutive dialyses per week. The peak (pk) predialysis concentration after 2 days without dialysis (Cpk) was used

Gotch

Fig. 2. A BUN profile calculated for 6 hemodialyses per week, each delivering eKt/V 1.2 with NPCR 1.00. Three definitions of equivalent steady-state urea concentration are shown.

Fig. 3. Correlation of % survivalwith calculated stdKt/V for data reported by Ronco et al. [8]. Total patients 425; randomized into 3 levels of CRRT as plottted; Apache II scores 22, 23 and 22 in the three groups.

by Keshaviah et al. [4] to define Css in three times weekly dialysis for calculation of a Kss defined as Kpk in figure 2. The physical meaning of Kpk is that if this clearance were continuously delivered it would result in steady-state concentration equal to Cpk for the patient therapy defined by the weekly profile in figure 2. The mean predialysis concentration (Cm) was used for the definition of standard clearance (stdK) [5] as shown in figure 2. The third definition of Css reported [6] is the time average concentration (CTAC) which has been used to define the ‘equivalent continuous renal clearance’ or EKR.

ered to represent adequate therapy. Thus these data indicate stdKt/V provides a clinically valid mathematical definition of equivalent continuous and intermittent clearances in CAPD and three times weekly hemodialysis. The stdKt/V model has also been evaluated from analysis of recent data reported in ARF by Ronco et al. [8]. Ronco reported % survival as a function of three doses of hemofiltration expressed as D = 20, 35 and 45 ml/kg/h. These doses can be directly transformed into weekly stdKt/V from

Clinical Evaluation of Kpk, stdK and EKR

The three mathematical definitions of equivalent steady-state clearance are empirically derived as described above. They have been subjected to clinical evaluation by calculation of the total weekly steady-state clearances normalized to V (Kpk*t/V, stdKt*t/V and EKR*t/V) for three times weekly hemodialysis which provides an eKt/V 1.05 for each treatment [7]. The results showed Kpk*t/V = 1.75, stdK*t/V = 2.00 and EKR*t/V = 3.0. The total weekly clearance in CAPD (Kpt/V) widely considered to provide adequate dialysis therapy is 2.00 and correlates very well with the stdKt/V predicted from three times weekly hemodialysis with eKt/V 1.05, also consid-

Daily Hemodialysis Is Unproven

stdKt/V = [0.001(D)(24)(7)BWt]/[0.55BWt]

(7)

where BWt is body weight in kg, 0.55 is body water fraction and the constants 0.001(24)(7) convert ml/h to liters/ week. Note that BWt cancels in equation 7 which when simplified shows that stdKt/V values for D 20, 35 and 45 are 6.1, 10.7 and 13.7 respectively. The authors state that the average delivered dose was 87% of that prescribed so the three adjusted stdKt/V values become 5.3, 9.3 and 12.0 respectively, extremely high levels of dialysis compared to stdKt/V 2.0 in CAPD and three times weekly hemodialysis. The % survival values fitted as a logarithmic regression on stdKt/V are shown in figure 3 for the Ronco data and indicate a good relationship between survival and stdKt/V for this very high intensity therapy.

Blood Purif 2001;19:211–216

213

Fig. 4. Interrelationships amongst eKt/V delivered each dialysis,

number of days dialyzed per week and the level of stdKt/V achieved.

Generalized Solution of the stdKt/V Model

A generalized solution for stdKt/V as a function of eKt/V delivered each dialysis and the number of days of dialysis each week [5] is shown in figure 4. Note that when n is constant, the increase in stdKt/V becomes progressively shallower as eKt/V increases but that for each increase in n there is a relatively constant increase in stdKt/V at all levels of eKt/V reflecting the first-order nature of solute removal discussed above. The two reference points for stdKt/V 2.0 in CAPD and three times weekly hemodialysis are shown in figure 4. The model can next be used to quantify the differences in stdKt/V resulting with the therapy options depicted in figure 1 above. The weekly stdKt/V was computed as a function of increasing eKt/V with n constant at 3 dialyses per week (panel B option in figure 1) and as a function of increasing n with eKt/V per treatment constant (panel C option in figure 1). The results are shown in figure 5 where it can be seen that there is minimal increase in stdKt/V from 2 to 2.6 when eKt/V is doubled and n held constant. In contrast, when eKt/V per treatment is held constant and n doubled the stdKt/V increases from 2.0 to 4.2 and reflects the greatly increased efficiency of solute removal with increased frequency compared to increased intensity of each treatment.

214

Blood Purif 2001;19:211–216

Fig. 5. The stdKt/V values calculated for the two therapy options in figure 1. By doubling the eKt/V with n constant at 3, stdKt/V increases only from 2.0 to 2.6. Doubling the frequency with constant eKt/V per treatment caused stdKt/V to increase from 2 to 4.2, a 210% increase, greater than the increase in summed eKt/V.

Problems with Kinetic Evaluation of Short Dialysis and Low eKt/V

One approach to daily dialysis has consisted of maintaining constant clearance, i.e., constant blood and dialysate flow and dialyzer, but reducing the treatment time from 4 to 2 h and increasing frequency from 3 to 6 times weekly. In these instances the single pool spKt/V has been typically reduced from 1.2 to 0.6 and in this range of very low spKt/V there will be predictable underestimation of the calculated urea distribution volume calculated from single pool kinetics (Vsp) and greater rebound with decrease in eKt/V relative to spKt/V. The effect on Vsp compared to the true volume reflecting double pool kinetics is illustrated in figure 6 which shows that with spKt/V of 0.6, the Vsp is only 75% of the true Vdp while at spKt/ V 1.2 the Vsp is essentially equal to Vdp. Further, the ratios of eKt/V/spKt/V will be 0.87 and 0.77 respectively at 4- and 2-hour treatment times. These corrections can readily be made to clinical data calculated with the variable volume single pool model [9] using the Tattersall algorithms [10]. If these kinetic errors are not addressed with short treatment time and low eKt/V therapy, there will be substantial overestimation of the delivered dose of dialysis.

Gotch

Fig. 6. Comparison of urea distribution volumes calculated with

double pool (Vdp) and single pool (Vsp) kinetics. The relationship is a highly predictable function of spKt/V delivered and the Vsp error can be readily corrected in clinical data.

Quantitative Evaluation with the stdKt/V Model of Several Reported Studies of Daily Dialysis

Fig. 7. The loci of stdKt/V calculated for two widely quoted studies of daily dialysis with 5–6 dialyses per week and the calculated change in stdKt/V when eKt/V 1.05, n = 3 therapy is changed to eKt/V 0.525, n = 6 dialyses per week.

addition to improved quality of life and blood pressure control, it was striking in this data that phosphate binders were completely eliminated and in fact phosphate was added to dialysate to prevent depletion. A third data set is shown in figure 7 depicting the calculated change in stdKt/V which would result if three times weekly dialysis with eKt/V 1.05 is converted to six times weekly dialysis with eKt/V 0.525 which is similar to data reported by Kooistra et al. [3]. In this case a modest increase in stdKt/V from 2.0 to 2.4 (20%) is calculated while the weekly summed eKt/V remains constant at 3.3 and reflects again the improved efficiency with more frequent and shorter dialyses. Kooistra et al. also observed improved quality of life and better blood pressure control but no change in phosphate metabolism.

The first study of increasing frequency of dialysis was reported by Bonomini in 1972 and reprinted in 1998 [11]. Bonomini et al. compared outcome with 2, 3 and 5 dialyses per week and found markedly improved outcome when n was increased from 2 to 3 and minimal further improvement with n increased to 5 times per week. An analysis of this data has previously been reported [12] and the results are plotted in figure 7 where it can be seen that the stdKt/V with n = 2 was grossly inadequate (1.2) and increased to 1.8 with n = 3 and to 2.1 with n = 5. Thus in these studies 28 years ago the normalized dialysis dose was increased from a level now known to be grossly inadequate to a level with 5 dialyses per week which is now readily achieved with modern three times weekly hemodialysis in which eKt/V 1.05 is delivered each treatment. In contrast to the Bonomini data, the level of stdKt/V calculated from data reported by Pierratos et al. [2] for long overnight dialysis 6 times per week is greatly increased above that achieved with three times weekly dialysis. The Pierratos data are also depicted in figure 7 where the long overnight therapy is shown to result in stdKt/V 4.2 or more than 200% greater than typically achieved with adequate three times weekly dialysis and CAPD. In

Defining the doses of variable frequency and intensity dialysis therapies by simple summation of the total weekly Kt/V results in meaningless comparisons of therapy. There are strong kinetic reasons to predict that efficiency of dialysis therapy substantially increases as frequency of dialysis increases. These relationships can be quantified with the stdKt/V model which has been verified to have

Daily Hemodialysis Is Unproven

Blood Purif 2001;19:211–216

Conclusions and Recommendations

215

clinical validity in the domain of current CAPD and three times weekly HD therapy (stdKt/V 2.0). The model has also been shown to be predictive of clinical outcome at very high levels of therapy (stdKt/V 5–12) in acute renal failure patients. Analysis of daily dialysis reports with the stdKt/V model indicates that the dose of dialysis varies enormously and ranges from 20 to 210% higher than adequate three times weekly HD and CAPD. In the light of these observations it would seem particularly important to design a prospective study of six times weekly dialysis with short and long treatment times (2 vs. 6–8 h) and well-defined and controlled levels of stdKt/V in both treatment arms, probably in a stdKt/V range 3–4. The study must be adequately powered to assess both mortality and morbidity. The stdKt/V model should be expanded to include evaluation of larger solutes such as creatinine, phosphorus and ß2-microglobulin.

The most striking effect of daily dialysis reported to date appears to be improved quality of life and better blood pressure control. These are very likely due to better control of Na and water metabolism with less predialysis volume overload and less intradialytic hypotension and morbidity due to ultrafiltration. Efforts should be directed toward modeling these aspects of therapy with better definition of the effects of increasing frequency on these abnormalities. It is now known that small patients with smaller body water volume have increased mortality on dialysis therapy [12]. In part this may represent greater swings in hydration in small patients with higher fluid intakes relative to body size. Thus V must also be considered in design of daily dialysis studies and it may be that small patients will benefit more than large patients from increased frequency of dialysis.

References 1 Kjellstrand C: Daily hemodialysis is best: Why did we stop at three? Semin Dial 1999;12:403– 406. 2 Pierratos A, Ouwendyk M, Francoeur R, Vas S, Raj D, Ecclesstone A, Langois V, Uldall R: Nocturnal hemodialysis: Three-year experience. J Am Soc Nephrol 1998;9:859–868. 3 Kooistra M, Vos J, Koomans A, Vos P: Daily home hemodialysis in the Netherlands: Effects on metabolic control, haemodynamics, and quality of life. Nephrol Dial Transplant 1998; 13:148–152. 4 Keshaviah P, Nolph K, Van Stone J: The peak concentration hypothesis: A urea kinetic approach to comparing the adequacy of continuous ambulatory peritoneal dialysis and hemodialysis. Perit Dial Int 1989;9:257–260.

216

5 Gotch: The current place of urea kinetic modeling with respect to different dialysis schedules. Nephrol Dial Transplant 1998;13(suppl 6):10–14. 6 Casino F, Lopez F: The equivalent renal urea clearance: A new parameter to assess dialysis dose. Nephrol Dial Transplant 1996;11:1574– 1581. 7 Gotch F: Modeling the dose of home hemodialysis. Home Hemodial Int 1999;3:37–41. 8 Ronco C, Bellomo R, Homel P, Brendolan A, Dan M, Piccini P, La Greca G: Effects of different doses in continuous veno-venous haemofiltration on outcomes of acute renal failure: A prospective randomized trial. Lancet 2000; 356:26–30.

Blood Purif 2001;19:211–216

9 Gotch F: Kinetic modeling in hemodialysis; in Nissenson A, Fine R, Gentille D (eds): Clinical Dialysis, ed 3. Norwalk, Appleton & Lange, 1995, pp 156–189. 10 Tattersall J, Detakats D, Chamney P, Greenwood R, Farrington K: The post-hemodialysis rebound: Predicting its effect on Kt/V. Kidney Int 1996;50:2094–2102. 11 Bonomini JV, Mioli V, Albertazzi A, Scolari P: Daily-dialysis program: Indications and results. Nephrol Dial Transplant 1998;13:2774– 2778. 12 Wolfe R, Ashby V, Hulber-Shearon V, Agodoa L, Port F: Body size, dose of hemodialysis and mortality: Results from USRDS special studies. Am J Kidney Dis 2000;35:1–11.

Gotch

Blood Purif 2001;19:217–221

Interdialytic Weight Gain and Dry Weight Nathan W. Levin a Fansan Zhu a Marcia Keen b a Renal

Research Institute, New York, N.Y., and b Amgen, Inc., Thousand Oaks, Calif., USA

The concept of dry weight in dialysis patients is more illusion than reality since, in general, the attainment of this state may, for reasons to be discussed below, be fraught with patient discomfort and morbidity. However as a concept it represents an ideal state which is more or less what can be observed in healthy persons. To some, dry weight is defined as the weight below which symptoms or hypotension develop during dialysis, but this description is of limited applicability. To others the weight achieved at the end of dialysis which is even below this, is the true dry weight and is targeted at maintenance of normotension and in order to reduce problems from subsequent interdialytic weight gains. Despite the severity implied by these definitions it can be safely stated that the majority of dialysis patients are at the other end of the spectrum in a constant state of extracellular fluid overhydration only relieved intermittently by the dialytic procedure. Some direct evidence for this is derived from the phenomenon of diuresis with subsequent weight loss often seen in the first few days after renal transplantation. The objective of this presentation is to discuss issues affecting interdialytic weight change and methods used to determine and achieve dry weight. Jaeger and Mehta [1] wrote an important review of this topic up to February 1999 entitled ‘Assessment of Dry Weight in Hemodialysis: An Overview’.

ABC

© 2001 S. Karger AG, Basel 0253–5068/01/0192–0217$17.50/0

Fax + 41 61 306 12 34 E-Mail [email protected] www.karger.com

Accessible online at: www.karger.com/journals/bpu

Interdialytic Weight Gain

There is evidence that the relationship of thirst to plasma osmolality varies from subject to subject, but in a given person, it correlates closely with the magnitude of the increase in plasma osmolality [2] between individuals. Zerbe et al. [3] showed that the plasma osmolality value at which thirst was activated was reproducible within a given subject and in monozygotic twins the individual twin values were highly correlated. These would suggest that the osmolality set points are genetically determined. In patients on chronic hemodialysis, only perceived thirst is left as the major regulatory mechanism to return plasma osmolality to the preferred set point. Over the past 15–20 years, there has been an increase in the prescribed dialysate Na+ concentration from the lower end of the normal plasma Na+ concentration to the upper end of the normal range. This increase has been promoted as a mechanism to reduce intradialytic symptoms such as hypotension and muscle cramps [4]. There is some evidence that higher dialysate Na+ concentrations are associated with a larger interdialytic weight gain and perceived thirst [5]. It is postulated that patients with plasma Na+ values (plasma osmolality) in the lower range (!135 mEq/l) dialyzed with dialysate Na+ concentration in the upper range of normal could have an obligatory increase in plasma Na+ with each dialysis treatment with consequent thirst and fluid intake sufficient to dilute plasma Na+ down to the desired

Nathan W. Levin, MD Renal Research Institute 207 East 94th Street, Suite 303 New York, NY 10128 (USA) Tel. +1 212 360 4954, Fax +1 212 996 5905, E-Mail [email protected]

Fig. 1. The interdialytic weight gain is strongly correlated to the predialysis blood to dialysate sodium gradient (Cdi Na – CpO Na).

Fig. 2. The predialysis map is positively related to the initial dialysate to blood sodium gradient which in this patient population is sodium set point dependent.

Na+ set point. Thus a study was undertaken to determine the relationship between the predialysis mean plasma Na+ concentration (or patient Na+ ‘set point’) and the Na+ gradient (dialysate Na+ concentration, mEq/l – mean plasma Na+ concentration, mEq/l), and interdialytic weight gain and blood pressure. Fifity-eight hemodialysis patients had data abstracted covering 9–16 months. Patients with 1500 ml urine/day, and diabetic patients were excluded. The facility used a fixed dialysate Na+ concentration of 143 mEq/l. Approximately 98% of patients had average predialysis plasma Na+ concentrations below the dialysate Na+ concentration, and the Na+ gradient averaged 5.66 mEq/l or the osmotic quivalent of 9.2 mosm/l. The mean predialysis plasma Na+ concentration appeared to be normally distributed. As shown in figure 1, there was a statistically significant correlation between the Na+ gradient and the interdialytic weight gain (p ! 0.002), showing that higher Na+ gradients were associated with larger interdialytic weight gain. The analysis of the relationship between the Na+ gradient and predialysis mean arterial pressure (MAP) showed a statistically significant positive relationship (p ! 0.005), with a higher Na+ gradient being associated with a higher predialysis MAP (fig. 2). These data support the premise that a high dialysate Na+ gradient is related to higher interdialytic weight gains and predialysis MAP in hemodialysis patients. This finding suggests that some of the interdialytic weight gain in hemodialysis patients may be the result of increasing plasma osmolality above the preferred set point, leading to activation of thirst and fluid intake. Further prospective studies need to be done.

This procedure obviously requires knowledge of the patient’s plasma Na+ concentration, a measurement made usually only monthly in the USA in most dialysis units. Although this value could be used, reliance on one measurement may be inappropriate. A more satisfactory alternative is the use of the on line conductivity clearance device which provides an exact equivalent to the urea clearance. It utilizes conductivity differences between the inlet and outlet dialysate ports after increasing and then reducing the dialysate Na+ concentrations. A value for plasma Na+ is available about 15 min after treatment begins. This value, available with every treatment, could be used as the basis for altering dialysate sodium, and which can be implemented on most modern dialysis machines. The second determinant of weight gain is the total intake of food. Sherman et al. [6] showed that patients with 13 kg interdialytic weight gain compared to !2 kg had higher normalized protein catabolic rates (equivalent to protein intake in metabolically stable patients) and albumin levels without differences in Kt/V [6]. Since the study was cross sectional it is not certain as to whether this effect would be as prominent in patients who were in metabolic equilibrium. Relatively, smaller patients eat more and would have a larger intradialytic weight gain from this mechanism. Much attention in the dialysis literature is focused on the dose of dialysis, particularly on the fractional clearance of urea (Kt/V) but to a lesser extent on the removal of middle molecules (in a range today which includes molecular weights of 300–12,000 daltons). It had been difficult enough to obtain consensus on dialysis dose but the

218

Blood Purif 2001;19:217–221

Levin/Zhu/Keen

DOQI Guidelines [7] convincingly summarize the evidence for the urea clearance targets. This cannot be said yet to the same extent for large molecules although epidemiologic data support the positive value of high-flux dialysis, when controlled for Kt/V dose. Presumably the HEMO Study [8] which compares dose and flux effects will answer this question. When it comes to the next important element of the prescription, namely ultrafiltration volume, the basis for a decision is lacking. Charra et al. [9] have pointed out for years the importance for blood pressure control and improved survival of achieving dry weight. He and colleagues have carefully described procedures involving severe Na+ restriction and ultrafiltration with slow dialysis, to the point of producing symptoms. However, over a period of time these are less troublesome and subsequently weight increases due to better nutrition [9]. The Tassin group has held that the low morbidity and mortality experienced by their patients (notably more compliant than US patients), who are dialyzed for 8 h three times per week, are largely due to the success of the above procedure. While it is difficult to rank reasons, another explanation may be the low incidence of coronary artery disease in France and Japan, countries well known for low dialysis mortality [10]. High interdialytic weight gains have been associated with problems including increased mortality [11], left ventricular hypertrophy and dilatation [12], and cardiac failure [13]. Kimmel et al. [14] performed an observational, multicenter study of 283 urban patients to determine the relationship of interdialytic weight gain to several factors, including survival. Somewhat suprisingly, given the information in the above references, there was no increased risk of mortality with increased weight gain in the general nondiabetic dialysis population when age variation, severity of illness and serum albumin were controlled. Diabetics, but only those who had recently begun dialytic therapy, were however at greater risk. Previously, Koch et al. [15] were unable to demonstrate an association between weight gain and increased mortality in diabetic patients on dialysis. These two studies differed in the populations examined (91.9% African-American in the Kimmel study) and presumably the prevalence of hypertension, a key factor in the discussion. Nevertheless, it can be argued that at the extremes, high interdialytic weight gain as a percentage of dry weight can only be disadvantageous to patients. Given that dialysis times are shorter in the USA, for reasons of patient preference and economic pressure, a major consideration is how to prescribe the duration of a dialysis to reach two largely independent goals. The first is the Kt/V,

preferably in this context the equilibrated (eKt/V), since this will incorporate the variable effect of the rate of dialysis (K/V) which correlates with the postdialysis urea rebound. The second is the achievement of dry weight without symptoms such as hypotension and cramps. Clearly the dialysis duration has to be prolonged to the point that the interdialytic weight gain is removed without symptoms and that any additional dialysis dose is accepted. Knowledge of the appropriate dry weight would be an essential aid to the prescription. Since this is not yet available, nephrologists will often use a previous history of target weights and over a period of time possibly miss nutritional changes so that fluid accumulates. Fluid gain has to be excessive and obvious when the jugular venous pressure is raised, peripheral edema is present (in the absence of cardiac failure), and the blood pressure is rising. One approach to safe fluid removal is temperature control. Increase in core temperature is usual when dialysate is heated to 37 ° C. Van der Sande and Leunissen [16] have made a series of contributions on the effect of cooled dialysate and increasing vascular resistance and venous tone. By this approach it is clearly possible to increase or maintain a preferred rate of ultrafiltration without the feared possibility of hypotension[17]. It is simple to merely decrease dialysate temperature to 35.5 ° C as a routine prescription. However, this will only be effective in some patients, because of the wide range of predialysis temperatures, so that use of the blood temperature monitor (Fresenius Medical Care USA) is more effective since this device provides a continuous record and control of arterial and venous temperatures and of extracorporeal energy loss or gain. The update to the DOQI hemodialysis guidelines emphasizes the use of dialysate cooling or Midodrine® in maintaining blood pressure during ultrafiltration.

Interdialytic Weight Gain and Dry Weight

Blood Purif 2001;19:217–221

Blood Volume

During ultrafiltration blood volume almost always decreases, sometimes to the point of resulting in hypotension. However, significant decreases can occur even with small reductions in blood volume in patients with left ventricular diastolic dysfunction. In patients with stable hypertension, only small blood pressure changes may be observed even with substantial blood volume changes. The major factors involved in blood volume maintenance include the ultrafiltration rate, the patient’s fluid status, dialysate Na concentration, buffer substrate, venous compliance and core temperature. Clearly this measurement of blood volume may not provide information about

219

Fig. 3. Use of segmental bioimpedance measurement to indicate the patient’s dry weight. Measurements were made over 195 min hemodialysis (last 30 min UF = 0). From this curve, the change in volume in the leg tends to be constant over the last 40 min, which is presumably due to the patient having reached dry weight so that no more fluid could be removed.

Fig. 4. Comparison of the mean value of resistivity between HD patients and healthy subjects.

Table 1. Techniques for assessment of dry weight

Methods

Practical value

Problems

Reference

On-line blood volume

Simple and easy to use, highly accurate to measure relative change in BV

Rate of refilling is determined rather than interstitial fluid volume

4, 19–21

Inferior vena cava diameter

Not practicable

Needs to be measured well after dialysis ends, not sufficiently accurate

24

Doppler using sonography of femoral vein

Not adequately investigated

False positives

25

Whole body bioimpedance – using ratio of ECV to ICV

Easy to use and inexpensive device

Underestimates ECV in the trunk, incorrect estimated ICV

22, 23

Bioimpedance vector analysis

Indication of relative change in body hydration

Partly accurate in individual patients, too large standard deviation

26

Segmental bioimpedance

Quantitative analysis of body hydration

May require continuous measurement, and needs special device

27

Biochemical measures, e.g. atrial natriuretic peptide, cGMP

Chemical analysis

Influenced by cardiac function, not useful in individual patients or if underhydration present

28, 29

excess volume in the extracellular compartment. While blood volume measurement may be useful for preventing hypotension by manual (or in some countries automated) reduction in ultrafiltration rate day-to-day variation in autonomic response makes it difficult to select a particular volume for intervention.

220

Blood Purif 2001;19:217–221

Dry Weight

These problems with reliance on blood volume measurement for the prescription of total ultrafiltrate to be removed suggests a need for more direct measures of extracellular fluid volume. Table 1 shows some of the

Levin/Zhu/Keen

techniques that have been suggested for measurement of dry weight. A promising concept is the continuous measurement of interstitial water in patients. Zhu et al. [18] have devised a method examining extracellular fluid in arm, legs and trunk using segmented bioimpedance (fig. 3). This has been suggested also by Steuer et al. [19] using an optical method on the fingertip which shows a

flattening of the curve of the extracellular fluid as dry weight is reached. An alternative approach is to compare resistivity at the end of the dialysis treatment with healthy subjects who are presumably at dry weight (fig. 4). Further developments in this field are required to obtain the next stage of prescription for fluid management.

References 1 Jaeger JQ, Mehta RL: Assessment of dry weight in hemodialysis: An overview. J Am Soc Nephrol 1999;10:392–403. 2 Robertson GL: Disorders of thirst in man; in Ramsey DJ, Booth DA (eds): Thirst. London, Springer, 1991, pp 453–477. 3 Zerbe RL, Miller JZ, Robertson GL: The reproduciblility and heritability of individual difference in osmoregularity function in normal human subjects. J Lab Clin Med 1991;117:51– 59. 4 Leunissen KLM, Kooman JP, van Kujik W, van der Sande F, Luik AJ, van Hoof JP: Preventing haemodynamic instability in patients at risk for intradialytic hypotension. Nephrol Dial Transplant 1996;11(suppl 2):11–15. 5 Daugirdas JT, Al-Kudsi RR, Ing TS, Norusis MJ: A double-blind evaluation of sodium gradient hemodialysis. Am J Nephrol 1985;5:163– 168. 6 Sherman RA, Cody RP, Rogers ME, Solanchick JC: Interdialytic weight gain and nutritional parameters in chronic hemodialysis patients. Am J Kidney Dis 1995;25:579–583. 7 DOQI Guidelines: Am J Kidney Dis 1997; 30(suppl 2). 8 Eknoyan G, Levey AS, Beck GJ, Agodoa LY, Daugirdas JT, Kusek JW, Levin NW, Schulman G, for the HEMO Study Group: The Hemodialysis (HEMO) Study: Rationale for selection of interventions. Seminars in Dialysis 1996;9:24–33. 9 Charra B, Bergstrom J, Scribner BH: Blood pressure control in dialysis patients: Importance of the lag phenomenon. Am J Kidney Dis 1998;32:720–724. 10 Levey D, Kannel WB: Searching for answers in ethnic disparities in cardiovascular risks. Lancet 2000;356:266–267. 11 Leggat JE, Orzol SM, Hulbert-Shearon TE, Golper TA, Jones CA, Held PJ, Port FK: Noncompliance in hemodialysis: Predictors and survival analysis. Am J Kidney Dis 1998;32: 139–145.

Interdialytic Weight Gain and Dry Weight

12 Schmeider RE: Dietary intake and left ventricular hypertrophy. Nephrol Dial Transplant 1997;12:245–248. 13 Harnett JD, Parfrey PS, The Management of Congestive Heart Failure in Uremia Patients: Cardiac Dysfunction in Chronic Uremia. Dordrecht, Kluwer, 1997, pp 221–246. 14 Kimmel PL, Varela MP, Peterson RA, Weihs KL, Simmens SJ, Alleyne S, Amarashinge A, Mishkin GJ, Cruz I, Veis JH: Interdialytic weight gain and survival in hemodialysis patients: Effects of duration of ESRD and diabetes mellitus. Kidney Int 2000;57:1141– 1151. 15 Koch M, Thomas B, Tschope W, Ritz E, et al: Survival and predictions of death in dialyzed diabetic patients. Diabetologia 1993;36:1113– 1117. 16 Van der Sande FM, Gladziwa U, Kooman JP, Bocker G, Leunissen K: Energy transfer is the single most important factor for the difference in vascular response between isolated ultrafiltration in hemodialysis. J Am Soc Nephrol 2000;11:1512–1527. 17 Kaufman A, Morris A, Lavarias V, Wang Y, Leung J, Glabman M, Yusuf S, Levoci A, Polaschegg H, Levin NW: Effects of controlled blood cooling on hemodynamic stability and urea kinetics during high efficiency hemodialysis. J Am Soc Nephrol 1998;9:877–883. 18 Zhu F, Schneditz D, Wang E, Martin K, Morris AT, Levin NW: Validation of changes in extracellular volume measured during hemodialysis using a segmental bioimpedance technique. ASAIO J 1998;44:M541–M545. 19 Steuer RR, Germain MJ, Leypoldt JK, Cheung AK: Enhanced fluid removal guided by blood volume monitoring during chronic hemodialysis. Artif Organs 1998;22:627–632. 20 Lopot F, Kotyk P, Blaha J, Forejt J: Use of continuous blood volume monitoring to detect inadequately high dry weight. Int J Artif Organs 1996;19:411–414.

21 Bogaard HJ, de Vries JPPM, de Vries PMJM: Assessment of refill and hypovolaemia by continuous surveillance of blood volume and extracellular fluid volume. Nephrol Dial Transplant 1994;9:1283–1287. 22 Spiegel DM, Bashir K, Fisch B: Bioimpedance resistance ratios for the evaluation of dry weight in hemodialysis. Clin Nephrol 2000;53: 108–114. 23 Katzarski K, et al: Multifrequency bioimpedance in assessment of dry weight in hemodialysis. Nephrol Dial Transplant 1996;11(suppl 2):20–23. 24 Kouw PM, Kooman JP, Cheriex EC, Olthof CG, de Vries PM, Leunissen KM: Assessment of post dialysis dry weight: A comparison of techniques. JAm Soc Nephrol 1993;4:98–104. 25 Reibe A, Langer T, Rosch R, Osten B: Doppler sonography of the femoral vein – A new tool to dry weight determination in Hemodialysis. Nieren- und Hochdruckkrankheiten 1998;27: S368–S372. 26 Piccoli A: Identification of operational clues to dry weight prescription in hemodialysis using bioimpedance vector analysis. Kidney Int 1998;53:1036–1043. 27 Zhu F, Ronco C, Morris AT, Bashir A, Gotch F, Levin NW: Quantitative analysis of dry weight by segmental bioimpedance, J Am Soc Nephrol 2000;11:307A. 28 Leunissen KLM, Menheere PP, Cheriex EC, van den Berg BW, Noordzij TC, van Hooff JP: Plasma alpha-human atrial natriuretic peptide and volume status in chronic hemodialysis patients. Nephrol Dial Transplant 1989;4:382– 386. 29 Lauster F, Drummer C, Fulle HJ, Gerzer R, Schiffl H: Plasma cGMP level as a marker of the hydration state in renal replacement therapy. Kidney Int 1993(suppl 1):57–59.

Blood Purif 2001;19:217–221

221

Blood Purif 2001;19:222–226

Acute Dialysis Quality Initiative Claudio Ronco a John Kellum b Ravindra L. Mehta c a Department

of Nephrology, St. Bortolo Hospital, Vicenza, Italy; b Department of Intensive Care, University of Pittsburgh, Pa., USA, and c Department of Nephrology, USC at San Diego, Calif., USA

Introduction

Objectives of ADQI

Despite the fact that several guidelines and directions for appropriate medical management have been developed for end-stage renal disease, little has been done so far concerning acute renal failure and its treatment. For this reason we decided to undertake a process seeking consensus and evidence-based guidelines in the field of acute renal failure. The name of this process is Acute Dialysis Quality Initiative (ADQI).

Background

The intent of ADQI is to provide an objective, dispassionate distillation of the literature and description of the current state of practice of dialysis and related therapies. The purpose is to develop a consensus of opinion, with evidence where possible, on best practice and to articulate a research agenda to focus on important unanswered questions. This approach is a blend of ‘expert panel’ and ‘evidence appraisal’ and was chosen in order to achieve the best of both methods. We recognize that the expert panel in absence of the literature can lead to statements at odds with high-quality research and that evidence appraisal without expert input to question the framework can lead to erroneous interpretation [1]. Furthermore, this combined approach has led to important practice guidelines with wide acceptance and adoption into clinical practice [2, 3].

ABC

© 2001 S. Karger AG, Basel 0253–5068/01/0192–0222$17.50/0

Fax + 41 61 306 12 34 E-Mail [email protected] www.karger.com

Accessible online at: www.karger.com/journals/bpu

The final objectives for ADQI are the following: (1) standardize the process of dialysis for the critically ill patient; (2) develop consensus recommendations for best practice; (3) establish evidence-based guidelines where applicable, and (4) identify questions for future research and consider study design options. To obtain these results, a first step was required and the first ADQI conference was organized (New York on August 28–30, 2000) specifically focusing on continuous renal replacement therapy (CRRT). The objectives for this first phase were: (1) to describe the current clinical applications of CRRT (therapeutic options, current practice, and evidence); (2) to identify and prioritize issues within each CRRT topic requiring standardization and to define current state of consensus, and (3) to propose a strategy to address unresolved issues (consensus development, further research) (fig. 1).

Rationale for ADQI

An increasing number of patients develop acute renal failure each year, and mortality is still higher than 50% despite new treatment strategies. In recent years, there have been considerable advances in our understanding and technical capabilities, but consensus over the optimal way to deliver care does not exist. Consequently, we decided to initiate a process which will include a series of

C. Ronco, MD Department of Nephrology St. Bortolo Hospital I–36100 Vicenza (Italy) E-Mail [email protected]

Fig. 1. Strategy to address unresolved issues. 1When conflicting studies were found, a full evaluation and appraisal was recommended. Individual work groups did not attempt to synthesize and combine data from individual studies (meta-analysis) nor were attempts made to adjudicate between individual studies on the basis of quality.

Fig. 2. The ADQI founding group.

Since among the several controversial points concerning CRRT there is the question of who should be in charge of patient’s care and what should be the specific contribution of intensive care and renal physicians, the founding group of ADQI in New York was constituted by a balanced group of scientists of both branches.

conferences and interactions with a large number of reviewers and experts entitled ADQI. ADQI aims at establishing an evidence-based appraisal and set of consensus recommendations to standardize care and direct further research. CRRT is being used at ever increasing rates worldwide. Today, approximately one quarter of all patients with acute renal failure are treated with CRRT. Despite the increasing use, there are presently no published standards for the application of this therapy and practice patterns vary widely between individual centers. Results from recent clinical trials on selection of dialysis membranes, and dialysis dose provide strong, yet conflicting evidence to guide therapy. Other areas of uncertainty have not been sufficiently addressed by clinical studies and directives for future research are needed. Finally, the success of multicentered clinical trials in supportive care in the ICU (transfusion thresholds and ventilator management) have intensified and renewed interest in the study of supportive care methods as a major target for future research. These developments have set the stage for the first ADQI conference held in New York on August 28– 30, 2000. The conference focused on the application of CRRT in the critically ill patient with acute renal failure. While the primary aims of this conference were to establish the methodology for the consensus process, to describe current clinical practice and to identify important clinical and research questions, the final objectives of ADQI are the development of evidence-based practice guidelines and directions for future research.

Chaired by John Kellum, Claudio Ronco and Ravindra Mehta as directors of the ADQI conference, the group featured 7 intensivists (Drs Murray, Stewart, Corwin, Bellomo, Skippen, Schetz and Angus) and 7 nephrologists (Drs Paganini, Leblanc, Bunchman, Levin, Depner, Palevsky and Davenport) (fig. 2). In the group a few members from the industry (Drs Tetta, Lazarus and Clark) and 2 representatives from the American National Institutes of Health (Drs Star and Kimmel) were included. Since the meeting took place in the USA, the American Society of Nephrology and the Society of Critical Care Medicine endorsed the scientific event. Nevertheless, for the future we are very much looking forward to receiving further sponsorships and endorsement from other scientific societies in Europe and Asia and possibly to organize focused conferences of ADQI on different specific issues in various countries. Specific objectives for this conference were: (1) To establish a methodology for the development of a series of evidence-based recommendations on dialytic intervention in the ICU. This methodological

Acute Dialysis Quality Initiative

Blood Purif 2001;19:222–226

The First ADQI Conference

223

analysis included: (a) the systematic search for evidence with review and evaluation of the available literature; (b) the establishment of clinical and physiologic outcomes as well as measure parameters to be utilized for comparison of different treatments; (c) the description of the current practice and the rationale for the use of current techniques; (d) the analysis of areas in which evidence is lacking and future research is required to obtain new information. (2) To focus first on CRRT and define the topics that will require development of evidence-based guidelines. Summary statements were drafted by seven different working groups to describe the current state of the art for each topic. The topics were as follows: Definitions/ Nomenclature; Patient Selection; Solute Control (Treatment Dose); Membranes; Operational Characteristics (Convection/Diffusion); Fluid Management/Composition, and Anticoagulation/Access.

ADQI Conference Methodology

Each work group was assigned a particular topic area. Topics were selected on the basis of the following criteria: Prevalence of the clinical problem; Variation in clinical practice; Potential influence on outcome; Potential for development of EBM guidelines, and Availability of scientific evidence. Studies were identified via MEDLINE search, bibliographies of review articles and participants’ files. Searches were limited to English language articles. However, articles written in other languages were used when identified and presented by members of the group. Evidence was classified according to levels per EBM methodology. Qualitative commentary was provided when deemed necessary by the group. However, there was no critical appraisal of individual studies during this phase. Outcomes were grouped into the following major categories: physiologic (e.g. blood pressure, BUN, etc.), clinical (short-term morbidity/mortality, long-term morbidity/ mortality, renal recovery, functional class/quality of life) and economic. Different types of outcomes were considered separately for each area. Animal research was not considered evidence except that it contributed to commentary. Each work group was composed of three members, one who served as the group facilitator. Summary statements were developed through a series of breakout sessions where individual work group members were required to identify key issues for which guidelines are needed and to

224

Blood Purif 2001;19:222–226

classify current state of consensus and identify supporting evidence for each issue. Work group members were then required to present their findings to the entire group, revising each statement as needed until a final version was agreed upon. The responsibilities for presenting the findings of the work group to the rest of the participants was shared by each member on a rotating basis. Group facilitators revised work group findings as needed after each plenary session. Directives for future research were achieved by asking the participants to: (a) identify deficiencies in the literature, (b) determine if more evidence is necessary, and (c) if more evidence is necessary, articulate general research questions. When possible, pertinent study design issues were also considered. Conference activities were divided into three steps: pre-conference, conference and post-conference. In the pre-conference step, the methodology was developed, work groups were assembled and assigned to specific topics. Each group identified a list of key questions, conducted a systematic literature search and generated a bibliography of key studies. During the next step, the conference itself, the methodology was approved by the group and the conference was divided into breakout sessions where work groups addressed the issues in their assigned topic area, and plenary sessions where their findings were presented, debated and refined. During the first plenary session, the key questions were discussed and debated. Revised versions (some added, some deleted and others re-written) of each question were then presented at the second plenary. At this point evidence was assembled for each question and summary statements were drafted. These statements were further refined in subsequent plenary sessions until final versions were agreed upon. A writing committee assembled the individual reports from the work groups. Each report was edited to conform to a uniform style and for length. The final reports were posted on the internet [www.ADQI.net] and mailed to each participant for comment and revision. Finally, international consultants were identified and reports were sent to them for comment. Once final reports were completed, the writing committee summarized the individual reports into a final conference document (fig. 3). In detail, the following steps can be summarized to identify the first ADQI conference:

Pre-Conference Activity Each participant was part of a working group to cover a single topic. Each member was required to perform the tasks listed below. Participants were encouraged to communicate with other group

Ronco/Kellum/Mehta

Fig. 3. Flow-sheet of the conference.

Table 1. Topics covered and not covered in the conference

Topics covered in the conference Definitions/nomenclature Patient selection for CRRT Solute control in CRRT Membranes Operational characteristics Access and anticoagulation Fluid composition and management Topics not covered in the conference Indications for renal replacement therapy Costs Drug dosing Blood purification in non-renal failure conditions Withholding and withdrawing dialysis

3. Compile a bibliography One complete set of references was brought to the meeting by the group members. A bibliography was compiled prior to the meeting, and this was organized in a single format. 4. Assess the current status of consensus It was determined for local institutions and regions what questions were already fairly settled vs. ones that were not. It was also determined what questions will be likely to be answerable with current literature vs. ones that have insufficient evidence. Each question was rated as either: (a) consensus already exists; (b) data exists but controversy and variability of practice is still present; (c) insufficient evidence is available. Note: we did not try to judge the quality of the evidence at this stage.

Conference Activities Part 1

2. Perform a systematic literature search Each participant was provided with a list of references from the directors. Participants were required to perform their own literature search to find any additional articles. Search strategy and terms had to be specified and participants had to be prepared to defend any exclusion criteria. In general, the trend was to be as broad and inclusive as possible.

The entire group was asked to consider methodology for the ADQI process. Specific tasks included the following: E The incorporation of evidence-based medicine principles into the literature review process. Definition of levels of evidence and terms to be used. Definition of what the literature sources should be and how far back the literature should be reviewed. E Determination of what clinical, physiologic, and health economic outcomes should be considered evidence of effectiveness in clinical trials in RRT. E Definition of how, physiologic outcomes (e.g. arterial blood pressure) should be rated in relation to clinical outcomes (e.g. survival, need for long-term dialysis) and health economic outcomes (e.g. total costs, length of hospitalization) in evidentiary tables used for further ADQI consensus statements. E Debate on the role of evidence from animal research in RRT. E Definition of how peer review should be done and who will be the peer group. E Discussion on how ‘best clinical practice’ will be established in the absence of evidence.

Acute Dialysis Quality Initiative

Blood Purif 2001;19:222–226

members via E-Mail or other means in order to streamline their efforts and work collectively. Participants were also encouraged to communicate with other experts both locally and internationally. 1. Define a list of questions within the topic For each topic, a list of questions was generated. For example under membranes: ‘Which membranes should be used for CRRT?’; ‘Should bio-compatible membranes be used for CRRT’; ‘How often should filters be changed during CRRT?’

225

E Determination of how consensus will be achieved on directions for future research. E Discuss who will be sponsoring bodies for overall effort and how we will approach this issue.

Conference Activities Part 2 Breakout sessions were used for each group to catalog and review the literature in each area and define areas of established consensus as well as areas where consensus is lacking. Each group reviewed preconference work and present a draft set of statements that summarize the questions for their topic and the state of the current literature. The specific tasks for each group were: E Create a list of individuals to serve as consultants (for each topic) for the final consensus. E Develop the final list of questions and identify key evidence that should be reviewed. (Each topic required the key references associated with it.) E A summary statement listing what is needed to proceed further for each question and the current state of the literature was drafted. The spokesperson presented the draft summary statements and presented findings from any key studies. Note, at this stage, we did not attempt to develop NEW consensus; that will be covered in the second stage of the ADQI process. Instead, summary statements listed questions, described current practices and noted the presence or absence consensus already existing. E The entire group evaluated the statements and suggested revisions. E Final statements were drafted ‘on line’ with all members present. A 2/3 majority vote was required to approve all statements. Indeed, unanimous agreement existed for most.

Post-Conference Activity and Future Plans A writing committee will include the conference directors and 2 other members nominated by the group to compile the findings of the conference. This document will be completed as soon as all the necessary revisions will be made from the original drafts and will be posted on the Internet [www.ADQI.net] for comment by the remainder of the participants. The period for comment will be limited in time and revisions will be made accordingly. The final product will be submitted as a manuscript for publication immediately following this process.

226

Blood Purif 2001;19:222–226

Conclusions

In conclusion, ADQI is a moving process that will produce evidence-based statements on different issues concerning acute dialysis. The first step was to try to reach consensus on CRRT, an area where major controversies are still present. The next step will be the development of consensus statements that should provide the basis for recommendations to be used in clinical practice. Our effort aims at obtaining a common ground where acute dialysis should be discussed and optimized. At the present time there is very little agreement on how much, when and how dialysis should be provided. We hope to move much further with the cooperation of all who may be interested in helping and becoming temporary or permanent members of the commission for the development of the ADQI tasks.

Acknowledgements The ADQI directors would like to express their gratitude to Renal Research Institute for the organizational and scientific support to the first ADQI conference. The generous support of the sponsors should also be acknowledged. In particular we would like to thank: Baxter, Bellco, B. Braun, Fresenius Medical Care, Gambro Renal Care, Kimal, Medica, Nextrom Med Tech and Renaltech.

References

1 Kellum JA, Ramakrishnan N, Angus D: Appraising and using evidence in critical care; in Grenvik A, Shoemaker PK, Ayers S, Holbrook PR (eds): Textbook of Critical Care. Philadelphia, Saunders, 1999, chapt 193, pp 2059– 2069. 2 Chestnut RM: Implications of the guidelines for the management of severe head injury for the practicing neurosurgeon. Surg Neurol 1998; 50:187–193. 3 National Kidney Foundation: KDOQI: www. kdoqi.org.

Ronco/Kellum/Mehta

Blood Purif 2001;19:227–232

Indications for Dialysis in the ICU: Renal Replacement vs. Renal Support Ravindra L. Mehta Department of Medicine, Division of Nephrology, University of California, San Diego, Calif., USA

Introduction

Whether or not to provide dialytic support, and if so, when, are two of the most fundamental questions facing nephrologists and intensivists in most cases of acute renal failure (ARF) in the ICU. Although these decisions are integral to the management of any critically ill patient with renal failure in the ICU, there is limited information on what should determine the decision to dialyze. Over the last decade significant advances have been made in the availability of different dialysis methods for replacement of renal function. The advances have ranged from modifications in intermittent dialysis, e.g. biocompatible membranes, bicarbonate dialysate and smarter dialysis machines with volumetric ultrafiltration controls, to the development of several modalities for continuous renal replacement therapy (CRRT) [1–6]. Several of these techniques may be used to treat ARF in the ICU, but there is little information on when dialysis should be offered and which therapies are most appropriate in a given circumstance. This article outlines the current concepts in the use of dialysis techniques for ARF in the ICU and suggests an approach for providing dialysis support for the critically ill patient.

Dialysis Decisions in the ICU: Current Practice

In general, most nephrologists will have no hesitation in offering dialysis in the presence of life-threatening situ-

ABC

© 2001 S. Karger AG, Basel 0253–5068/01/0192–0227$17.50/0

Fax + 41 61 306 12 34 E-Mail [email protected] www.karger.com

Accessible online at: www.karger.com/journals/bpu

ations, e.g. severe hyperkalemia, marked acid-base disturbances or diuretic-resistant pulmonary edema. However, in the absence of these factors there is generally a tendency to avoid dialysis as long as possible, a thought process that reflects the decisions made for patients with endstage renal disease, in whom the initiation of dialysis signals the start of dialysis dependency. Two factors tend to dissuade nephrologists from initiating dialysis in the ICU. First, there are well-known risks of the dialysis procedure, including hypotension, arrhythmia, and complications of vascular access placement. Second, there is a strong concern that some element of the dialysis procedure may slow the recovery of renal function, and increase the risk of developing end-stage renal failure [7–9]. This contention is supported by experimental data showing renal lesions consistent with fresh ischemia in animals dialyzed without systemic hypotension, long after their initial renal injury. Thus, in current practice, the decision to dialyze is based most often on clinical features of volume overload and biochemical features of solute imbalance (azotemia, hyperkalemia) [10–12]. As shown in figure 1, the decision to dialyze is often based on an estimation of the likelihood for and timing of renal functional recovery. Factors that influence the likelihood of renal functional recovery include a knowledge of the nature and timing of renal insult, the severity of the underlying illness and associated co-morbidities and the presence of other factors known to adversely influence renal function, e.g. prolonged hypotension (table 1). When the nature and timing of the renal insult is known,

Ravindra L. Mehta, MD, FACP UCSD Medical Center, 8342 200 W Arbor Drive San Diego, CA 92103 (USA) Tel. +1 619 294 6083, Fax +1 619 291 3353, E-Mail [email protected]

Fig. 1. Decision process for initiation of dialysis in the ICU.

Table 1. Factors to assess likelihood of spontaneous reversal of renal dysfunction in ARF

Factor

Influence

Nature and timing of renal insult

Both nature and timing well defined, e.g. antibiotic nephrotoxicity (20%) Possible knowledge of insult and timing, e.g. postoperative ARF (30%) Nature and timing unknown, e.g. multiorgan failure (50%)

Presence of oliguria

Affected by diuretic use Inaccurate marker for estimating level of renal function Unreliable as an indicator for recovery

Change in BUN and creatinine

Affected by multiple factors Imprecise in detecting impending recovery May lag behind recovery

Underlying disease

Is ARF an epiphenomenon? Does ARF contribute to outcome?

Other factors

Demand exceeds renal excretory capacity, e.g. volume resuscitation Intensivist demand Logistics

e.g. a contrast load during an angiographic procedure, it may be possible to estimate the duration of renal dysfunction and withhold dialysis until evidence of worsening or improvement is obtained. In most circumstances, however, the nature and timing of renal injury are not precisely determined and it is thus difficult to predict the course.

228

Blood Purif 2001;19:227–232

Fig. 2. Factors affecting initial timing and frequency of dialysis. Results from a survey of US nephrologists.

Most clinicians base their estimate of the severity of renal dysfunction on the presence or absence of oliguria, the rate of change in blood urea nitrogen (BUN) and knowledge of the underlying disease. A recent survey of practicing nephrologists showed that the main determinants for the timing of intervention were an estimate of catabolic state and a need to treat specific levels of BUN (level 1100 mg/dl) (fig. 2) [13]. Unfortunately, these parameters are subject to wide variation and do not always reflect the underlying state. While oliguric renal failure has been associated with a poor outcome in most studies and dialyzed patients are more likely to be oliguric, urine volume is influenced by diuretic use and is an inaccurate marker of renal function and is unreliable as an indicator for impending renal recovery. Similarly the change in BUN and creatinine is influenced not only by the change in renal function but also by the underlying volume status, catabolic state and nutritional supplementation. Thus these markers become inaccurate to estimate changes in renal function and may lag behind recovery. Some other factors that influence the decision to dialyze include volume requirement in excess of urinary output, demand from the consulting physician to provide dialysis and the availability of dialysis based on logistical issues. It is thus evident that at the current time there is no standardization of the indications or timing of dialysis in the ICU and in most instances the decision to offer dialysis uses a thought process similar to that used for ESRD. I believe that this approach is problematic and fails to recognize the varying needs of the critically ill patient and thereby limits the application of dialysis techniques.

Mehta

Dialysis Decisions in the ICU: Need for Change?

It is quite evident that despite the availability of several dialysis techniques, including CRRT, we have been unable to demonstrate an improvement in survival in critically ill patients who develop ARF in the ICU setting. While several factors influence survival from ARF, there is emerging evidence that the development of ARF itself imparts a significant risk for an adverse outcome. Levy et al. [14] found a fivefold increased risk of mortality in patients who developed contrast-induced ARF in comparison to those who did not. The increased risk could be attributed to worsening of pre-existing and development of new onset sepsis, respiratory failure, mental status changes and bleeding. Similarly, Chertow et al. [15] found 30-day mortality in patients with ARF post-cardiac surgery was 63.7% compared with 4.3% in patients without ARF. Most recently, Vincent et al. [16] have shown that ICU mortality was three times higher in patients with ARF and that these patients had a greater incidence of infections and multiorgan failure. We have previously shown that the time to consultation for ARF and institution of therapy has a significant effect on mortality, renal functional recovery and length of stay in the ICU [17]. Thus these studies all support the notion that if ARF is unrecognized or untreated it contributes to an adverse outcome. If ARF is deleterious, it is tempting to postulate that earlier intervention with dialysis may be of benefit; however relatively few studies have carefully examined this question [18–22] (none in the modern dialysis era), and most studies on timing are confounded by differences in intensity as a result of a chosen therapeutic strategy. Moreover, changes in the severity of illness, especially in later years, make comparisons of studies extremely difficult. Conger [21] and Gillum et al. [22] performed prospective studies of the intensity of dialysis, the former showing beneficial trends in improved survival, although neither were well controlled for timing of intervention. Case mix clearly differs today compared with these studies dating back several decades. A much smaller proportion of patients require dialysis because of obstetric complications, and many more are elderly and suffer from multiple organ dysfunction accompanying ARF. Furthermore, what was considered early in the previous studies would not be considered early by most criteria today. Many modern nephrologists intervene when the serum urea nitrogen concentration is in excess of 100 mg/dl, for fear of impending uremic complications. Our data from a randomized controlled trial comparing intermittent hemodialysis (IHD) to CRRT [23] sug-

Indications for Dialysis in the ICU

gest that the indication for dialysis is an important determinant of outcome. In our trial, patients who were dialyzed predominantly for solute control had a better outcome than those who were dialyzed predominantly for volume overload. Patients dialyzed for control of both azotemia and volume overload experienced the worst outcome. Mukau et al. [24] found that 95% of their patients with postoperative ARF had fluid excess of more than 10 liters at the time of dialysis. The amount of fluid overload was a strong determinant of outcome independent of other factors. Volume resuscitation is a common strategy used in the treatment of multiple organ dysfunction, particularly when accompanied by sepsis syndrome and hypotension. It is often applied indiscriminately in the setting of oliguric ARF, where it is assumed that providing additional volume will lead to increased renal perfusion, and prompt correction of renal dysfunction. While this may be of great benefit in patients with prerenal azotemia, excessive fluid administration can lead to pulmonary edema, compromising oxygenation and ventilation, and may hasten the need for dialysis. These factors collectively suggest that we need to develop evidence-based, patient-specific, nonbiased indications for the initiation of dialysis in ARF.

Dialysis in the ICU: Suggested Approach

We favor utilizing an approach that recognizes that critically ill patients have different needs than stable patients with ESRD and that dialysis in the ICU can serve several purposes. The following factors should be considered in defining the indications and timing of dialysis. Characteristics of the Critically Ill Patient When ARF complicates the course of a critically ill patient in the ICU it is usually associated with multiple organ failure (MOF), which can influence the course of the patient in two ways. There may be a rapid decline of renal function that does not permit much of an adaptive response, which characterizes the course of the patient with ESRD. Second, therapeutic interventions designed to support other organ function, e.g. volume resuscitation, may exceed the renal excretory capacity, contributing to a worsening of the underlying state. In these circumstances it is apparent that the goal for any therapeutic intervention is to provide support for various organs and compensate for the adverse effects of other therapeutic interventions to provide an opportunity for the patient to recover from the underlying illness. Renal functional recovery is thus largely influenced by recovery of other organ func-

Blood Purif 2001;19:227–232

229

Table 2. Strategies for management of renal failure

Goals of therapy Desired outcome Determining factor Indication for dialysis

ARF

ESRD

Improve organ system failure Survival, recovery of renal function Other organ function Renal support

Ameliorate uremia Long-term survival, quality of life Renal process Renal replacement

Table 3. The changing paradigm of dialysis in the ICU

Purpose Timing of intervention Indications for dialysis Dialysis dose Application

Renal replacement

Renal support

Replace renal function Based on level of biochemical markers Narrow Extrapolated from ESRD Renal failure

Support other organs Based on individualized need Broad Targeted for overall support Renal and nonrenal indications

tion. Dialysis in this setting thus has the primary goal to provide adequate renal support for other organ function. This is in contrast to the patient with ESRD in whom the goal for dialysis is to ameliorate the effects of uremia and the determinant for long-term outcome is primarily the delivery of dialysis (table 2). Goals of Therapy and Therapeutic Potential of Dialysis The broad goals for treating ARF with dialysis are to (1) maintain fluid and electrolyte, acid-base and solute homeostasis, (2) prevent further insults to the kidney, (3) promote healing and renal recovery and (4) permit other support measures (e.g. nutrition) to proceed without limitation. Dialysis techniques differ in their operational characteristics and their ability to provide renal support and this should be considered in the dialysis decision. For instance, dialysis is largely viewed as a method to remove solute and fluids, as these are the main options utilized for ESRD patients. However, in the ICU setting dialysis techniques could be used not only to remove substances but also to add them and selectively manipulate the internal milieu. For instance, CRRT techniques can be successfully utilized for fluid regulation [25], selective replacement of specific electrolytes, e.g. bicarbonate, without the addition of sodium or fluid [26] or to add substances back to the blood [27]. Similarly the use of combined techniques (e.g. combined plasma filtration adsorption for sepsis, cell-based and hemoperfusion devices for hepatic support

230

Blood Purif 2001;19:227–232

Table 4. Potential indications for dialysis in critically ill patients

Renal replacement Life-threatening indications

Hyperkalemia Acidemia Pulmonary edema Uremic complications

Solute control Fluid removal Regulation of acid-base and electrolyte status Renal support Nutrition Fluid removal in congestive heart failure Cytokine manipulation in sepsis Cancer chemotherapy Treatment of respiratory acidosis in ARDS Fluid management in multiorgan failure

devices) is emerging [28–31]. As a consequence, the traditional indications for renal replacement may need to be redefined. For instance, excessive volume resuscitation, a common strategy used for MOF, may be an indication for dialysis even in the absence of significant elevations in BUN. In this respect it may be more appropriate to consider dialytic intervention in the ICU patient as a form of renal support rather than renal replacement (table 3). This terminology serves to distinguish between the strategy for replacing individual organ function and one to provide support for all organs. Table 4 lists some of the revised

Mehta

Table 5. Renal replacement therapy for ARF: initial choice

Indication

Clinical setting

Modality

Uncomplicated ARF Fluid removal Uremia Increased intracranial pressure Shock Nutrition Poisons Electrolyte abnormalities

Antibiotic nephrotoxicity Cardiogenic shock, CP bypass Complicated ARF in ICU Subarachnoid hemorrhage, hepatorenal syndrome Sepsis, ARDS Burns Theophylline, barbiturates Marked hyperkalemia

IHD, PD SCUF, CVVH CRRT (CVVHD, CVVH, CVVHDF), IHD CRRT (CVVH, CVVHDF) CRRT (CVVH, CVVHDF) CRRT (CVVHD, CVVHDF, CVVH) Hemoperfusion, IHD, CVVHD IHD, CVVHD

indications for dialytic intervention using this approach. It is thus possible to widen the indications for renal intervention and provide a customized approach for the management of each patient. It is also apparent that this approach will become increasingly the norm as we move into the era of using dialysis for nonrenal problems [32– 33].

Recommendations for Initial Choice of Dialysis in the ICU

Despite the lack of definitive results derived from randomized clinical trials [34], it is possible to develop a rational approach to the selection of a dialysis modality for the initial treatment of ARF in critically ill patients. A primary consideration is the availability of a technique at the center and familiarity and comfort of personnel with the technique. The latter point is extremely important with respect to continuous techniques as infrequent use may be associated with a higher incidence of iatrogenic complications [35, 36]. The next consideration is the complexity of the patient, the location in the hospital and need for mobility. Patients with uncomplicated ARF can be treated with IHD or acute peritoneal dialysis (APD) and the choice between these is based on other patient characteristics (e.g., pregnancy, hemodynamic tolerance, access and urgency for treatment). Patients with MOF and ARF can be treated with CRRT or IHD. In general, hemodynamically unstable, catabolic and excessively fluid overloaded patients are ideally treated with CRRT, whereas IHD may be better suited for patients requiring early mobilization and who are more stable. Table 5 depicts a potential therapy for several different clinical scenarios. Amongst continuous therapies, those that include hemofiltration

Indications for Dialysis in the ICU

(CVVH, CVVHDF) may be superior in sepsis or the systemic inflammatory response syndrome because of the ability to more efficiently remove larger molecular weight solutes [23, 37]. For most clinical scenarios, we favor the use of hemodiafiltration techniques that combine dialysis and ultrafiltration and thus are ideal for both small and large molecule clearances [37, 38]. It is important to stress that one of the key factors in the choice of renal replacement is to tailor the therapy to the patient. This implies an ongoing assessment of the patient and modification of the therapy used based on clinical criteria (e.g., in a hemodynamically unstable patient CRRT may be an initial choice, however, when the patient is more stable and needs to be mobilized IHD may be more appropriate). We would suggest that flexibility in utilizing the entire range of renal replacement therapies is an important adjunct to the management of ARF. In summary, management of ARF in the ICU is different from that of ESRD and the dialysis prescription should incorporate the unique characteristics of each patient. Several new methods of dialysis are now available to treat ARF. The application of these techniques is rapidly expanding. Rational use of these techniques should be based on the overall concept of providing renal support. The choice of therapy requires an understanding of the therapeutic potential of a modality and appreciation of the advantages and disadvantages of each technique. Therapeutic alternatives to traditional IHD now permit nephrologists to match the method to the patient. This approach will allow better management of ARF patients and may improve outcome.

Blood Purif 2001;19:227–232

231

References 1 Mallick NP, Gokal R: Haemodialysis. Lancet 1999;353:737–742. 2 Briglia A, Paganini EP: Acute renal failure in the intensive care unit. Therapy overview, patient risk stratification, complications of renal replacement, and special circumstances. Clin Chest Med 1999;20:347–366. 3 Canaud B, Bosc JY, Leray H, Stec F, Argiles A, Leblanc M, Mion C: On-line haemodiafiltration: State of the art. Nephrol Dial Transplant 1998;13(suppl 5):3–11. 4 Ronco C: Haemodialysis filters: What’s new? Curr Opin Nephrol Hypertens 1999;8:709– 713. 5 Leblanc M, Moreno L, Robinson OP, Tapolyai M, Paganini EP: Bicarbonate dialysate for continuous renal replacement therapy in intensive care unit patients with acute renal failure. Am J Kidney Dis 1995;6:910–917. 6 Manns M, Sigler MH, Teehan BP: Continuous renal replacement therapies: An update. Am J Kidney Dis 1998;32:185–207. 7 Adams PL, Adams PF, Bell PD, Navar LG: Impaired renal blood flow autoregulation in ischemic acute renal failure. Kidney Int 1980; 18:68–76. 8 Conger J: The role of blood flow autoregulation in pathophysiology of acute renal failure. Circ Shock 1983;11:235–244. 9 Conger JD: Does hemodialysis delay recovery from acute renal failure? Semin Dial 1990;3: 146–147. 10 Bellomo R, Ronco C: Indications and criteria for initiating renal replacement therapy in the intensive care unit. Kidney Int Suppl 1998;66: S106–S109. 11 Star RA: Treatment of acute renal failure. Kidney Int 1998;54:1817–1831. 12 Paganini EP: Dialysis is not dialysis is not dialysis! Acute dialysis is different and needs help! Am J Kidney Dis 1998;32:832–833. 13 Mehta RL, Letteri J: Renal replacement therapy for ARF, current trends: A survey of US Nephrologists. Am J Nephrol 1999;19:377– 382. 14 Levy EM, Viscoli CM, Horwitz RI: The effect of acute renal failure on mortality. A cohort analysis JAMA 1996;275:1489–1494. 15 Chertow GM, Levy EM, Hammermeister KE, Grover F, Daley J: Independent association between acute renal failure and mortality following cardiac surgery. Am J Med 1998;104: 343–348.

232

16 de Mendonca A, Vincent JL, Suter PM, Moreno R, Dearden NM, Antonelli M, Takala J, Sprung C, Cantraine F: Acute renal failure in the ICU: Risk factors and outcome evaluated by the SOFA score. Intensive Care Med 2000; 26:915–921. 17 Mehta RL, Farkas A, Fowler W, Pascual M: Effect of delayed consultation on outcome from acute renal failure in the ICU (abstract). 28th Annual Meeting of the American Society of Nephrology, Nov 1995. 18 Parsons FM, Hobson SM, Blagg CR, McCracken BH: Optimum time for dialysis in acute reversible renal failure. Description and value of an improved dialyzer with large surface area. Lancet 1961;i:129–134. 19 Fischer RP, Griffen WO, Reiser M, Clark DS: Early dialysis in the treatment of acute renal failure. Surg Gynecol Obstet 1966;123:1019– 1023. 20 Kleinknecht D, Jungers P, Chanard J, Barbanel C, Ganavel D: Uremic and non-uremic complications in acute renal failure: Evaluation of early and ferquent dialysis on prognosis. Kidney Int 1972;1:190–196. 21 Conger JD: A controlled evaluation of prophylactic dialysis in post-traumatic acute renal failure. J Trauma 1975;15:1056–1063. 22 Gillum DM, Dixon BS, Yanover MJ, Kelleher SP, Shapiro MD, Benedetti RA, Dillingham MI, Paller MS, Goldberg JP, Tomford R, Gordon JA, Conger JD: The role of intensive dialysis in acute renal failure. Clin Nephrol 1986;25: 249–255. 23 Mehta RL, McDonald B, Pahl M, Farkas M, Pascual M, Fowler B and the ARF Collaborative Study Group: Continuous vs. intermittent dialysis for acute renal failure in the ICU: Results from a randomized multicenter trial. J Am Soc Nephrol 1996;7:1456. 24 Mukau L, Latimer RG: Acute hemodialysis in the surgical intensive care unit. Am Surg 1988; 54:548–552. 25 Mehta RL: Fluid management in continuous renal replacement therapy. Semin Dial 1996;9: 140–144. 26 Kierdorf HP, Leue C, Arns S: Lactate- or bicarbonate-buffered solutions in continuous extracorporeal renal replacement therapies. Kidney Int 1999;56(suppl 72):S32–S36.

Blood Purif 2001;19:227–232

27 Ronco C, Ghezzi PM, La Greca G: The role of technology in hemodialysis. J Nephrol 1999; 12(suppl 2):S68–S81. 28 Schetz M, Ferdinande P, Van den Berghe G, Verwaest C, Lauwers P: Removal of proinflammatory cytokines with renal replacement therapy: Sense or nonsense? Intensive Care Med, 1995;21:169–176. 29 Tetta C, Bellomo R, Brendolan A, Piccinni P, Digito A, Dan M, Irone M, Lonnemann G, Moscato D, Buades J, La Greca G, Ronco C: Use of adsorptive mechanisms in continuous renal replacement therapies in the critically ill. Kidney Int 1999;56(suppl 72):S15–S19. 30 Ronco C, Brendolan A, Dan M, Piccinni P, Bellomo R, De Nitti C, Inguaggiato P, Tetta C: Adsorption in sepsis. Kidney Int 2000;58(suppl 76):S148–S155. 31 Rahman TM, Hodgson HJ: Review article: Liver support systems in acute hepatic failure. Aliment Pharmacol Ther 1999;13:1255–1272. 32 Schetz M: Non-renal indications for continuous renal replacement therapy. Kidney Int 1999;56(suppl 72):S88–S94. 33 Tetta C, Ronco C, Brendolan A, Irone M, Digito A, Cioffi M, Piccinni P, Dan M, Inguaggiato P, La Greca G: Present and future options in continuous renal replacement therapies of sepsis and MOF. Minerva Anestesiol 1999;65: 419–426. 34 Mehta RL, McDonald B, Pahl M, Farkas M, Pascual M, Fowler B and the ARF Collaborative Study Group: Continuous vs. intermittent dialysis for acute renal failure in the ICU: Results from a randomized multicenter trial. Kidney Int, in press. 35 Manns M, Sigler MH, Teehan BP: Intradialytic renal haemodynamics – Potential consequences for the management of the patient with acute renal failure. Nephrol Dial Transplant 1997;12:870–872. 36 Mehta RL: Renal replacement therapy for acute renal failure: Matching the method to the patient. Semin Dial 1993;6:253–259. 37 Clark WR, Mueller BA, Kraus MA, Macias WL: Dialysis prescription and kinetics in acute renal failure. Adv Renal Replace Ther 1997;4 (suppl 1):64–71. 38 Bellomo R, Ronco C: Acute renal failure in the intensive care unit: Adequacy of dialysis and the case for continuous therapies Nephrol Dial Transplant 1996;11:424.

Mehta

Blood Purif 2001;19:233–237

Design Issues for Clinical Trials in Acute Renal Failure Robert Star Division of Kidney, Urologic and Hematologic Diseases, NIH, Bethesda, Md., USA

Acute renal failure (ARF) is a life-threatening illness whose mortality has remained high despite hemodialysis and other advances in supportive care. Understanding of the pathophysiology of ARF has advanced because of information gained from animal models. However, translation of these advances to the patient bedside has been slow. Because of this difficulty, the NIH sponsored a workshop on ‘Design Issues for Clinical Trials in Acute Renal Failure’ held in Bethesda, MD on September 10– 12, 2000. The meeting brought many issues into the open and enhanced communication between basic and clinical scientists in the field. The following is an initial synthesis of the main conclusions of the conference; however, a detailed conference report will be published elsewhere.

Outside Advice

We asked experts from the areas of ARDS, sepsis, and acute heart failure to tell us why interventions have succeeded or failed in their areas. Drs. Taylor Thompson and Claude Bernard suggested that large-scale interventional trials succeed because information from sound preclinical studies and pilot clinical trials is used to select trial hypotheses and aid in the design of clinical trials. Knowledge about the natural history of the disease allows accurate estimation of control group event rates; information about the intervention allows assessment of the golden window of opportunity and effect size. The third dis-

ABC

© 2001 S. Karger AG, Basel 0253–5068/01/0192–0233$17.50/0

Fax + 41 61 306 12 34 E-Mail [email protected] www.karger.com

Accessible online at: www.karger.com/journals/bpu

cussant, Dr. Robert Califf, suggested that animal studies of acute myocardial infarction provided the wrong targets, thus impeding the development of effective interventions. Instead, plausibility for subsequent clinical trials was provided by an astute clinical observation that opening the disease artery helps; i.e., the heart likes blood flow. He stressed the importance of large simple trials where the entry criteria are consistent with clinical practice. All discussants agreed that successful interventional trials require a network of experienced clinical investigators and study coordinators, access to a large number of patients, an adequate network size to achieve statistical power, and careful and continuous attention to implementation issues. The most common reasons for failed trials include poor design because of low statistical power, inadequate definition of disease, improper selection of endpoints, and failure to control for patient heterogeneity and nonstandardized concurrent therapies. Negative trials may occur because of problems with the intervention including inactive drugs, adverse drug effects that overpower a beneficial effect, missing the golden window of opportunity, wrong or redundant targets, and a misunderstanding of the mechanism of disease. This list was long and daunting. However, recent successes with low volume ventilation in ARDS [1] and an unpublished trial using activated protein C in sepsis indicate that patients with severe illnesses can be successfully treated.

Robert Star, MD Division of Kidney, Urologic and Hematologic Diseases NIDDK, NIH, Building 31, Room 9A35 31 Center Drive MSC 2560, Bethesda, MD 20892-2560 (USA) Tel. +1 301 594 7715, Fax +1 301 496 2830, E-Mail [email protected]

Learn from Past Successes and Mistakes

Drs. Steve Miller, Robert Schrier, Ravi Mehta , and Bert Kasiske reviewed past clinical trials in ARF. Ronco et al. [10] recently found that more ultrafiltration during continuous renal replacement therapy improved survival. This is the most direct evidence to date that treatment of ARF can alter mortality in ICU patients with ARF. The recent report that N-acetylcysteine treatment prevents GFR from falling after radiocontrast [12] promotes antioxidant pathways as new targets for other interventions in ARF. The speakers also reviewed trials that failed to show statistically significant effects. Drug trials have had trouble with delayed diagnosis and randomization [2, 6], delayed administration of study drug [2, 5–7], study druginduced hypotension [2, 9], low statistical power caused by small sample size [5], confounding effects of nonstudy drugs [4], and poorly defined endpoints (e.g., when to start dialysis). Dialysis studies have been hampered by unbalanced randomization caused by small sample size and the heterogeneity of human ARF [8]. Dialysis trials have been difficult to implement because there is a lack of standard criteria for initiating dialysis, multiple methods of delivering dialysis, and difficulties in measuring and comparing dialysis dose across the different methods.

Identify Gaps and Fill Them

The speakers stressed that many critical tools are missing which would greatly enhance the design and implementation of clinical trials in ARF. These missing tools include, for example, an adequate definition of ARF, noninvasive diagnostic and prognostic tests, real-time measurements of renal function, accurate severity of illness scores that measure both the renal and nonrenal components of illness, and an accurate method to measure dialysis dose. There is insufficient epidemiological data for accurate estimation of the event rate in ICU patients with ARF, and inadequate knowledge of the pathophysiology of human ARF, although recent advances in our understanding of delayed graft function are encouraging. Successful tool development will likely require collaboration of basic science and clinical investigators with access to modern tools (genomics, gene arrays, proteomics, etc.), sufficient well-characterized patients, and human biopsy material.

234

Blood Purif 2001;19:233–237

Advice from Clinical Trialists

A clinical trial explores the cause-and-effect relationship between an intervention and an outcome. The ideal trial is a prospective comparison of intervention X to a control group; randomized to reduce confounding third factors, and double blind (masked) to decrease bias. Drs. Ray Bain, Joel Gore, William Macias, and Tom Greene indicated that conduct of successful clinical trials will likely require (1) a testable hypothesis usually based upon a sound understanding of the mechanism of disease, (2) a feasible intervention usually based upon preclinical animal data or pilot human studies, (3) practical entry criteria based upon knowledge of the population at risk, (4) identification of clinically meaningful outcomes, and (5) pilot data to estimate event rate, effect size, and sample size. As discussed by Dr. Califf, the first two criteria are helpful but not necessary; in rare cases, other information can be substituted for mechanism and preclinical animal data. Tom Greene described the difference between internal and external validity. Studies aiming for internal validity are designed to determine the effect of treatment on the primary outcome. These studies are typically Phase II studies with narrow entry criteria that control for extraneous variation in the patient population, have a tightly controlled intervention, and have a sensitive outcome measure (change in GFR). Because this type of study minimizes variation, it can be quite small. In contrast, studies aiming for external validity attempt to generalize to broad clinical practice patterns. Thus, they have broad entry criteria, a realistic population that matches the clinical population of interest, realistic interventions that are clinically achievable, and clinically relevant outcomes (death or dialysis). This tension between demonstrating a treatment effect and having external validity arises during the design phase of every clinical trial. A clinical trial may be summarized as follows: ‘To evaluate the efficacy of [intervention regimen X] vs. control regimen in participants with [population to be studied] as assessed by [primary outcome measurement].’ We will not discuss potential interventions, since they have been reviewed in several recent reviews [3, 11, 13]. The study population is defined by inclusion and exclusion criteria. Inclusion criteria define the disease to be studied, whereas exclusion criteria narrow the population to exclude patents who will not make it through the trial because of noncompliance or serious underlying medical conditions. Because the ARF field lacks a standard definition of disease, this critical portion of the clinical trial design process is currently very difficult.

Star

Many outcomes have been used in interventional trials in ARF. The FDA requires clinically meaningful and compelling outcomes, which include living longer and feeling better; i.e., decreased need for dialysis, and decreased dialysis-free days, but does not necessarily include improved renal function (GFR). Small pilot trials commonly employ GFR as a functional surrogate endpoint, however, the primary outcome in larger trials is death or dialysis. For example, an intervention that improves GFR by 20% would not be clinically compelling. However, if the same intervention decreased the need for dialysis from 50 to 40%, that would be clinically compelling. Obviously, the outcome should be amenable to therapy. This sounds simple, but may be a problem in ARF since nonrenal factors may dominate. Binary outcomes (dead/ alive; dialysis/no dialysis), time to event (death, dialysis), or organ failure-free days may be useful as primary outcome variables because they are structured to pick up events that occur in only a portion of the population (death, dialysis), or are likely amenable to therapy (dialysis-free days). Statistical Issues

tice. Second, one needs an accurate estimate of the treatment effect size. The outcome of a large trial should be clinically relevant, compelling, and sufficient to change clinical practice. The estimate should also be based upon a realistic expectation of the treatment effect. In ARF trials, the estimate must consider the effect of the intervention on renal and nonrenal pathways to the measured outcome. Thus, a drug may decrease renal events by 50%, but if only 20% of the patient outcome is determined by renal effects, the outcome will only change by approximately 10%. Sample size calculations are traditionally performed using an alpha level of 0.05 and a power of 90%. With these four pieces of information, one can estimate the sample size. For example, consider the case of an intervention that changes the mortality of ICU ARF from 50 to 40%. At an alpha of 0.05 and a power of 90%, 540 patients per group are needed to achieve statistical significance. This example illustrates that large trials will be needed for adequate power. A common response to obtaining a large sample size estimate is to increase the hypothesized effect size, and thus lower the sample size. All the statisticians and clinical trialists emphasized that this should not be done; trials should be planned with realistic effect sizes. A statistician can aid in correcting this initial estimate for reduction in event rate, effect of covariates, crossovers, dropouts, etc. These corrections can nearly double the sample size estimate.

Drs. Tom Greene, Bill Macias, and Charles Fisher discussed the statistical issues to consider when designing a clinical trial, including patient heterogeneity, sample size calculations, prespecified data analysis plans, and interim stopping rules. The first two items will be considered here. The extreme patient heterogeneity of human ARF has led to unbalanced randomization in several ARF trials. This problem can be lessened in larger trials because of the law of large numbers. Smaller trials may need to stratify patients during randomization, or adjust the analysis for prespecified covariates measured before randomization. Stratification should more evenly allocate the patients to the different treatment. There was general agreement that stratification will be critical for small and moderatelysized trials; however, stratification may unnecessarily complicate large simple trials. Calculating sample size is a critical design element of any clinical trial. Many previous ARF trials have been underpowered because of overgenerous hypothesized treatment effects and failure to correct sample size for deviations from protocol. The sample size calculation requires accurate knowledge of the control outcome which is usually based upon epidemiologic information of patients with the same inclusion/exclusion criteria used for the trial. Ideally, this should be a conservative estimate that includes recent improvements in clinical prac-

Drs. Ann Willoughby and William Macias described the design issues for the formulation of a successful clinical trial network. A clinical trial network and an individual clinical trial have a well-defined ‘life cycle’ that ends with a publication; this end should be kept in mind at all times. The process begins with the formation of a small planning committee that explicitly defines in writing the goals and objectives of the network along with the roles and responsibilities of every member of the project. It is important to define in writing who has the responsibility for defining the scientific agenda, pacing the activities, distributing resources, and establishing sanctions. Strong leadership is absolutely essential; clinical trialists and statisticians should be involved from the start. The planning committee, along with the data coordinating center, may be formed before the rest of the network is developed. The sequencing of studies and trials is defined by discussions of the state of the art, and key informative questions including what is known about the epidemiology, patho-

Design Issues for Clinical Trials in ARF

Blood Purif 2001;19:233–237

Design of Clinical Trial Networks

235

genesis, and natural history of the disease. For example, is the field ready for therapeutics clinical trials? Are pilot studies needed? These discussions may take 1 year, but must yield definable pieces of work including specifications for epidemiologic studies, hypothesis generating pilot studies, and definitive trials. These mechanisms will be used throughout the lifetime of the network to select and prioritize studies and trials. For each trial, the network will develop a detailed written protocol, a prespecified statistical analysis plan, a monitoring plan, and a validation plan. These plans will specify the data collection needs, systems (database) support, and the structure of the final report. Finally, the network must develop a welldesigned case report form for capturing data. The network will include basic scientists, clinical researchers, clinical trialists, and statisticians. A well-crafted external advisory board, liaison to professional societies, high visibility at key scientific meetings, and advocacy groups are also essential. Once the clinical and basic science centers have been added, the network sets up a training structure to train the principal investigators, study coordinators, and other study personnel. Ideally, this training allows all study personnel to understand the issues of the trial and ‘buy into’ the trial. This training creates a cadre of excellent and dedicated study personnel that can properly execute the network protocols. The implementation of the trial protocol, monitoring plan, and validation plan are discussed below. The data should be transmitted from each clinical center via the Internet to a versatile data warehouse. This centralized data collection allows the trial to be continuously monitored; at the end of the trial, the prespecified analysis to be rapidly performed, allowing for timely publication of the clinical trial results.

gov/cder]. These provide a unified standard for designing, conducting, recording, and reporting trials that involve the participation of human subjects. For example, all trial procedures, policies, and actions should be documented in writing. The data coordinating center should monitor ‘patient quality’ in real time to prevent randomization of patients who do not meet the inclusion and exclusion criteria of the trial. The trial must monitor drop-ins, dropouts, and noncompliant patients. Standard operating procedures must be designed to maintain the trial intervention (e.g., level of dialysis dose) and standardize other nontrial interventions (e.g., dopamine). The data quality should be monitored by validating a portion of the submitted data. The conference also considered regulatory (e.g., Institutional Review Board, Food and Drug Administration) issues that will not be discussed here.

Conclusions

Management of patients with ARF is challenging because the existing evidence base is weak or lacking. Researchers must aim to obtain conclusive evidence at the level of evidence-based medicine because this type of information is right for patients, is critical for health policy decisions, and is increasingly used for reimbursement decisions. Gathering the necessary tools and clinical evidence will require significant work and resources. The design of a large-scale interventional trial network in ARF will be difficult. It is up to the community and NIH to obtain advice from a broad range of opinions, and adhere to the highest standards when designing the network. However, recent clinical trials indicate that ARF is amenable to treatment.

Implementation Issues

While planning and data analysis phase of a clinical trial is critical, the majority of the effort is placed into implementation of the trial, including recruitment, retention, outcome assessment, and monitoring. The trial team must be ready to face and rapidly solve all problems that occur during this stage. This is commonly aided by training workshops for the study personnel and by computerized methods to monitor the trial in real time. Recruitment is always an issue, since it commonly lags far behind the initial goals. Clinical trials must follow good clinical practice guidelines as outlined in CFR Title 21, CFR Title 45, and the International Conference on Harmonization [www.fda.

236

Blood Purif 2001;19:233–237

Star

References 1 The Acute Respiratory Distress Syndrome Network: Ventilation with lower tidal volumes as compared with traditional tidal volumes for acute lung injury and the acute respiratory distress syndrome. N Engl J Med 2000;342:1301– 1308. 2 Allgren RL, Marbury TC, Rahman SN, Weisberg LS, Fenves AZ, Lafayette RA, Sweet RM, Genter FC, Kurnik BRC, Conger JD, Sayegh MH: Anaritide in acute tubular necrosis: Auriculin Anaritide Acute Renal Failure Study Group. N Engl J Med 1997;336:828–834. 3 Bonventre JV: Mechanisms of ischemic acute renal failure. Kidney Int 1993;43:1160–1178. 4 Chertow GM, Sayegh MH, Allgren RL, Lazarus JM: Is the administration of dopamine associated with adverse or favorable outcomes in acute renal failure? Am J Med 1996;101:49– 53. 5 Franklin SC, Moulton M, Sicard GA, Hammerman MR, Miller SB: Insulin-like growth factor I preserves renal function postoperatively. Am J Physiol 1997;272:F257–F259.

Design Issues for Clinical Trials in ARF

6 Hirschberg R, Kopple J, Lipsett P, Benjamin E, Minei J, AlbertsonT, Munger M, Metzler M, Zaloga G, Murray M, Lowry S, Conger J, McKeown W, O’Shea M, Baughman R, Wood K, Haupt M, Kaiser R, Simms H, Warnock D, Summer W, Hintz R, Myers B, Haenftling K: Multicenter clinical trial of recombinant human insulin-like growth factor I in patients with acute renal failure. Kidney Int 1999;55: 2423–2432. 7 Lewis J, Salem MM, Chertow GM, Weisberg LS, McGrew F, Marbury TC, Allgren RL: Atrial natriuretic factor in oliguric acute renal failure. Am J Kidney Dis 2000;36:767–774. 8 Mehta R, McDonald B, Gabbai F, Pahl M, Farkas A, Pascual M, Fowler W, ARF Collaborative Study Group: Continuous versus intermittent dialysis for acute renal failure in the ICU: Results from a randomized multicenter trial (abstract). J Am Soc Nephrol 1996;7:1457.

9 Rahman SN, Kim GE, Mathew AS, Goldberg CA, Allgren R, Schrier RW, Conger JD: Effects of atrial natriuretic peptide in clinical acute renal failure. Kidney Int 1994;45:1731–1738. 10 Ronco C, Bellomo R, Homel P, Brendolan A, Dan M, Piccinni P, La Greca G: Effects of different doses in continuous veno-venous haemofiltration on outcomes of acute renal failure: A prospective randomized trial. Lancet 2000; 356:26–30. 11 Star RA: Treatment of acute renal failure. Kidney Int 1998;54:1817–1831. 12 Tepel M, van der Griet M, Schwarzfeld C, Laufer U, Liermann D, Zidek W: Prevention of Radiographic-contrast-agent induced reductions in renal function by acetylcysteine. N Engl J Med 2000;343:180–184. 13 Thadhani R, Pascual M, Bonventre JV: Medical progress – Acute renal failure. N Engl J Med 1996;334:1448–1460.

Blood Purif 2001;19:233–237

237

Blood Purif 2001;19:238–244

Prescription of Adequate Renal Replacement in Critically Ill Patients E.P. Paganini a N.S. Kanagasundaram a B. Larive b T. Greene b a Section

of Dialysis & Extracorporeal Therapy, Department of Nephrology/Hypertension, and of Biostatistics and Epidemiology, The Cleveland Clinic Foundation, Cleveland, Ohio, USA

b Department

Background

One of the more useful techniques in the field of chronic dialysis support is the ability of defining a delivered dialysis dose. This has been traditionally accomplished using urea kinetic modeling either via direct dialysis quantification [1, 2] or using formulae or urea reductions [3–8] to derive a specific urea removal rate in the chronic dialysis population. Improved survival in chronic patients who received a higher delivered dialysis dose has also been described. Although the rather simplistic approach offered in the chronic population has used the pre/post urea concentration as a method of arriving at total uremic control, problems with single pool vs. multiple pool kinetics [9–12] as exhibited by urea rebounding post procedure, have given rise to various formulas applied to this measurement. To date, there has been no verification of this dosage technique applied to the delivery of acute dialytic intervention. While authors pointed to a lack of correlation between the maintenance of patients at various levels of ‘steady-state’ urea and outcome differences, neither the ‘dose’ of delivered dialysis being utilized in this patient population, nor the potential impact of that dose upon outcome has been recently studied. Membrane structure as a basis for differing outcome in acute renal failure

ABC

© 2001 S. Karger AG, Basel 0253–5068/01/0192–0238$17.50/0

Fax + 41 61 306 12 34 E-Mail [email protected] www.karger.com

Accessible online at: www.karger.com/journals/bpu

(ARF) has also been suggested [13, 14], but no clear description of delivered dialysis dose was studied. Mortality statistics in patients with ARF in the ICU requiring dialytic support have remained high ranging from 50 to 100%, especially in patients requiring ventilatory support [15–17]. Attempts at classifying patients according to co-morbid parameters, physiologic status or disease history resulting in survival outcome scoring systems have been made and have had limited success in some subgroups of patients [18–21]. Varying methods of dialytic support have also been utilized in the care of patients with ARF with varying levels of outcome improvement [22]. In none of these studies, however, has data on the quantity of dialysis or dialysis adequacy been described. Dialysis treatment regimens have largely been dictated by practical convention, financial considerations, personnel availability or local resource and have generally been extrapolated from the chronic dialysis treatments delivered to patients with chronic renal failure. Validation of these extrapolations has been wanting. Such diverse issues as dialysate composition, hemodynamic consequence of the dialytic therapy, drug kinetic differences, or variation in dialysis-induced compartmental volume changes have been described as adding to improved patient outcome in acute renal failure. Gillum et al. [23] in a prospective study of 17 pairs of patients

E.P. Paganini, MD The Cleveland Clinic Foundation, Department of Dialysis and Extracorporeal Therapy 9500 Euclid Avenue Cleveland, OH 44106 (USA) Tel. +1 216 444 5792, Fax +1 216 444 7577, E-Mail [email protected]

found no advantage of intensive dialysis (maintaining predialysis BUN !60 mg/dl) over the non-intensive therapy (predialysis BUN !100 mg/dl) in the ICU for patients with ARF. However, neither control over the type or amount of nutrition received by the patients, nor the metabolic characterization of patients was undertaken, assuming that the randomization process would suffice. Notwithstanding, the intensive therapy group showed less GI bleeding and a trend of better survival though not significant perhaps because of its small sample size. Three retrospective trials done before this had shown benefit however [24–26], and recent prospective trials using aggressive intermittent dialysis [27] or hemofiltration [28] have clearly demonstrated benefit. Membrane bio-incompatibility has been linked to activation of complement and a worse outcome in ARF. Hakim et al. [13] had shown in a prospective randomized study of 72 patients that 62% of patients using polymethyl-methacrylate membranes recovered renal function compared to 37% of those using cuprophane membranes, a difference that was significant after adjustment for co-morbidities. Similar results have been shown by Schiffl et al. [14] in a group of 52 patients who were randomized to receive dialysis with cuprophane or polyacrylonitrile membranes. The former group had greater mortality (65 vs. 35%) and lesser renal recovery. Activation of C3a and LTB4 as well as altered neutrophil function were demonstrable. Any study of ARF would therefore have to ensure that biocompatible membranes alone were used. The consequences of unstable hemodynamics in the setting of ARF are linked to the loss of vascular autoregulation that accompanies ARF. In animal models of ARF, hypotension was associated with the development of new renal injury and infarction [24], and all methods of dialysis currently used aim to minimize the occurrence of such events. In this context, one naturally assumes that continuous renal replacement therapies (CRRT) would maintain superior stability and in general this is true [25]. However, improved techniques of intermittent dialysis delivery including variable sodium and ultrafiltration modelling and dialysate modifications allow safe therapy in the great majority of patients who are not initially unstable. Sandroni et al. [22] had only a 5% incidence of significant hypotension in 547 consecutive ICU-based intermittent hemodialysis (IHD) therapies. In a sense, patients will preselect themselves to the method of RRT applicable to them based on their hemodynamic status, and this has to be factored into any study. Dialysate composition is now well standardized for IHD, acetate being rarely used in the ICU setting. It is therefore unlikely that

this will contribute significantly to future assessments of outcome, but it is indubitable that acid-base disturbances may have contributed to the poor outcomes in ARF noted in the past. It may seem intuitive that the method of dialysis delivery is a major outcome predictor in all salvageable patients. There are several theoretical reasons for this supposition. The safer hemodynamic effects of CRRT have been mentioned. Fluid removal is greater with CRRT than intermittent therapy for the same reason [29]. Greater delivery and more flexibility with parenteral nutrition is also possible [30] and thereby CRRT supports any positive impact that nutrition has in the outcome of ARF. Greater solute clearances have been noted in CRRT [31] and in addition, mediators of inflammation are removed. Polyacrylonitrile membranes have been shown to deplete factor D by adsorption, thus inhibiting the alternate complement pathway [32]. Bellomo et al. [33] showed that CVVHD was capable of clearing both IL-6 and IL-8 from the blood of septic patients. Tonneson et al. [34] demonstrated extraction of IL-1 and TNF-· from septic patients treated by CAVHD. CRRT also provides for superior regulation of drug pharmacokinetics. Despite these obvious facts, there is a remarkable lack of data on head-to-head comparison between IHD and CRRT. There is at this stage no truly prospective randomized study looking at outcomes in these two groups, though Mehta et al. [35] have just reported a completed study of the same. Based upon an extensive literature review, Jakob et al. [36] compiled 15 trials which incorporate comparisons of outcome made between a total of 522 patients who received IHD and 651 patients who received CRRT. Only 4 studies had prospective data and none were randomized. Remarkably, no clear evidence of benefit could be established despite adjustment for co-moribidities. Preliminary data from Mehta suggests the same. This issue deserves a closer look in a major trial but at this time, it can be safely assumed that the magnitude of difference in outcome seems minimal.

Adequate Renal Replacement in Critically Ill Patients

Blood Purif 2001;19:238–244

Method of Establishing a Dialysis Dose Measurement

A mechanism for determining dialysis dose in the patients with ARF has not been established. Methods have been extrapolated from the ESRD population. Many of the assumptions which were possible in the ESRD patient may not be possible in the patient with ARF, especially in the ICU setting where change is both dramatic

239

Fig. 1. Weight variability during the support period with intermittent hemodialysis. a Unstable patient. b Stable

patient.

and frequent. Comparison of blood-side and dialysateside techniques for this measurement was done. We used the direct dialysis quantification method as our standard since this reflects the actual removal of marker substance (urea, creatinine). This method is not subject to variations in therapy technique, and can be used to calculate actual urea volume and generation. The classical techniques are, however, quite cumbersome, and thus not applicable to large clinical studies. Newer methods of collecting dialysate values will also be examined for feasibility. The clearance-based methods to be evaluated are: (1) equilibrated variable volume double pool urea kinetic model using timed (30 and 60 min) postdialysis samples; (2) an estimate of the equilibrated variable volume double-pool Kt/VUREA from immediate postdialysis sample; (3) an equivalent renal clearance derived from both the generation (G) and equilibrated time averaged concentrations (TAC) of the marker substance. This final method could form the basis for attempts at unifying the dose description for both intermittent and continuous hemodialysis.

Results

After obtaining IRB and patient consent, a series of consecutive ICU patients requiring dialytic support were studied using the various methods described above. Weight among these patients varied significantly during their dialysis support as noted in figure 1a and b. The eKt/Vdp is Daugirdas’ [11] estimate for the dpwKt/ VUREA that would be measured if an ‘equilibrated’ postdialysis BUN were drawn. The formula for this estimate is: eKt/Vdp

= Kt/Vsp – (0.4 W

Kt/V th

+ 0.02)

where Kt/Vsp is calculated according to Daugirdas II (see below) and th is the treatment time in hours. This is the formulation being used in the NIH Modification of Mortality in Hemodialysis (HEMO) Trial currently underway to estimate the dose of delivered dialysis achieved in the chronic dialysis patient. The calculation for the single pool Kt/V by Daugirdas follows the formula: Kt/Vsp = –ln

冉CC

post pre

冊 冉

– 0.008 W th + 4 – 3.5 W

Cpost Cpre

冊 WUF W

post

where Cpre and Cpost are the pre- and immediate postdialysis BUN, th is time in hours of the treatment, UF is the weight loss (kg) and Wpost is the postdialysis patient weight (kg). Focusing on the dialysis session, dpvvKt/

240

Blood Purif 2001;19:238–244

Paganini/Kanagasundaram/Larive/Green

Fig. 2. Intradialytic deviation based on postdialysis curve fit.

Fig. 3. 30- or 60-min post-rate estimate of eKt/V vs. eKt/V using the four-point curve fit rate equation.

VUREA describes the effectiveness of a single dialysis treatment session. Thus with this term, one describes an effective session clearance, rather than a specific blood toxin level. Another method of total treatment measurement is based upon solute removal formulations. During an intermittent dialysis procedure, solute mass balance can be derived from pre- and postdialysis blood values of a substance according to the concept:

patient at the time of the session, the standard rate equation seemed to be a good estimate of the session dose delivery. Postdialysis draws of 0, 5, 15, 30, 60 min and an immediate predialysis BUN determination were used to define the curve. The agreement with these values and the predicted curve was excellent and is shown using the fourpoint curve fit in figure 3.

Total MarkerPRE + Generated Marker =

Discussion

Total MarkerPOST + Removed Marker CePRE W VPRE + G W tDIAL = CePOST W VPOST + R

where Ce is the equilibrated blood concentration of the marker, V is the volume of distribution of the marker, G is the marker appearance rate, and tDIAL is the time of dialysis. The individual sessions are then evaluated both as predicted and also measured and plotted against one another. Several of these sessions reveiled a dramatic intra-therapy variation in urea decline reflecting a significant urea mass balance error during that session (fig. 2). These errors may be accounted for by regional flow variations, blood access recirculation or varying pressor agent doses utilized during the particular session. However, using the rebound blood results, after having accounted for the variation in urea generation in the particular

Adequate Renal Replacement in Critically Ill Patients

The issue of dialysis dose has never been addressed in a major clinical trial of ARF. Part of the difficulty lies in defining delivered dialysis dose. In chronic dialysis, this has been traditionally accomplished using urea kinetic modeling based upon direct dialysis quantification methods or simplified formulae. Improved survival in chronic patients who received a greater delivered dialysis dose has been described [37–39]. However, dosing in ARF needs to account for highly variable body water volumes and varying urea generation rates, as well as different methods of dialysis. Hence, it has never been properly quantified or studied. Preliminary data from our group seem to suggest that survivors of ARF in the ICU setting did indeed receive a greater dialysis dose, but dose estimation using Kt/VUREA and URR varied widely from

Blood Purif 2001;19:238–244

241

Fig. 4. Patient mortality and relationship to

delivered dialysis dose – Retrospective Requistry Data.

actual dose prescribed (fig. 4). The first task is to define dialysis dosing in ARF and show how it could be applied to both methods of dialysis. Pragmatic dosing would take into account volume overload, electrolyte disturbances and acidosis. It also makes intrinsic sense to use a ‘uremic toxin’ if such could be defined. Notwithstanding, urea kinetics still remain one possible standard. Kt/VUREA has been used by Leblanc et al. [40]. When applied to CRRT, K or dialyzer urea clearance was calculated as: K = Qdo W (Cdo/Cp), where Qdo is dialysate outflow rate, Cdo is outlet urea concentration and Cp is plasma urea concentration. Error resulted from simplified measures of V or body water content which can vary greatly from the norm of 55% of body weight. As mentioned, Kt/V applied to IHD would not only incorporate this error but also that due to imperfect estimation of K. The latter is estimated by blood side measurements and accuracy depends on constant urea generation rates, which may not be valid in ARF. A more sophisticated dose assessment of CRRT was undertaken by Clark et al. [31] – the so-called ‘three-point urea kinetic modeling’. Steady-state and pre-steady-state data while on CRRT was used to derive the body water volume (V) and urea generation (GUREA). An expression by Ofsthun et al. [41] which uses blood flow, ultrafiltration and replacement fluid flow rates and the urea sieving coefficient (assumed to be unity) was used to derive K. Thus the Kt/V was obtained. Protein catabolic rate (PCR) was derived from G and non-urea nitrogen generation rates and normalized PCR (nPCR) calculated as per

242

Blood Purif 2001;19:238–244

Borah [42]. The expression nPCR/(Kt/V) described urea appearance and clearance and could be related to steadystate blood urea. Applying this expression however to IHD for comparison resulted in assumptions about urea generation and volume status. Plasma clearances of urea and creatinine giving a virtual GFR have also been used to describe dialysis dose in CRRT. The dialysate outflow rate in this case is considered to be equal to the volume of plasma cleared of urea or creatinine. The generalization of this approach to patients on IHD is difficult. Calculated dialysis dose does not match prescribed dose for several reasons: in-vivo clearances of urea are 10–15% lower than the values supplied [43]. Patient instability often results in shorter treatments. Heparinfree dialysis or CRRT with low-dose heparin often results in progressive clotting of the dialyzer fibers, and despite routine checks and changes, some reduction in urea clearances is to be expected. Golper et al. [44] showed that the sieving coefficients of small solutes do not fall in 43 h of use. We have demonstrated recirculation in temporary catheters are of the order of 5–38% depending on site, blood flow and reversal of lines [45, 46]. There is no significant data on postdialysis urea rebound (PDUR) in ARF patients on IHD, which represents late urea entry into the blood pool from other body pools and therefore diminishes urea clearances. Initial data on 5 patients [47] suggests that PDUR may not be complete even after 30 min, however our more recent data seems to point toward a final equilibration at this time, if urea generation is accounted for in the absolute determination. Hence,

Paganini/Kanagasundaram/Larive/Green

IHD would result in less clearance of any solute following two-pool kinetics for a given amount of dialysis than CRRT.

Conclusion

In summary, there is an obvious need to develop a reliable and practicable dosing formula for dialytic interventions in ARF. This formula should be applicable to both intermittent and continuous forms of dialysis delivery,

and should relate to an acceptable standard. Once reliable dosing formulae are developed, there is an urgent need of examining the effects of delivered dialysis dose on outcome of patients with ARF. Prospective data from both Ronco and Shiffel point toward a reaffirmation of the retrospective data from our registry showing a definitive outcome influence of delivered dialysis dosage. If this is indeed the case, the long-standing high mortality of the critically ill patient requiring dialytic support should begin to fall once ‘adequate’ dialysis is delivered.

References 1 Garred LJ: Dialysate-based kinetic modeling. Adv Ren Replace Therapy 1995;2:305–318. 2 Ing TS, Yu AW, Wong FKM: Collection of a representative fraction of total spent dialysate. Am J Kidney Dis 1995;25:810–812. 3 Keshaviah P, Star RA: A new approach to dialysis quantification: An adequacy index based on solute removal. Semin Dial 1994;7:85–90. 4 Gotch FA: Kinetic modeling in hemodialysis; in Nissensen AR, Gentile DE, Fine RN (eds): Clinical Dialysis, ed 2. Norwalk, Appleton & Lange, 1989, pp 118–146. 5 Daugirdas JT: Simplified equations for monitoring Kt/V, PCRn, eKt/V, and ePCRn. Adv Ren Replace Ther 1995;2:295–304. 6 Daugirdas JT: The post:pre-dialysis plasma urea nitrogen ratio to estimate Kt/V and nPCR: Mathematical modelling. Int J Artif Organs 1989;12:411–416. 7 Depner TA, Cheer AY: Modelling urea kinetic with two vs. three BUN measurements: A critical comparison. ASAIO Trans 1989;35:499– 502. 8 Depner TA: Prescribing Hemodialysis: A Guide to Urea Kinetic Modeling. Boston, Kluwer Academic, 1991. 9 Schneditz D, Van Stone JC, Daugirdas JT: A regional blood circulation alternative to inseries two-compartment urea kinetic modeling. ASAIO J 1993;39:M573–M577. 10 Daugirdas JT: Second-generation logarithmic estimates of single pool variable volume Kt/V: An analysis of error. J Am Soc Nephrol 1993;4: 1205–1213. 11 Schneditz D, Daugirdas JT: Formal analytic solution to a regional blood flow and diffusionbased urea kinetic model. ASAIO J 1993;40: M667–M673. 12 Schneditz D, Zaluska WT, Morris AT, Fan Z, Kaufman AM, Levin NW: Effect of ultrafiltration on peripheral urea sequestration in hemodialysis patients. J Am Soc Nephrol 1996;7: 1525. 13 Hakim RM, Wingard RL, Parker RA: Effect of the dialysis membranes in the treatment of patients with acute renal failure. N Engl J Med 1994;331:1338–1347.

Adequate Renal Replacement in Critically Ill Patients

14 Schiffl H, Lang SM, Konig A, Strasser T, Haider MC, Held EL: Biocompatible membranes in acute renal failure: Prospective case controlled study. Lancet 1994;344:570–572. 15 French Multicenter Study Group – Mezzarobba P, Agostini MM, Loirat P, Kleinknecht D, Brivet F, Landais P: Comparison of SAPS, Apache II and OSF in the evaluation of severe acute renal failure. Intensive Care Med 1992; 18(suppl 2):S65. 16 Bosteels V, Verberckmoes R, Vandenbroucke J, Michielsen P: Prognosis of acute renal insufficiency; in Pathogenesis and Clinical Findings with Acute Renal Failure. Stuttgart, Thieme, 1969, pp 234–236. 17 Mehta RJ: Therapeutic alternatives to renal replacement for critically ill patients in acute renal failure. Semin Nephrol 1994;14:64–82. 18 Lohr JW, McFarlane MJ, Grantham JJ: A clinical index to predict survival in acute renal failure patients requiring dialysis. Am J Kidney Dis 1988;11:254–259. 19 Cioffi WG, Ashikaga T, Gamelh RL: Probability of surviving postoperative acute renal failure. Development of a prognostic index. Ann Surg 1984;200:205–211. 20 Rasmussen HH, Pitt EA, Ibels LS, McNeil DR: Prediction of outcome in acute renal failure by discriminant analysis of clinical variables. Arch Intern Med 1985;145:2015–2018. 21 Paganini EP, Haistenberg WK, Goormastic M: Risk modeling in acute renal failure requiring dialysis. The introduction of a new model. Clin Nephrol 1996;44:206–211. 22 Sandroni S, Arora N, Powell B: Performance characteristics of contemporary hemodialysis and venovenous hemofiltration in acute renal failure. Ren Fail 1992;14:571–574. 23 Gillum DM, Dixon BS, Yanover MJ, et al: The role of intensive dialysis in acute renal failure. Clin Nephrol 1986;25:249–255. 24 Kleinknecht D, Jungers P, Chanard J, Barbanel C, Ganeval D: Uremic and non-uremic complications in acute renal failure: Evaluation of early and frequent dialysis on prognosis. Kidney Int 1972;1:190–196.

25 Conger JD, Schultz MF, Miller F, Robinette JB: Responses to hemorrhagic arterial pressure reduction in different ischaemic renal failure models. Kidney Int 1994;46:318–323. 26 Parsons FM, Hobson SM, Blagg CR, McCracken BH: Optimum time for dialysis in acute reversible renal failure. Lancet 1961;i:129– 134. 27 Schiffl H, Lang SM, Haider M: Effects of biocompatibility and frequency of hemodialysis in acute renal failure (abstract). ASAIO Journal 1998;44:69A. 28 Ronco C, Bellomo R, Homel P, et al: Effects of different doses in continuous veno-venous haemofiltration on outcomes of acute renal failure: A prospective randomized trial. Lancet 2000; 355:26–30. 29 Davenport A, Will EJ, Davidson AM: Improved cardiovascular stability during continuous modes of renal replacement therapy in critically ill patients with acute hepatic and renal failure. Crit Care Med 1993;21:328–338. 30 Simpson HKL, Allison MEM, Telfer ABM: Improving the prognosis in acute renal and respiratory failure. Ren Fail 1987;10:45–54. 31 Clark WR, Mueller BA, Alaka KJ, Macias WL: A comparison of metabolic control by continuous and intermittent therapies in acute renal failure. J Am Soc Nephrol 1994;4:1413–1420. 32 Gasche Y, Pascual M, Suter PM, Favre H, Chevrolet JC, Schifferli JA: Complement depletion during hemofiltration with polyacrylonitrile membranes. Nephrol Dial Transplant 1999;11:117–119. 33 Bellomo R, Tipping P, Boyce N: Interleukin-6 and interleukin-8 extraction during continuous venovenous hemofiltration in septic acute renal failure. Ren Fail 1995;17:457–466. 34 Tonneson E, Hansen MB, Hohndorf K, et al: Cytokines in plasma and ultrafiltrate during continuous arteriovenous hemofiltration. Anaesth Intens Care 1993;21:752–758. 35 Mehta RJ, McDonald B, et al: Continuous vs intermittent dialysis for acute renal failure in the ICU: Results of a randomized multicenter trial (abstract). J Am Soc Nephrol 1996;7: 1457.

Blood Purif 2001;19:238–244

243

36 Jakob SM, Frey FJ, Uehlinger DE: Does continuous renal replacement therapy favourably influence the outcome of patients? Nephrol Dial Transplant 1996;11:1250–1255. 37 Charra B, Calemard E, Ruffet M, et al: Survival as an index of adequacy of dialysis. Kidney Int 1992;41:1286–1291. 38 Parker TF, Husni L, Huang W, et al: Survival of hemodialysis patients in the United States is improved with a greater quantity of dialysis. Am J Kidney Dis 1994;23:670–680. 39 Hakim RM, Breyer J, Ismail N, et al: Effects of dose of dialysis on morbidity and mortality. Am J Kidney Dis 1994;23:661–669. 40 Leblanc M, Bonnardeaux A, Cardinal J: Kt/V in continuous dialysis techniques. Semin Dial 1995;8:51–52.

244

41 Ofsthun N, Colton C, Lysaught M: Determinants of fluid and solute removal removal rates during hemofiltration; in Henderson L, Quelhorst E, Baldamus C, Lysaught M (eds): Hemofiltration. Berlin, Springer, 1986, pp 17–39. 42 Borah MD, Schonfeld PY, Gotch FA, Sargent JA, Wolfson M, Humphreys MH: Nitrogen balance during intermittent dialysis therapy of uremia. Kidney Int 1978;14:491–500. 43 Barth R, Caruso C, Li J, et al: Accurate in vivo measurement of dialyser urea clearance and mass transfer coefficient: A daunting task (abstract). J Am Soc Nephrol 1992;3:353.

Blood Purif 2001;19:238–244

44 Golper T, Erbeck K, Roberts P, et al: Small solute sieving coefficients using hemodialysers as filters in continuous venovenous hemofiltration (abstract). J Am Soc Nephrol 1992;3:367. 45 Twardowski ZJ, Van Stone JC, Jones ME: Blood recirculation in intravenous catheters for hemodialysis. J Am Soc Nephrol 1993;3:1978– 1981. 46 Kelber J, Delmez JA, Windus DW: Factors affecting delivery of high efficiency dialysis using temporary vascular access. Am J Kidney Dis 1993;22:24–29. 47 Leblanc M, Tapolyai M, Paganini EP: What dialysis dose should be provided in acute renal failure? A review. Adv Ren Replace Ther 1995; 2:255–264.

Paganini/Kanagasundaram/Larive/Green

Blood Purif 2001;19:245–250

Validation of the Blood Temperature Monitor for Extracorporeal Thermal Energy Balance during in vitro Continuous Hemodialysis Shahriar Rahmati a Federico Ronco a Margaret Spittle a Alice T. Morris a Christian Schleper a Laura Rosales a Alan Kaufman b Richard Amerling b Claudio Ronco a, b, c Nathan W. Levin a a Renal

Research Institute and b Beth Israel Medical Center, New York, N.Y., USA; of Nephrology, St. Bortolo Hospital, Vicenza, Italy

c Department

Introduction

Continuous renal replacement therapies (CRRT) are today considered a well-tolerated and efficient group of treatments for acute renal failure in critically ill patients [1–12]. The evolution in technology of CRRT has only partially followed the more sophisticated evolution that took place in the equipment for chronic hemodialysis patients. In such patients, the increased morbidity and the progressively increased age, require a gentle and carefully monitored hemodialysis therapy. To achieve such results, on-line monitoring techniques have been developed including urea sensors, temperature sensors, blood volume sensors and biofeedback systems [13]. We will try to analyze how this new technology could have a positive impact in acute patients and how it could be implemented in the present equipment for CRRT [14–16].

calculation of the thermal balance in a single session, but it also permits a target value of body temperature to be achieved in the patient at the end of the session. This is obtained by a controlled variation in the temperature of the dialysis solution. The possibility to deliver a thermal bolus by injection of cold saline in the venous line or by a sudden decrease in the temperature of dialysate, allows for the measurement of the recirculation in the vascular access. These features of the blood temperature monitor (BTM) would be of great interest for the acute patient: the critically ill patient may frequently be hyperthermic and he may require a progressive cooling to improve hemodynamic stability and prevent thermal damage. Alternatively, when large volumes of fluid are exchanged in hemofiltration, the risk of a negative thermal balance is greatly enhanced. For all these aspects, the use of the BTM may represent a definite improvement in the conduction of hemodialysis treatment in acute patients.

On-Line Temperature Sensors

This technology has been proposed in some of the latest dialysis machines for chronic patients. The technology consists of the placement of isolated temperature sensors on the arterial and venous lines of the extracorporeal circulation. Not only does this equipment allow for the

ABC

© 2001 S. Karger AG, Basel 0253–5068/01/0192–0245$17.50/0

Fax + 41 61 306 12 34 E-Mail [email protected] www.karger.com

Accessible online at: www.karger.com/journals/bpu

Characteristics of the Blood Temperature Monitor

The studied Blood Temperature Monitor (BTM, Fresenius Medical Care, Bad Homburg, Germany) is a plugin module that fits in the 2008H series of dialysis ma-

Shahriar Rahmati, MD Renal Research Institute 207 East 94 Street, Suite 303, New York, NY 10128 (USA) Tel. +1 212 360 4900, Fax +1 212 360 7233, E-Mail [email protected]

al flow rates and has never been tested for flow rates typical for CRRT, a study was designed to test the functionality and accuracy of the BTM at low dialysate and blood flow rates.

Methods

Fig. 1. Diagram of test equipment circuit.

chines and serves to control the extracorporeal thermal energy balance of patients. The BTM offers two programs for temperature control to increase hemodynamic stability (body temperature control and energy control). Additionally it measures recirculation noninvasively by thermodilution. Independently of the chosen program, the BTM measures the thermal energy balance of the patients at any given time during treatment. The body temperature control algorithm of the BTM allows automated decrease or increase of patients’ core temperature at a constant rate relative to the body temperature measured at the beginning of treatment. The body temperature control program requires several recirculation measurements every 30–60 min during treatment. The energy control program allows the user to choose how much thermal energy per hour (kJ/h) shall be fed to or withdrawn from the patient. Energy control does not require recirculation measurement; it can be done manually if necessary. A processor-controlled dialysis machine communicates with the BTM during the treatment. One of the important data transferred between the dialysis machine and the BTM is the actual dialysate temperature coming from the dialysis machine. The BTM works correctly only when the actual dialysate temperature reading is accurate. The lowest dialysate flow rate of a 2008H dialysis machine for conventional dialysis treatment is limited to 300 ml/min. A modified 2008H dialysis machine for the CRRT enables the user to decrease the dialysate flow to 100 ml/min. Since the BTM was designed for convention-

246

Blood Purif 2001;19:245–250

A 19-liter water bath at a constant temperature simulated the patient. The dialysis machine was set as for a regular dialysis treatment in CRRT mode. The ends of both the arterial and venous lines were immersed in the water bath, and the distance between them was fixed (ca. 35 cm). The machine’s blood pump brought the fluid from the reservoir (200 ml/min), through the BTM arterial-measuring head, to the dialyzer. Fluid then passed through the BTM venous-head back to the water bath. An external blood pump (recirculation pump) diverted part of the returning fluid from the venous line to the arterial line, simulating an access recirculation. The pump speed was set at 50 ml/min (fig. 1). Based on manufacturer’s recommendation, part of the modification of the 2008H machine for CRRT involved physical removal of the Diasafe on-line filter normally used to provide ultrapure dialysate. The first step of the study was to re-calibrate the modified dialysis machine for CRRT operation especially the temperature and conductivity sensors. A test was performed to verify the temperature behavior of the dialysis machine at three different dialysate flow rates (Qd = 100, 200 and 300 ml/min) in order to determine the optimal dialysate flow rate at which the studies should be performed. During this test the user-set temperature (TSet) of the dialysis machine was set to 37.0 ° C and kept constant while the dialysate flow rate was changed from 100 to 200 and 300 ml/min respectively. The dialysate temperature was additionally monitored with a calibrated external temperature meter. The second step was the analysis of an 8-hour simulated CRRT session in which the BTM was run in a temperature control mode where recirculation tests were attempted. The third part of the study included the analysis of the thermal energy balance during 8-hour simulated CRRT sessions with different scheduled rates of cooling (from +50 to –100 kJ/h). Furthermore, we wanted to explore the impact of the environmental temperature on the thermal balance by conducting some selected experiments at two different room temperatures (22.5 B 2.5 vs. 24.5 B 2.5 ° C) but similar rates of cooling. Finally we also investi-

Rahmati et al.

gated the effect of thermal losses through the blood lines by comparing sessions of CRRT with and without insulated blood lines (fig. 2).

Results

Figure 3 represents the dialysate temperature behavior at three different dialysate flow rates in which the temperature was set at 37.0 ° C. At Qd = 300 ml/min, the actual dialysate temperature was 0.6°C below the set point (TSet = 37.0 ° C). A small degree of fluctuation (¢T = B 0.1 ° C) was also observed. At Qd = 200 ml/min the difference of the actual dialysate temperature from the set point was smaller (0.3 ° C) but the temperature fluctuation was similar (¢T = B 0.15 ° C). At Qd = 100 ml/min, the actual dialysate temperature was only 0.1 ° C from the set point. However, the temperature fluctuation was greater than at the other dialysate flow rates (¢T = B 0.25 ° C). The optimal dialysate flow for the subsequent tests was therefore considered to be 200 ml/min. Calibration results indicated that the dialysate flow rate of 200 ml/min would be the best compromise between the lowest degree of temperature fluctuation and achieving the closest actual dialysate temperature to the set temperature. The next step was to find out whether the control programs of the BTM could work at Qd = 200 ml/ min. The attempt to run a dialysis session in temperature control mode, failed because of the impossibility to per-

Fig. 2. The experimental setup including the thermostated bath, the machine and the insulated blood lines.

Fig. 3. Temperature behavior of the dialysis

machine at three different dialysis flow rates.

Blood Temperature Monitor in CRRT

Blood Purif 2001;19:245–250

247

Fig. 4. Interruption of positive temperature

bolus during BTM recirculation measurement. Ufr = Ultrafiltration rate was set to 0 ml/min; Qbl = blood flow was 200 ml/min; Qdia = dialysate flow was set at 200 ml/min; Cd-dia = dialysate conductivity was 14.2 mS/cm; T-dia = dialysate temperature was set at 37 ° C. Values displayed on the right side of the panel correspond to the actual value at the level of the dotted vertical line.

Table 1. Summary of the results from the energy balance studies

Test, kJ/h

+50 +50 0 0 –30 –30 –100 –100

Insulation

No Yes No Yes No Yes No Yes

Environment temperature °C 27.4 26.9 26.5 26.5 22.0 25.0 26.8 25.5

Energy balance scheduled kJ

obtained kJ

+400 +400 0 0 –240 –240 –800 –800

+80.51 +65.50 –337.4 +2.1 –484.6 –240.0 –795.7 –880.0

form successful recirculation measurements. Figure 4 shows the captured data of this test. In detail all manually set recirculation measurements with different types of temperature bolus were interrupted, due to the slow dialysate temperature change in CRRT mode. Therefore, BTM could not be used in temperature control mode, since this mode requires successful access recirculation measurements. A subsequent set of experiments were carried out in energy control mode to investigate the accuracy of thermal energy balance measured by BTM at a dialysate flow rate of 200 ml/min. The tests were done for different rates of thermal energy flux (+50 to –100 kJ/h). The water bath was kept constant at 37.0 B 0.2 ° C. The set temperature of the dialysis machine was adjusted to 37.0 ° C for all

248

Blood Purif 2001;19:245–250

tests. Tests were started when the actual dialysate temperature came to within 0.3 ° C of the set temperature (37.0 B 0.3 ° C). The environment temperature of the laboratory was measured in each experiment. Table 1, displays the results obtained in the different experimental conditions including the effect of insulation of the blood lines. It is evident from the reported data that in the absence of blood line insulation, the thermal energy balance was always more negative than that expected from the scheduled program. This is especially evident when neutral or positive thermal balance was programmed (fig. 5). The interpretation of these results suggests that a possible heat loss may occur in the hemodialyzer or in the blood lines. The study performed with the insulated blood lines showed that in fact this is the case. With the insulation, the results obtained were always very close to the scheduled values. The environmental temperature as displayed in figure 6 may also play a slight effect. A lower temperature of the environment may in fact cause an exceedingly negative thermal balance.

Discussion

Our study showed that some of the features of the BTM are not available in CRRT mode such as the BTM access recirculation measurement and temperature control mode. The bolus detection time for the BTM is limited to 4–5 min. In our tests, respective error messages generated by the BTM at the measurement termination time indicated that the detection time had been exceeded.

Rahmati et al.

Fig. 5. An 8-hour treatment test in CRRT mode using BTM in energy control program without insulation of the blood lines. In spite of the scheduled negative thermal balance of –33.7 kJ/h (expected energy loss = 269.6 kJ/8 h) the final thermal loss was 484.6 kJ.

Fig. 6. Effect of room temperature on the final thermal balance.

Blood Temperature Monitor in CRRT

Blood Purif 2001;19:245–250

249

The dialysate temperature change during the recirculation process in CRRT mode is slower than the temperature change during conventional hemodialysis. The dialysis machine at low dialysate flow rates, typical for CRRT, cannot achieve the desired temperature change necessary for the BTM within this amount of time. This results in interruption of the recirculation measurements. The temperature control program of the BTM requires successful recirculation measurements to correct the dialysate temperature changes. Since no recirculation measurements can be done, the temperature control of the BTM cannot operate in CRRT mode. Our study also showed that the BTM energy control program functions more accurately when it is set to higher rates of cooling (· = –100 kJ/h) during CRRT. The range of accuracy can be extended if insulation of the blood lines is provided. Analysis of the results shows that at the blood and dialysate flow rates typical for CRRT, the temperature loss in the extracorporeal circuit increases. More heat is required to compensate the heat loss and to keep the venous temperature at the desired level. This also explains why the BTM controls the thermal energy balance more accurately when higher rates of cooling are programmed. The extra-

corporeal circuit can also contribute to the problem with lower blood and dialysate flows as we could demonstrate that massive heat loss could occur through the blood lines. The insulation decreases the temperature loss in extracorporeal circuit significantly. This improves the accuracy of the BTM measurements at lower rate of cooling.

Conclusions

The blood temperature monitor is an extremely valuable tool for hemodialysis treatment. Although BTM was not designed to be used at low flow rates typical for CRRT, it can easily be integrated for this mode. Important data can be collected via BTM during lengthy CRRT treatments. The energy control mode of the BTM measures the total energy balance with high accuracy provided that specific conditions of application are fulfilled. The BTM may therefore become a powerful tool for patients undergoing continuous venovenous hemodialysis in order to control their thermal energy balance and possibly to establish the most favorable conditions for improved hemodynamic stability.

References 1 Ronco C, Burchardi H: Management of acute renal failure in the critically ill patient; in Pinsky MR, Dhaunaut JFA (eds): Pathophysiobiologic Foundations of Critical Care. Baltimore, Williams & Wilkins, 1993, pp 630–676. 2 Bellomo R, Ronco C: Acute renal failure in the intensive care unit: Which treatment is the best? In Update in Intensive Care and Emergency Medicine. Heidelberg, Springer, 1995, vol 20, pp 385–407. 3 Ronco C, Bellomo R: Adequacy of renal replacement therapy; in Update in Intensive Care and Emergency Medicine. Heidelberg, Springer, 1995, vol 20, pp 364–385. 4 Ronco C: Continuous renal replacement therapies in the treatment of acute renal failure in intensive care patients. 2. Clinical indications and prescription. Nephrol Dial Transplant 1994;9(suppl 4):201–209. 5 Ronco C, Bellomo R: Complications with continuous renal replacement therapies. Am J Kidney Dis 1996;28(suppl 3):100–104.

250

6 Kramer P, Wigger W, Rieger J, Matthaei D, Scheler F: Arteriovenous haemofiltration: A new and simple method for treatment of overhydrated patients resistant to diuretics. Klin Wochenschr 1977;55:1121–1122. 7 Lauer A, Saccaggi A, Ronco C, Belledonne M, Glabman S, Bosch JP: Continuous arteriovenous hemofiltration in the critically ill patient. Ann Intern Med 1983;99:455–460. 8 Kaplan A, Longnecker RE, Folkert VW: Continuous arteriovenous hemofiltration. Ann Intern Med 1984;100:358–364. 9 Ronco C: Continuous renal replacement therapies in the treatment of acute renal failure in intensive care patients. 1. Theoretical aspects and techniques. Nephrol Dial Transplant 1994; 9(suppl 4):191–200. 10 Ronco C, Bellomo R: Kontinuierliche Nierenersatzverfahren: Wirkungsprinzipien. Dialyse J 1995;50:19–24. 11 Ronco C: Continuous renal replacement therapies for the treatment of acute renal failure in intensive care patients. Clin Nephrol 1994;4: 187–198.

Blood Purif 2001;19:245–250

12 Ronco C, Bellomo R: Critical Care Nephrology. Dordrecht, Kluwer Academic, 1998. 13 De Vries JP, Van der Meer BJ, Vonk Noordegraaf CA, Beukhof JR, Janssen MJ, Van der Meulen J, De Vries PM: Combined measurement of tissue fluid, blood volume and hemodynamics in hemodialysis. Int J Artif Organs 1995;18:705–711. 14 Ronco C, Bellomo R, Wratten ML, Tetta C: Today’s technology for continuous renal replacement therapies. Clin Intens Care 1996;7: 198–205. 15 Ronco C, Brendolan A, Bellomo R: Continuous versus intermittent renal replacement therapy in the treatment of acute renal failure. Nephrol Dial Transplant 1998;13(suppl 6):79– 85. 16 Silvester W: Outcome studies of continuous renal replacement therapy in the intensive care unit. Kidney Int 1998;53(suppl 66):S138– S141.

Rahmati et al.

Blood Purif 2001;19:251–254

Two Years’ Experience with Dialock® and CLSTM (A New Antimicrobial Lock Solution) Klaus Sodemann a Hans-D. Polaschegg b Beate Feldmer a a Dialysis

Center, Lahr-Ettenheim, Germany and b Koestenberg, Austria

Introduction

Materials

Central venous catheters are reluctantly used as blood access for hemodialysis because of safety concerns and frequent complications, e.g., sepsis, thrombosis and vessel stenosis. Nevertheless, 20% or more of all patients are relying on atrial catheters for chronic dialysis because of lack of any other blood access. Two recent developments tackle the most critical problems of the catheter: safety and sepsis. An implantable access port ameliorates or eliminates safety problems as air embolism and blood loss to the environment and the antimicrobial locking solution eliminates sepsis related to biofilms in catheters. This paper reports about results of a German Dialock®/CLS study, the first study using an antimicrobial catheter locking solution in a large number of patients. Dialock and CLS have been developed by Biolink Corp., USA. The study was approved in May 1998 by the ethics committee of the ‘Landesärztekammer Baden-Württemberg’. Eligible patients were at least 18 years of age requiring hemodialysis. Inclusion criteria were vessel exhaustion and/or congestive heart failure. Pregnancy and inability to provide informed consent were exclusion criteria. The purpose of this pilot trial was to demonstrate the safety and efficacy of the Dialock in combination with a new heparin-free antimicrobial lock solution.

ABC

© 2001 S. Karger AG, Basel 0253–5068/01/0192–0251$17.50/0

Fax + 41 61 306 12 34 E-Mail [email protected] www.karger.com

Accessible online at: www.karger.com/journals/bpu

The original Dialock system has been described elsewhere [1]. Dialock is a subcutaneous device consisting of a titanium housing with two passages with integrated valves connected to two silicone catheters. The system is implanted subcutaneously below the clavicle. The tips of the catheters are placed in the right atrium. The port is accessed percutaneously with needle-cannulas. Half of the patients originally received the first version of Dialock. The current version of Dialock has a modified body allowing for easier needle access and improved wire reinforced catheters without side holes. This version was implanted beginning in the middle of 1999. Between dialysis sessions the housing and catheters are filled with the catheter locking solution (CLS). CLS is a mixture of taurolidine – an antimicrobial substance –, citric acid and citrate for anticoagulation. Taurolidine is a unique nontoxic substance with antimicrobial [2], antimycotic [3], antiadherent [4] and antiendotoxic [5] properties. It is effective against multiresistant bacteria [6] and also inhibits staphylococcus coagulase [7], a clotting activator not inhibited by heparin, hirudin or antithrombin. Systemically, the infusion of even large volumes of taurolidine solution showed no measurable effect in a large randomized study [8]. Taurolidine has no systemic activi-

Dr. med. Klaus Sodemann Dialyse-Centrum Lahr, Schillerstrasse 6–8 D–77933 Lahr/Schw. (Germany) Tel. +49 7822 8977 113, Fax +49 7821 9166 66 E-Mail [email protected]

ty on the coagulation system or on laboratory samples. For this reason citrate was added to CLS. The pH of CLS is titrated with citric acid to increase the biocidal activity of taurolidine [9]. Risk Analysis No adverse effects related to the CLS have been surfaced so far. Taurolidine dissociates into taurultam and this decays with a half-time of t½ (·) = 0.14 h and t½ (ß) = 2.22 h respectively. The final products are taurine (an amino acid), water and carbon dioxide [10]. The CLS is injected at the end of the treatment into the catheter and aspirated again before the next treatment. During this procedure it is possible that a small part of the CLS bolus leaves the catheter. Furthermore, taking user error into account, larger amounts of the CLS may be injected. As for any device, risk analysis must show that the combination of likelihood and severity of such an event is below an acceptable threshold. The likelihood of double injection into the same lumen must be assumed as probable. The pre-filled syringe contains 6 ml. Under worst case fault condition the full syringe is injected into one passage. This means that approximately 3.5 ml of the CLS are injected as bolus. Effect of Taurolidine. Blenkharn [11] mentions studies where up to 20 g of taurolidine per day have been applied. 3.5 ml of CLS contains only 0.2% of this amount. As total dose this amount is clearly negligible. Assuming a short term distribution in the plasma volume of 3 liters only, the concentration is F16 mg/l compared to 50 mg/l plasma concentration mentioned as typical by Blenkharn [11]. In the same paper, bolus application of 1 g of taurolidine is mentioned. It is therefore concluded that taurolidine in CLS poses no risk in case of unintended bolus injection. Effect of Citrate. Again, short-term distribution in plasma space is assumed. Citrate complexes ionized calcium and the related hazard is hypocalcemia. Two molecules of citrate bind three molecules of calcium. 3.5 ml of CLS contain 0.47 mmol of citrate able to bind F0.7 mmol of ionized calcium. For comparison: 250 ml of ACD(A) blood for transfusion purposes contain F5.6 mmol of citrate, which is a tenfold higher amount but cannot be infused within 2 s. The normal value of Ca2+ is 1.26 mmol/l under physiological conditions and the total amount in 3 liters of plasma is 3.78 mmol. This amount is reduced by citrate to 3.08 mmol and the resulting plasma concentration is F1.03 mmol/l that can be regarded as safe under the worst-case assumptions made for the distribution of ci-

252

Blood Purif 2001;19:251–254

trate. The citrate will rapidly distribute in the extracellular space and will be metabolized within a few minutes by the liver. Biolink has confirmed the safety of CLS with large animal acute studies. However, the same calculation with 3.5 ml done for a 1,600 mmol/l citrate concentrate (467 g/l or 46.7% solution, triCitrasol®) would result in a complete depletion of Ca2+. This complete depletion of ionized calcium even for a short period is life-threatening [12]. Patient Demographics The study began in Ettenheim and Lahr where most of the study patients are dialyzed. After the initial success, patients from all parts of Germany were accepted. Between June 1998 and September 2000, 70 patients (29 male, 41 female) were enrolled into the study. The mean age at implantation was 63 years (range 30–88). The distribution is shown an figure 1. Thirty-one patients were diabetics and 13 had implanted pacemakers. Other comorbidities were short-bowel syndrome with colostomy bag, previous cardiac surgery, existing infection at the time of implantation, hypercoagulability, liver cirrhosis with low platelet count, anomaly of the vena cava and venous vessel stenosis. Twenty-four patients died from causes not related to Dialock. Six patients were transferred to other treatments during 76 patient-years. Catheter Placement The right internal jugular vein was the preferred vessel for catheter placement. In more than 50% of cases, however, another vein had to be used and in 12 cases the femoral vein was the only remaining choice. Table 1 shows the veins used for the first catheter implantation. Six patients required a catheter replacement and 1 patient required two catheter replacements because of flow problems. The femoral route was used as a last resort. Intended placement of catheter tips was the right atrium. This was achieved in each patient at least on half of the time in all upper torso placements. For femoral placement, the catheters only reached the kidney level in the IVC. Now using longer catheters, the atrial position is achieved. Access Procedure Accessing the Dialock has previously been described [1]. The puncture site is first cleaned with alcohol and then disinfected with an alcohol pad for 3 min. The needles are placed and the trocars are withdrawn. The clamps are closed and the needle caps removed. The locking solution is aspirated and the remaining procedure is identical to the common access technique. Neither patients nor

Sodemann/Polaschegg/Feldmer

Fig. 1. Distribution of patients at implantation.

medical staff wear masks during the procedure. The use of gloves is optional. The whole procedure from the start until the beginning of dialysis takes less then 10 min. At the end of the treatment, the port and the catheters are rinsed with saline and each side is filled with 3 ml of CLS from a pre-filled syringe. Treatment of Sepsis Fever and shivering were indicators for sepsis that was confirmed by elevated white blood cell counts, increase of C-reactive protein (CRP) and positive blood culture. Pocket infection was identified by redness, pain or swelling around the port. Treatment was initiated immediately after taking blood samples by systemic antibiotic treatment through the Dialock (vancomycin, cefotiam). For pocket infections this treatment alone was insufficient and seven devices were explanted between May and November 1999. Subsequently, losses of implants were avoided by injection of 160 mg gentamicin into the pocket in addition to the systemic antibiotic treatment. Blood samples following this application revealed a slow release of gentamicin into the blood stream. The highest concentration observed was 4.2 Ìg/ml.

Fig. 2. Distribution of infection episodes among patients (13 months average patient experience using Dialock/CLS).

Table 1. Veins used for the first catheter implantation

Jugular internal right Jugular internal left Jugular external left Subclavian right Subclavian left Femoral right Femoral left Total

33 12 2 5 6 10 2 70

were transferred to a spread sheet program for further evaluation. Infection events occurring during the first 30 days after implantation (surgery-related) or following a previous infection within 30 days were not counted as independent infection events.

Results

Data Evaluation Death and/or explantation of the Dialock as of October 2, 2000 were the endpoint for data evaluation. The number of catheter days for each patient on the study was calculated from the difference between the endpoint and the implantation date. Data from treatment protocols were put into a relational database and all adverse events

Forty-two patients (60%) had no infection and this number increases to 45 (64%) when infections occurring within 30 days after implantation are omitted. Excluding these early events, 25 patients experienced a total of 30 infections (fig. 2). The majority (22 events in 20 patients) were pocket infections. The first 7 of these pocket infections caused the loss of the Dialock. After initiation of local treatment with gentamicin no further devices were lost due to pocket infection. No infection events were recorded within the last 3 months although the expecta-

Dialock

Blood Purif 2001;19:251–254

253

tion rate calculated on the previous occurrences would be approximately 4 pocket infections. This may be related to increased nursing care to avoid infection. The study results in a normalized pocket infection rate of 0.8 per 1,000 catheter days and a normalized rate of bloodstream infections of 0.29 per 1,000 days.

Discussion

Safety concerns and high complication rates have caused catheters to be regarded as temporary or last resort blood access. Potentially fatal risks related to central venous catheters include air embolism [13], severe blood loss [14] and electric shock [15]. These specific risks have been substantially eliminated by the inherent design and implantation of Dialock. The complications of infection have been quite low in this study. Catheter-related infection or bacteremia was less than 0.3 episodes per 1,000 catheter days of use, which is considerably below recent reviews of catheter

studies. Schwab and Beathard [16] reported rates of 4 episodes per 1,000 days and Canaud [17] reported 3.5 episodes per 1,000 days with permanent catheters. Pocket infection in this study were 0.8 episodes per 1,000 days. This rate is lower than analogous catheter exit site infections reported as 1.8 episodes per 1,000 days when the site is protected with mupirocin in the treatment group and 14.3 in the control group [18]. We have observed that the incidence of pocket infection has been reduced in the last 10 months of the study following greater attention to accessing technique. Currently, we strive to treat all infections to avoid removal of Dialock and now have a 77% eradication of infection over the full study. In addition, we have instituted an infection prevention strategy that has eliminated pocket infections in all of the currently enrolled patients (40) over the last 4 months. Our experience with the Dialock/CLS has been gratifying. We believe the Dialock system is ready for widespread evaluation by clinicians to further elucidate its role in hemodialysis access.

References 1 Levin NW, Yang PM, Hatch DA, Dubrow AJ, Caraiani NS, Ing TS, Gandhi VC, Alto A, Davila SM, Prosl FR, Polaschegg HD, Megerman J: New access device for hemodialysis. ASAIO J 1998;44:M529–M531. 2 Blenkharn JI: In-vitro antibacterial activity of noxythiolin and taurolidine. J Pharm Pharmacol 1990;42:589–590. 3 Nösner K, Focht J: In-vitro-Wirksamkeit von Taurolidin und 9 Antibiotika gegen klinische Isolate aus chirurgischem Einsendegut sowie gegen Pilze. Chir Gastroenterol 1994;10(suppl 2):80–89. 4 Gorman SP, McCafferty DF, Woolfson AD, Jones DS: A comparative study of the microbial antiadherence capacities of three antimicrobial agents. J Clin Pharm Ther 1987;12: 393–399. 5 Pfirrmann RW, Leslie GB: The anti-endotoxin activity of taurolin in experimental animals. J Appl Bacteriol 1979;46:97–102. 6 Traub WH, Leonhard B, Bauer D: Taurolidine: In vitro activity against multiple-antibioticresistant, nosocomially significant clinical isolates of Staphylococcus aureus, Enterococcus faecium, and diverse Enterobacteriaceae. Chemotherapy 1993;39:322–330.

254

7 Reinmüller J, Mutschler W, Meyer H: Hemmung der Staphylokokken-Koagulase durch Taurolin. Hämostaseologie 1999;19:94–97. 8 Willatts SM, Radford S, Leitermann M: Effect of the antiendotoxic agent, taurolidine, in the treatment of sepsis syndrome: A placebo-controlled, double-blind trial. Crit Care Med 1995; 23:1033–1039. 9 Brodhage H, Pfirrmann RW: Taurolin – Bakteriologie in vitro; in Brückner WL, Pfirrmann RW (eds): Taurolin. Munich, Urban & Schwarzenberg,1985, pp 38–47. 10 Blenkharn JI: Antibakterielle und verwandte Eigenschaften von Taurolin – ein Überblick. Chir Gastroenterol 1991(suppl):143–151. 11 Blenkharn JI: Prevention of septic complications associated with TPN. J Parenter Enteral Nutr 1986;10:436–437. 12 Bunker JP, Bendixen HH, Murphy AJ: Hemodynamic effects of intravenously administrated sodium citrate. N Engl J Med 1962;266:372– 267. 13 Orebaugh SL: Venous air embolism: Clinical and experimental considerations. Crit Care Med 1992;20:1169–1177. 14 Lau G: Iatrogenically-related, fatal haemorrhage occurring in end-stage renal failure: A series of three cases. Forensic Sci Int 1995;73: 117–124.

Blood Purif 2001;19:251–254

15 Jonsson P, Stegmayr BG: Current leakage in haemodialysis machines varies. Int J Artif Organs 1999;22:425. 16 Schwab SJ, Beathard G: The hemodialysis catheter conundrum: Hate living with them, but can’t live without them. Kidney Int 1999; 56:1–17. 17 Canaud B: Haemodialysis catheter-related infection: Time for action. Nephrol Dial Transplant 1999;14:2288–2290. 18 Sesso R, Barbosa D, Leme IL, Sader H, Canziani ME, Manfredi S, Draibe S, Pignatari AC: Staphylococcus aureus prophylaxis in hemodialysis patients using central venous catheter: Effect of mupirocin ointment. J Am Soc Nephrol 1998;9:1085–1092.

Sodemann/Polaschegg/Feldmer

Blood Purif 2001;19:255–259

Sorbent Augmented Dialysis: Minor Addition or Major Advance in Therapy? James F. Winchester a Claudio Ronco b, d James A. Brady a Ellen Golds a Jonathan Clemmer a Larry D. Cowgill c Thomas E. Muller a Nathan W. Levin b a RenalTech

International, b Renal Research Institute, New York, N.Y., and c Veterinary Medicine Teaching Hospital, University of California Davis, Sacramento, Calif., USA; d Department of Nephrology, St. Bortolo Hospital, Vicenza, Italy

Introduction

The Changing Understanding of Toxins in Uremia (Small Molecular Weight Proteins)

Modern dialysis, used in approximately 1 million people throughout the world, carries a mortality (unadjusted for co-morbidity) in the first year of anywhere between 14 and 23% (EDTA/USRDS). Clearly, mortality is governed by co-morbid conditions, such as diabetes and heart disease, but also by the quantity of dialysis delivered (Kt/ Vurea). Even with high efficiency dialysis and consequently high Kt/Vurea, patients continue to feel unwell, have demonstrable suppression of immunity, and increased risk of the following; infection, atherosclerosis, circulating abnormalities of proteins (glycosylation) and lipids, as well as endocrine disturbances in parathyroid, pituitaryhypothalamic, and gonadal function. In addition, specific diseases associated with dialysis develop with the length of time a patient has been treated (e.g. dialysis-related amyloidosis, DRA), with hemodialysis or continuous ambulatory peritoneal dialysis (CAPD). Without the use of exogenous recombinant human erythropoietin, patients would remain anemic, and without the use of exogenous vitamin D analogues bone disease and hyperparathyroidism would be much more prevalent.

ABC

© 2001 S. Karger AG, Basel 0253–5068/01/0192–0255$17.50/0

Fax + 41 61 306 12 34 E-Mail [email protected] www.karger.com

Accessible online at: www.karger.com/journals/bpu

High efficiency dialysis, and on-line hemodialfiltration (with or without adsorption of substances in the filtrate) does correct certain of the above abnormalities, but it has become apparent in the last decade that certain known toxins (e.g. ß2-microglobulin (ß2M)) can cause serious medical problems such as DRA. Recently discovered (e.g. reactive oxygen species (ROS) [1], advanced lipoxygenation end-products (ALEs), C-reactive protein (CRP)) toxins accumulate in the dialyzed uremic patient and eventually produce serious consequences. Modern dialysis membranes and CAPD are inefficient or unable to remove these toxins by virtue of their molecular size, which impairs transfer across dialysis membranes. Some molecules (ß2M) can adsorb (or can be removed more efficiently with convection), to certain dialysis membranes (polysulfone or polyacrylonitrile), but for the most part the ‘new’ uremic toxins demand a different approach for removal. Hence the rekindled interest in sorbent technology applied to treatment of the dialyzed uremic patient. As mentioned above, the classical uremic toxins have been supplemented by growing knowledge of new bio-

James F. Winchester RenalTech International 320 East 65th St, #116 New York, NY 10021 (USA) Tel. +1 212 717 8644, Fax +1 212 717 8643, E-Mail [email protected]

Table 1. Classical [23] and ‘new’ [24] uremic toxins

Classical

‘New’

Urea Cyanate, ammonium Carbonate ‘Organic base’ Creatinine Creatine Guanidinoacetic acid Guanidinosuccinic acid Methylguanidine Guanidinoproprionic acid Glycocyamine Guanidinobutyric acid Taurocyanamine N-Acetylarginine Uric acid Amino acids Polypeptides Middle molecules Polyamines Indicans (indoles) Hippuric acid Phenols and conjugates Organic acids Aliphatic amines Dimethylamine Trimethylamine Pseudouridine Acetoin 2,3-Butyleneglycol Sulfates ß-Hydroxybutyrate 4-Amino-5-imidazole Carboxamide N-methyl-2-pyridoxine Carboxamide 3·-Androst-5-ene-17-one Glucuronic acid Myoinositol Oxalic acid Water Hydrogen ion Phosphate Hormones Growth hormone Renin Glucagon Natriuretic peptide Parathyroid hormone Minerals Calcium Magnesium Arsenic

256

Cystatin C Clara cell protein Leptin AGEs ALEs ROS ß2-Microglobulin GIP-I GIP-II ß-Endorphin ß-Lipotropin Adrenomedullin Carboxymethylpropyl furanpropionic acid

Blood Purif 2001;19:255–259

chemical species either unknown a decade ago, or seen in a new light (table 1). Glycosylation of proteins (long known for glycosylated hemoglobin as a measure of control of diabetes) brings about profound changes in nearly all tissues of the body (‘browning’), a process that seems to be accelerated in dialysis patients [2]. Glycosylation of lipids occurs, proinflammatory cytokines circulate in high concentration, ROS abound, complement factors (factor D) impair leukocyte physiology, and many other biochemical substances, including markers of inflammation (CRP), granulocyte inhibitory proteins (GIP), etc., have been detected in chronic renal failure and dialyzed patients. Most of the ‘new toxins’ are in the middle molecular weight range (8–15 kD), not amenable to dialytic removal.

Brief History of Sorbents in Uremia

Oral Sorbents Orally administered sorbents such as charcoal or oxystarch which bind nitrogen compounds have been studied in uremia. Metabolism of nitrogen compounds by ingestion of enzymes has also been studied with inconclusive results. Some strains of bacteria which synthesize enzymes that recycle urea and other nitrogen compounds have also been studied, in an attempt to control uremia. To date the use of oral sorbents remains questionable. Peritoneal Dialysis Sorbents have been investigated for their role in enhancing solute removal,but since abandoned, due to lack of clear-cut efficacy [3, 4]. More recent interest in increasing the efficiency of automated peritoneal dialysis has centered around sorbent regeneration of peritoneal dialysate – no clinical data is available [5]. Hemodialysis Membranes with a layer of adsorptive compounds attached (carbon) were investigated in the past – the low efficiency of these membranes (cuprophane) has delayed further clinical investigation [6]. Convective removal of solutes is increased by including sorbents (charcoal and zeolites) in the dialysate path. This offers several advantages, the most important of which are: no need for sterilization of the sorbent, no blood contact and therefore no risk of hemo-incompatibility [7]. However, for larger molecules of current interest their removal is limited by the pore structure of the membrane. In the latter regard, using on-line hemodiafil-

Winchester/Ronco/Brady/Golds/Clemmer/ Cowgill/Muller/Levin

tration with polysulfone membranes allowing passage of middle molecules in the 8–15 kD range, the ultrafiltrate can be subjected to adsorption prior to reinfusion [8, 9]. Hemoperfusion Typical sorbents used in hemoperfusion devices are the activated carbons (charcoals), ion-exchange resins or nonionic macroporous resins [10]. Clinical devices contain 70–300 g of activated charcoal coated with polymer coatings ranging in thickness from 0.05 to 0.5 microns. Maximal adsorptive capacity is achieved by inducing a high surface porosity and large surface area (approximately 1,000 m2/g). The pores are classified by the size of their radius, which principally determines the efficiency of adsorption, into micropores (!20 Å), transitional pores (20–500 Å) and macropores (radius x500 Å). In general, nonpolar solutes are better adsorbed from aqueous solution than are polar solutes. Diffusion must occur through the membrane, through the macropores, then into the micropores where the adsorption process is finalized. For uncoated carbon, the rate-limiting step is pore diffusion, while with coated carbon the rate-limiting step is diffusion through the polymer coating. Removal by activated carbon of solutes ranging in molecular mass from 60 to 21,500 Daltons (D) has been demonstrated in vitro and in vivo. As mentioned above, diffusion of solutes into the microporous structure of coated carbon depends on the polymer membrane thickness and for substances of low molecular mass (creatinine (113 D), uric acid (168 D), hippuran (363 D), vitamin B12 (1,355 D)) a thin cellulose coating reduces the adsorption only slightly. At higher molecular weights (13,500 D), however, substantial reduction of adsorption occurs with polymer coating [6]. Nevertheless, it is this capacity to adsorb molecules of the ‘middle molecular weight’ size (8–15 kD) that has stimulated interest in activated charcoal hemoperfusion in uremia. Adsorption of molecules 11,500 D is limited by the pore structure of the specific semipermeable membrane coating. Adsorption of biologically important small solutes also occurs, namely glucose, calcium and amino acids, all of which exhibit finite saturation rates for adsorption. Even 25-hydroxycholecalciferol and other hormones have been removed by charcoal hemoperfusion in vivo [11, 12].

ment activation by surface contact, with margination of leukocytes similar to that observed during hemodialysis. No appreciable changes in coagulation factors II–XII have been observed as a response to charcoal hemoperfusion in uremic patients. The side effects outlined above, although minor in nature, have stimulated the search for more biocompatible sorbents. Clinical Studies of Charcoal Hemoperfusion in Uremia In uremic patients, the classical uremic toxins creatinine, uric acid, guanidine, indoles, phenolic compounds and organic acids were removed more efficiently than with the available dialysis equipment (table 1). Most short-term studies have not reported any clinical improvement with the use of charcoal hemoperfusion in uremia, except when cardiac function, pericarditis, gastrointestinal symptoms and lethargy occurred. Middle molecule removal with charcoal hemoperfusion is equivalent to polyacrylonitrile hemodialysis, and the rebound in plasma concentrations is delayed, indicating release of middle molecules from more than one body pool [14, 15]. Several long-term studies have shown an improvement in mean nerve conduction velocity, electromyogram, pruritus, and pericarditis. Specifically, Stefoni et al. [16] and Bonomini et al. [17] have commented on a 30% added annual cost for dialysis, with combined hemoperfusion and hemodialysis. The major reason that charcoal hemoperfusion is not added routinely to dialysis relates to the spectrum of solutes adsorbed, particularly the poor performance in removing molecules 11,500 kD. Hence, the renewed interest in developing sorbents with larger pore size distribution.

New Sorbent Design for Removal of Middle Molecular Weight Proteins, Polypeptides, etc.

Side Effects of Charcoal Hemoperfusion Current charcoal hemoperfusion devices produce platelet losses of 30% or less [13]. Transient leukopenia, similar to that observed during hemodialysis, occurs during hemoperfusion in man, and may be a result of comple-

Recently, several scientific approaches have been made to increase the internal pore structure of resins. Using a proprietary polymerization process it is possible to create large adsorptive surface area polymers of divinylbenzene in resin bead form and coat the surface of theses beads with a biocompatible coating of polyvinylpyrrolidone [18]. Such a resin has a pore size distribution of 2–20Å, which is suitable for removing molecules in the weight range 8–15 kD, and restricting adsorption of larger molecules such as serum albumin. The bead size is around 600 Ìm, and the surface for adsorption is 600 m2/g of resin. Hemoperfusion devices contain 300 ml resin for

Sorbent Augmented Dialysis

Blood Purif 2001;19:255–259

257

Table 2. Putative uremic toxins removed by sorbents (molecular

weight range 60–21,500 D) Adrenocorticotrophin (Aldosterone)1 Amino acids1 ß2-Microglobulin Calcium2 25-OH-cholecalciferol1 Creatinine1 Cyclic AMP Epinephrine Folic acid2 Fibronectin Gastrin Glucagon Glucose1 (Growth hormone)1 Trace metals: As, Co, Cr, Se (Mg)1 Tumor necrosis factor (TNF) Triglycerides1 Interleukin-1 Thyroxine1, triiodothyronine1 (Urea)1 Uric acid1 1 2

Middle molecule peaks1 Myoinositol Non-protein nitrogen Norepinephrine Organic acids1 Oxalate1 Parathyroid hormone Phenols1 (Phosphate)1 Polyamino acids (Renin)1 Ribonuclease Serotonin Guanidines Indoles1 Insulin1 L-Dopamine Lysozyme Vitamin B12

Studied during hemoperfusion in uremic patients. Unpublished. ( ) = Incompletely removed.

human use (BetaSorb™, RenalTech International, New York, N.Y., USA) and 100 ml for experimentation in animals. Smaller devices are also used for in vitro and in vivo testing. In normal and uremic canines the devices produce about a 30% fall in platelet and leukocyte counts, but no changes occurred ex vivo using human volunteer blood passed over the resin [19] or in vivo in man during clinical hemodialysis [20]. This confirms the bio-compatibility of the coating, and opens the way for extensive clinical testing. In a uremic dog it was confirmed that ß2-M removal was F95% with a single combined hemodialysis/hemoperfusion session, and predialysis ß2-M was normalized after three sessions of the combined procedure. In addition, parathyroid hormone concentrations in blood were reduced over that obtained with hemodialysis alone. Similarly in 2 dialysis patients, ß2-M was reduced 69 and 79%, with no changes in platelet count nor leukocyte count over the 3-hour combined treatment period. It has also been demonstrated that in addition to ß2-M removal this resin also adsorbs the cytokines TNF-· and IL-1ß [21].

258

Blood Purif 2001;19:255–259

Another resin device has been used in the treatment of DRA (Kaneka Corp., Osaka, Japan) – the authors demonstrated a reduction in symptoms as well as radiographic amelioration of bone cysts [22]. No data has been published on the biocompatibility of this device. The BetaSorbTM device will undergo extensive clinical testing, under an Investigational Device Exemption (IDE), from the US Food and Drug Administration. The initial study will examine and quantify the amount of ß2M removed over a 12-week period.

Conclusions

New insights into the toxicity of nonclassical uremic toxins have prompted a reappraisal of the efficiency of hemodialysis in removing these toxins. It has become clear that current dialysis membranes may have reached their limit for transport of middle molecular weight substances and that sorbent devices of sufficient capacity to adsorb molecules in the range 8–15 kD may be a necessary addition to the dialysis regimen. The biocompatible polymer used in this study, the first not to reduce platelets in man during treatment, adsorbs ß2-M but also several toxins relevant to uremia. The addition of sorbents to dialysis is a major advance in improving removal of a spectrum of solutes, otherwise poorly or inefficiently mobilized by hemodialysis alone.

Winchester/Ronco/Brady/Golds/Clemmer/ Cowgill/Muller/Levin

References 1 Miyata T, Kurokawa K, Van Ypersele de Strihou C: Relevance of oxidative and cabonyl stress to long-term uremic complications. Kidney Int 2000;58(suppl 76):S120–S125. 2 Friedman EA: Advanced glycation end-products in diabetic nephropathy. Nephrol Dial Transplant 1999;14(suppl 3):1–9. 3 Lang HL, Nolph KD, McGary TJ: Enhancement of clearances by activated charcoal in an in vitro model of peritoneal dialysis. Clin Exp Dial Apheresis 1982;6:85–95. 4 Blumenkrantz MJ, Lewin AJ, Gordon A, Roberts M, Pecker EA, Coburn JW, Maxwell MH: Development of a sorbent peritoneal dialysate regeneration system – A progress report. Proc Eur Dial Transplant Assoc 1978;15:213–219. 5 Roberts M, Ash SR, Lee DB: Innovative peritoneal dialysis: Flow-thru and dialysate regeneration. ASAIO J 1999;45:372–378. 6 Randerson DH, Gurland HJ, Schmidt B, Farrell PC, Hone PW, Stokoe C, Zuber A, Blogg A, Fateh-Moghadam A, Marschner I, Kopcke W: Sorbent membrane dialysis in uremia. Contrib Nephrol. Basel, Karger, 1982, vol 29, pp 53– 64. 7 Ash SR, Barile RG, Thornhill JA, Sherman JD, Wang NH: In vivo evaluation of calciumloaded zeolites and urease for urea removal in hemodialysis. Trans Am Soc Artif Intern Organs 1980;26:111–115. 8 Tetta C, Bellomo R, Brendolan A, Piccinni P, Digito A, Dan M, Irone M, Lonnemann G, Moscato D, Buades J, La Greca G, Ronco C: Use of adsorptive mechanisms in continuous renal replacement therapies in the critically ill. Kidney Int 1999;56(suppl 72):S15–S19.

Sorbent Augmented Dialysis

9 Marinez de Francisco AL, Ghezzi PM, Brendolan A, Fiorini F, La Greca G, Ronco C, Arias M, Gervasio R, Tetta C: Hemodiafiltration with online regeneration of the ultrafiltrate. Kidney Int 2000;58(suppl 76):S66–S71. 10 Samtleben W, Gurland HJ, Lysaght MJ, Winchester, JF: Plasma exchange and hemoperfusion; in Jacobs C, Kjellstrand CM, Koch KM, Winchester JF (eds): Replacement of Renal Function by Dialysis. Dordrecht, Kluwer Academic, 1996, pp 472–500. 11 Winchester JF, Ratcliffe JG, Carlyle E, Kennedy AC: Solute, amino acid, and hormone changes with coated charcoal hemoperfusion in uremia. Kidney Int 1978;14:74–81. 12 Kokot F, Pietrek J, Seredynski M: Influence of haemoperfusion on plasma levels of hormones and ß-methyldigoxin. Proc Eur Dial Transplant Assoc 1978;15:604. 13 Winchester JF: Haemostatic changes induced by adsorbent haemoperfusion; in Kenedi RM, Courtney JM, Gaylor JDS, Gilchrist T (eds): Artificial Organs. London, MacMillan Press, 1977, pp 188–195. 14 Asaba H: Uremic middle molecules. Accumulation, renal excretion and elimination by extracorporeal treatment. Scand J Urol Nephrol Suppl 1982;67:1–8. 15 Asaba H, Bergstrom J, Furst P, Gunnarsson B, Neuhauser M, Oules R, Yahile V: Removal of endogenous middle molecules by hemoperfusion. Artif Organs 1979;3:132–136. 16 Stefoni S, Coli L, Feliciangeli G, Baldrati L, Bonomini V: Regular hemoperfusion in regular dialysis treatment. A long-term study. Int J Artif Organs 1980;3:348–354. 17 Bonomini V, Stefoni S, Feliciangeli G, Coli L, Scolari MP, Orsi C, Nanni-Costa A, Prandini R, Galanti S: Shortened treatment time by combined hemodialysis and hemoperfusion. Contrib Nephrol. Basel, Karger, 1985, vol 44, pp 57–63.

18 Davankov V, Pavlova L, Tsyurupa M, Brady J, Balsamo M, Yousha E: Polymeric adsorbent for removing toxic proteins from blood of patients with kidney failure. J Chromatogr B Biomed Sci Appl 2000;739:73–80. 19 Bosch T, Wendler T, Duhr C, Brady J, Samtleben W: Ex-vivo biocompatibility of a new ß2microglobulin adsorbent during hemoperfusion with human whole blood. J Am Soc Nephrol 2000;11:257A. 20 Ronco C, Brendolan A, Winchester JF, Golds E, Clemmer J, Polaschegg HD, Muller TE, La Greca G, Levin N: First clinical experience with an adjunctive hemoperfusion device designed specifically to remove ß2-microglobulin in hemodialysis patients. Blood Purif 2001;19: 260–263. 21 Brady JA, Potempska A, Ronco C, Yousha E, Muller T, Levin N: Ex vivo cytokine clearance with a new adsorbent material from endotoxinstimulated whole blood. J Am Soc Nephrol 2000;11:586A. 22 Homma N, Gejyo F, Hasegawa S, Teramura T, Ei I, Maruyama H, Arakawa M: Effects of a new adsorbent column for removing beta-2microglobulin from circulating blood of dialysis patients; in Maeda K, Shinzato T (eds): Dialysis-Related Amyloidosis. Contrib Nephrol. Basel, Karger, 1995, vol 112, pp 164–171. 23 Winchester JF, Schreiner GE: The continuing search for the uremic toxin(s); in Heintz R (ed): Uremia as a State of Intoxication. Oberursel, Fresenius Stiftung, 1978, vol 6, pp 5–15. 24 Dondt A, Vanholder R, Van Biesen W, Lameire N: The removal of uremic toxins. Kidney Int 2000;58(suppl 76):S47–S59.

Blood Purif 2001;19:255–259

259

Blood Purif 2001;19:260–263

First Clinical Experience with an Adjunctive Hemoperfusion Device Designed Specifically to Remove ß2-Microglobulin in Hemodialysis Claudio Ronco a, b Alessandra Brendolan b James F. Winchester d Ellen Golds d Jonathan Clemmer d Hans Dietrich Polaschegg c Thomas E. Muller d Giuseppe La Greca b Nathan W. Levin a a Renal

Research Institute, New York, N.Y., USA; b Ospedale San Bortolo, Vicenza, Italy; c Koestenberg, Austria; International, New York, N.Y., USA

d RenalTech

Introduction

ß2-Microglobulin, a middle molecular weight protein (11.8 kD), has increasingly been the focus of attention in its role as the precursor of dialysis-related amyloidosis (DRA) [1]. Current evidence favors the alteration of ß2microglobulin by glycosylation [2], and subsequent deposition of glycosylated ß2-microglobulin as amyloid fibrils in many tissues, particularly in large joints (shoulders), and in the carpal tunnel. DRA in chronic dialysis patients can produce bone cysts, severe carpal tunnel syndrome and crippling arthritis [3]. While there is poor correlation between blood concentrations of ß2-microglobulin and DRA, there is direct correlation of this complication with duration of dialysis. Moreover, the disposition of ß2microglobulin, follows a 3-pool kinetic model with distribution in plasma water, extracellular fluid and a slow equilibration compartment [4]. Dialyzers constructed with cuprophan dialysis membranes (cellulosic), long the standard therapy for chronic dialysis, do not reduce ß2-microglobulin concentrations and may in fact alter the conformational structure of ß2microglobulin, promoting amyloidosis. On the other hand, noncellulosic (e.g. polyacrylonitrile and polysul-

ABC

© 2001 S. Karger AG, Basel 0253–5068/01/0192–0260$17.50/0

Fax + 41 61 306 12 34 E-Mail [email protected] www.karger.com

Accessible online at: www.karger.com/journals/bpu

fone) dialyzer membranes may reduce ß2-microglobulin concentrations during treatment, by a combination of adsorption (the majority) to the membrane and by convective removal (the minority), the latter of which is dependent on high rates of ultrafiltration. The efficiency of the latter membranes is not considered optimal, and usually adsorption is complete early in the dialysis procedure. Moreover, dialyzer reuse significantly impairs the removal of ß2-microglobulin [5]. Hemoperfusion devices containing adsorbents have been used to enhance ‘middle molecule’, amino acid, and creatinine removal, in dialysis patients, using nonspecific agents such as activated charcoal [6]. A porous resin hemoperfusion device has also been shown to reduce ß2microglobulin concentrations during extracorporeal treatment, with improvement in the clinical manifestations of DRA [7]. In this study we present a new adsorbent device with improved biocompatibility by virtue of a polymer coating [8] , designed specifically to adsorb ß2-microglobulin. In vitro and in vivo data are reported as far as efficacy and biocompatibility performances are concerned. We report the case of two long-term hemodialysis patients who volunteered to undergo combined hemodialysis/ hemoperfusion at a single session, in order to assess the

Claudio Ronco, MD Department of Nephrology St. Bortolo Hospital I–36100 Vicenza (Italy) E-Mail [email protected]

Fig. 1. ß2-Microglobulin concentration throughout the session of

combined hemoperfusion-hemodialysis.

effects of the hemoperfusion device on ß2-microglobulin removal, as well as hemocompatibility.

Materials and Methods

The combined hemodialysis/hemoperfusion procedure was approved by the Ospedale San Bortolo Institutional Review Board for Investigation in Human Subjects. The procedure and risks were explained to both patients prior to combined treatment. The device will be subject to trial in human subjects in the USA and Italy under an Investigational Device Exemption (IDE) by the US Food and Drug Administration. The cylindrical hemoperfusion devices were constructed of polysulfone and contained 300 g hydrated polystyrene resin beads coated with polyvinylpyrollidone sealed with end-caps (BetaSorbTM, RenalTech International, New York, N.Y., USA). The devices were steam sterilized, inspected, primed prior to use with sterile normal saline containing 1,000 IU heparin, and placed in line with the dialysis circuit, upstream of the dialyzer. Combined hemodialysis/hemoperfusion was performed with a Fresenius 4008B controlled ultrafiltration dialysis machines (Fresenius AG, Bad Homburg, Germany), using an FH80 (polysulfone high flux) dialyzer (Fresenius AG). Blood flow rate was maintained at the patient’s customary values (380 and 405 ml/min, respectively) during the dialysis period, as was dialysate flow rate (500 ml/min). Pressures (mm Hg) across dialyzer and hemoperfusion were measured at timed intervals to detect potential flow disturbances within the device. Heparin anticoagulation was given as an intravenous bolus (2,000 IU) at the beginning of the procedure, continued by infusion of 1,000 IU/ h and supplemented where appropriate to maintain an

ß2-Microglobulin Removal by Hemoperfusion

Fig. 2. Platelet count during different moments of the combined hemoperfusion-hemodialysis session.

activated clotting time (ACT) 1120 s. Blood samples were drawn from the dialyzer lines before and after the hemoperfusion device, and after the dialyzer to assess the contribution of the dialyzer and hemoperfusion component to changes in platelets, leukocytes, ß2-microglobulin and albumin concentrations. Vital signs were measured at 30-min intervals. The combined procedure was carried out for 3 h and an additional blood sample was drawn 30 min after the procedure was discontinued for determination of ß2-microglobulin concentration rebound. Patient 1 was a stable 64year-old white male with adult polycystic kidney disease who had been on dialysis 15 years, with a history of a failed transplant 3 years earlier. Patient 2 was a stable African male aged 41 years who had been on dialysis for 8 years.

Results

The combined hemoperfusion/hemodialysis procedure was well tolerated in both patients, with neither exhibiting changes in vital signs attributable to the addition of the hemoperfusion device. ß2-Microglobulin concentrations were reduced substantially (79 and 69% in patients 1 and 2 respectively) during the combined hemodialysis/hemoperfusion procedure (fig. 1). Rebound in ß2microglobulin concentration of 49 and 31% respectively in patient 1 and 2 from the nadir was observed (fig. 1). Platelet and leukocyte counts remained stable, as did serum albumin (uncorrected for ultrafiltration required for routine patient management during treatment) (fig. 2, 3). Serum albumin concentrations are depicted in figure 4. The complete setup of the combined hemoperfusion-hemodialysis treatment is displayed in figure 5.

Blood Purif 2001;19:260–263

261

Fig. 3. Leucocyte count during different moments of the combined

hemoperfusion-hemodialysis session.

Fig. 4. Serum albumin concentration throughout the session of combined hemoperfusion-hemodialysis.

Discussion

Fig. 5. The combined hemoperfusion-hemodialysis treatment

mounted on a dialysis machine. The adsorbent catridge is placed in series with the hemodialyzer just before it.

Pressure changes measured by both standard transducers, as well as a pressure transducer connected to a computer through an analog interface, across the hemoperfusion and hemodialysis devices remained stable throughout the procedure (uncorrected for changes in blood volume).

262

Blood Purif 2001;19:260–263

DRA is not corrected by standard, nor high-flux dialysis. Serum ß2-microglobulin concentrations do appear to be reduced after a period of high-flux dialysis following a period of low-flux dialysis, although normal concentrations of ß2-microglobulin are never achieved in long-term dialysis patients. Retention of ß2-microglobulin occurs as renal function declines. Continuous ambulatory peritoneal dialysis patients also exhibit elevations in ß2-microglobulin concentration [9]. Transplantation is associated with a fall in ß2-microglobulin concentration when near normal renal function is achieved. Leypoldt et al. [10] have demonstrated that there is a survival advantage to reduced ß2-microglobulin concentrations in dialysis patients, prompting investigation of methods to remove ß2-microglobulin. Using a hemoperfusion column containing 350 ml of a porous resin, Homma et al. [7] have demonstrated substantial removal of ß2-microglobulin, with partial regression of DRA over 6–13 months. While our primary interest is in ß2-microglobulin reduction, the adsorptive resin used here is also capable of removing other ‘middle molecular’ weight toxins, such as TNF-·, and IL-1ß. To our knowledge, the biocompatibility of the adsorbent resin used in this device, as measured by platelet and leukocyte counts, is superior to devices used currently or abandoned. This property encourages the application of hemoperfusion for continual use in dialysis to achieve sustained reduction in ß2-microglobulin, to measure kinetics of ß2-microglobulin behavior, and to study the effects of ß2-microglobulin removal on prevention and reversal of dialysis related amyloidosis. Lornoy et al. [12] have recently shown that after 10 years of either hemodiafiltration (which removes about

Ronco/Brendolan/Winchester/Golds/ Clemmer/Polaschegg/Muller/La Greca/ Levin

340 mg ß2-microglobulin per session), or biocompatible membrane hemodialysis that ß2-microglobulin-associated bone disease is present in 25%, and carpal tunnel syndrome in 12.5%. Clinical hemodialysis using the synthetic membranes Arylane (Hospal Renal Care, Lyon, France) or Fresenius Polysulfone (Fresenius Medical Care, Bad Homburg, Germany) in addition to removing ß2-microglobulin (170 and 110 mg per session, respectively), is also associated with moderate removal of other small molecular weight proteins (10 and 7 g, respectively) [13]. The minimal adsorption of albumin in our study suggests that there is no added removal of protein as described in the above study. This has clinical benefits in reducing the contribution of protein removal to malnutrition. We have shown that a 53% reduction of ß2-microglobulin can be achieved using synthetic membranes (T-sulfone, Toray, Japan) [14]. In the long term this resulted in a

lower predialysis ß2-microglobulin concentration in hemodialysis patients treated with synthetic high-flux membranes compared to those treated with low-flux membranes. Our preliminary results with the coupled hemoperfusion-hemodialysis technique suggest that a further reduction should be expected after 6–12 weeks of this combined therapy.

Conclusions

The substantial reduction of ß2-microglobulin concentrations in two end-stage renal disease patients with a hemoperfusion device as part of a hemodialysis procedure, without changes in formed elements of blood, encourages us to pursue formal evaluation of the device as a treatment for DRA or its prevention.

References 1 Gejyo F, Yamada T, Odani S, Nakagawa Y, Arakawa M, Kunitomo T, Kataoka H, Suzuki M, Hirasawa Y, Shirahama T, Cohen AS, Schmid K: A new form of amyloid protein associated with hemodialysis was identified as ß2-microglobulin. Biochem Biophys Res Commun 1985;129:701–706. 2 Miyata T, Wada Y, Maeda K: ß2-Microglobulin modified with AGE products of the Maillard reaction in dialysis-related amyloidosis; in Maeda K, Shinzato T (eds): Dialysis-Related Amyloidosis. Contrib Nephrol. Basel, Karger, 1995, vol 112, pp 52–64. 3 Drueke TB: Beta-2-microglobulin and amyloidosis. Nephrol Dial Transplant 2000;15 (suppl 1):17–24. 4 Vincent C, Chanard J, Caudwell V, Lavaud S, Wong T, Revillard JP: Kinetics of 125I-ß2microglobulin turnover in dialyzed patients. Kidney Int 1992;42:1434–1443. 5 Leypoldt JK, Cheung AK, Deeter RB: Effect of dialyzer reuse: Dissociation between clearances of small and large solutes. Am J Kidney Dis 1998;32:295–301.

ß2-Microglobulin Removal by Hemoperfusion

6 Winchester JF, Ratcliffe JG, Carlyle E, Kennedy AC: Solute, amino acid, and hormone changes with coated charcoal hemoperfusion in uremia. Kidney Int 1978;14:74–81. 7 Homma N, Gejyo F, Hasegawa S, Teramura T, Ei I, Maruyama H, Arakawa M: Effects of a new adsorbent column for removing beta-2microglobulin from circulating blood of dialysis patients; in Maeda K, Shinzato T (eds): Dialysis-Related Amyloidosis. Contrib Nephrol. Basel, Karger, 1995, vol 112, pp 164– 171. 8 Bosch T, Wendler T, Duhr C, Brady J, Samtleben W: Ex-vivo biocompatibility of a new ß2microglobulin adsorbent during hemoperfusion with human whole blood. J Am Soc Nephrol 2000;11:257A. 9 Tan SY, Baillod R, Brown E, Farrington K, Soper C, Percy M, Clutterbuck E, Madhoo S, Pepys MB, Hawkins PN: Clinical, radiological and serum amyloid P component scintigraphic features of beta-2-microglobulin amyloidosis associated with continuous ambulatory peritoneal dialysis. Nephrol Dial Transplant 1999; 14:1467–1671.

10 Leypoldt JK, Cheung AK, Carrol CE, Stannard D, Pereira BJG, Agodoa LY, Port FK: Effect of dialysis membranes and middle molecule removal of chronic hemodialysis patient survival. Am J Kidney Dis 1999;33:349–355. 11 Brady JA, Potempska A, Ronco C, Yousha E, Muller T, Levin N: Ex vivo cytokine clearance with a new adsorbent material from endotoxinstimulated whole blood. J Am Soc Nephrol 2000;11:586A. 12 Lornoy W, Because I, Billiouw JM, Sierens L, Van Melderen P, D’Haenens P: On-line haemodiafiltration. Remarkable removal of beta2-microglobulin. Long-term clinical results. Nephrol Dial Transplant 2000;15(suppl 1):49– 54. 13 Hoenich NA, Stamp S: Clinical performance of a new high-flux synthetic membrane. Am J Kidney Dis 2000;36:345–352. 14 Ronco C, Brendolan A, Cappelli G, Ballestri M, Inguaggiato P, Fortunato L, Milan M, Pietribiato G, La Greca G: In vitro and in vivo evaluation of a new membrane for hemodialysis. Reference methodology and clinical results. 2. In vivo study. Int J Artif Organs 1999;22: 616–624.

Blood Purif 2001;19:260–263

263

Author Index Vol. 19, No. 2, 2001

Amerling, R. 245 Aslam, M.A. 168 Backus, G. 189 Besarab, A. 168 Brady, J.A. 255 Brendolan, A. 260 Burkart, J.M. 179 Clemmer, J. 255, 260 Conlon, P.J. 152 Cowgill, L.D. 255 Di Filippo, S. 195 Feldmer, B. 251 Golds, E. 255, 260 Gotch, F.A. 211 Greene, T. 238 Himmelfarb, J. 200 Kanagasundaram, N.S. 238 Kaufman, A. 245 Keen, M. 217 Kellum, J. 222 Kleophas, W. 189 La Greca, G. 260 Larive, B. 238 Levin, N.W. 137, 217, 245, 255, 260 Locatelli, F. 195 London, G.M. 139

ABC Fax + 41 61 306 12 34 E-Mail [email protected] www.karger.com

© 2001 S. Karger AG, Basel

Accessible online at: www.karger.com/journals/bpu

Macdougall, I.C. 157 McMonagle, E. 200 Manzoni, C. 195 Mehta, R.L. 222, 227 Morris, A.T. 245 Muller, T.E. 255, 260 Owen, W.F., Jr. 152 Paganini, E.P. 238 Pierratos, A. 206 Polaschegg, H.D. 251, 260 Rahmati, S. 245 Reddan, D. 152 Ronco, C. 137, 222, 245, 255, 260 Ronco, F. 245 Rosales, L. 245 Schleper, C. 245 Schulman, G. 175 Sodemann, K. 251 Spittle, M. 245 Star, R. 233 Stenvinkel, P. 143 Szczech, L.A. 152 Tattersall, J. 185 Winchester, J.F. 255, 260 Zhu, F. 217

Subject Index Vol. 19, No. 2, 2001

Access, vascular, see also Infection, bacterial central venous catheter 251–254 long-term 318 short daily hemodialysis 207 thrombosis 160, 164, 170 N-Acetylcysteine treatment 234 Acute phase reactants 143, 147 Acute renal failure Acute Dialysis Quality Initiative 222–226 adequate renal replacement 238–243 blood temperature monitoring 245–250 clinical trial design issues 233–236 renal replacement vs renal support 227–231 Advanced glycation end products 147 Advanced lipoxygenation end products 255 Advanced oxidation protein products 149 Age, patient arteriosclerosis 140 co-morbidity 154 hemoglobin, target 159 mortality 144, 176, 180 Albumin, serum 145 batch hemodialysis system 192 cardiovascular mortality 146 endothelial function 148 plasma protein oxidation 204 removal and adsorption 263 short daily hemodialysis 207 stabilization 261 Anemia 158 cardiovascular function 162, 169, 170 daily and nocturnal hemodialysis 206, 207 morbidity 157 pre-end-stage renal disease 165, 166 treatment 168, see also Epoetin therapy Arterial function, end-stage renal disease 139–141 Arthritis 260 Atherogenesis 139, 140, 143, 144, 148 Atherosclerosis 143, 144, 147 dialysis-related 255 oxidant stress 201 Automated peritoneal dialysis 175, 176 Azotemia 229 Bacteria, see Infection, bacterial Batch system, hemodialysis 189, 191–193 BetaSorb 258 Blood lines, thermal balance 246–248, 250 Blood loss 251 Blood pressure 197, 219, see also Hypertension, Hypotension, Intradialytic complications batch system 193 coronary heart disease 144

ABC Fax + 41 61 306 12 34 E-Mail [email protected] www.karger.com

© 2001 S. Karger AG, Basel

Accessible online at: www.karger.com/journals/bpu

hematocrit normalization 172 hemoperfusion combination 262 nocturnal 69, 207 residual renal function 180 short daily 206 Blood temperature 219 control 192 modeling 186 monitoring 245–250 Blood urea nitrogen 228, 230, 239 Blood viscosity, whole blood 164 Blood volume 186, 187, 219, 220 Bone disease 255, 260, 263 nocturnal hemodialysis 207 ‘Browning’ 256 C-reactive protein 149, 256 ‘browning’ 255 cardiovascular disease 148 elevation 146, 253 inflammation 147, 201 Calcium concentration, plasma 252 Canadian Multicentre Study 161–168, 172 access-related thrombosis 160 gender 159 quality of life 170, 171 Cannulation, buttonhole 206 Cardiovascular disease 169, 170, 182, 183, 197, 198, see also Atherogenesis, Atherosclerosis, Left ventricular hypertrophy calcification 140 electrolyte balancing 195 malnutrition and inflammation 143–149 Cardiovascular Reduction Early Anemia Treatment with Epoetin Beta Study 166 Carpal tunnel syndrome 193, 260, 263 Case Mix Adequacy Study 182, 183 Central venous catheter 251–254 Cholesterol, serum 144 Citrate 252 Clearance constant 214 continuous vs intermittent 212, 213 peritoneal vs renal 177 Clinical trials continuous renal replacement therapy vs intermittent hemodialysis 239 desing for acute renal failure studies 233–236 Coagulation factors 257 Co-morbidity 153, 154, 160 anemia 168 central venous catheter Dialock study 252 hemodialysis vs peritoneal dialysis 176, 180 ‘unphysiology’ 208

265

Complement 239, 256 Complications, intradialytic, see Intradialytic complications Conductivity dialysate 197, 198, 246 on-line clearance device 218 Continuous ambulatory peritoneal dialysis 182, 207, 214 ß2-microglobulin elevation 262 outcome improvement 208 total weekly clearances 213 Continuous dialysis kinetics 212, 213 Continuous flow peritoneal dialysis 208 Continuous renal replacement therapy 222–231 acute renal failure, adequate dosing 238–243 blood temperature monitoring 245–250 continuous venovenous hemodialysis 239, 250–254 vs intermittent hemodialysis 239, 242, 243 Cost batch dialysis system 193 country-by-country healthcare expenditures 155 daily hemodialysis 208 epoetin therapy 157, 163, 169 hemodialysis/hemoperfusion combination 257 hospitalization 181 ‘integrated care’ approach 179, 181 peritoneal dialysis vs hemodialysis 183 reimbursement rates 185–187 Creatinine 228 Critically ill patients, see Acute renal failure Crit-Line monitor 187 Cytokines, pro-inflammatory cardiovascular role 143, 145, 147–149 glycosylation 256 removal from blood 239 Daily hemodialysis 206–209, 211–216 Death, see Mortality Diabetes mellitus 143, 144, 255 mortality 176, 180, 219 target hemoglobin 160 vascular endothelial activation 148 Dialock system 251–254 Dialysate 246–248, 250 Dialysis 208, see also Acute renal failure, Cost, Dialysis Outcome Initiative guidelines, Dialyzer membrane, End-stage renal disease, Intradialytic complications, specific kinds of dialysis and dialysis devices complications overview 255 dose 207, 211, 218, 219 acute renal failure 238–243 stdKt/V 214, 215 efficiency 164, 215 fluid composition 191 frequency and length of treatment 211, 214 2008H Dialysis machines 246–248 Dialysis Outcome Quality Initiative Guidelines hematocrit 157 initiation of dialysis 181 quality of life 182 urea clearance 219 Dialysis-related amyloidosis 254, 260

266

Blood Purif Vol. 19, No. 2, 2001

Dialyzer membrane 119, 239, 255–257 biocompatibility 258, 260, 262, 263 bioincompatibility 146, 187, 239 Diasafe on-line filter 246 Diet, renal 209 Dry body weight 198, 217–221 Early referral 175 Economic issues, see Cost Electrolyte balance 195–198, 230 Endotoxin transfer 187 End-stage renal disease cardiovascular disease 139–141 geographical disparities in care and outcome 152– 156 Epoetin therapy 157, 160, 168 blood pressure control 172 elderly patients 159 hemoglobin fluctuations 164, 165 morbidity and mortality 162 pre-end-stage renal disease patients 166 short daily hemodialysis 206, 207 vascular access clotting 164 Erythropoetin, see Epoetin therapy European Best Practice Guidelines 157, 158 European Dialysis and Transplantation Association data 153, 154 Exercise capacity 159, 169, 171, 172 hematocrit normalization 161 Extracellular fluid 189–191, 220, 221 Factor D 239, 256 Fibrinogen 147, 148 Fluid gain 219 overload 229 removal 239 Food intake 218, see also Nutrition Gender, patient mortality 176, 180 target hemoglobin 159, 169 Genius batch machine 192, 193 Glomerular filtration rate, acute renal failure 234, 235 Glycosylation 255, 256 Granulocyte inhibitory proteins 255 Health Care Financing Administration data 153 Hematocrit normalization 157–166, 168–172 HEMO trial 240 Hemodiafiltration 187, 262, 263 acute renal failure 231 on-line 256, 257 sodium modeling 196 Hemodialysis, see also Anemia advances 189 cardiovascular disturbances 146, 182, 198 central venous catheter 251–254 dialysis fluid 190 early initiation 175–178

Subject Index

frequency daily 206–209, 211–216 nocturnal 206–209, 215 three times weekly 207, 208, 213 hemoglobin concentration 160, 164 normalization 169 hemoperfusion combination 257, 260–263 home 177, 178, 186 interdialytic weight gain and blood pressure 218 malnutrition 145 mortality 146, 176, 177 and morbidity by country 143–149 batch dialysis system 192, 193 oxidative stress 149 plasma protein oxidation 204 posttransplantation concerns 82 quality of life 182 residual renal function, preserving 176 simplification 189–193 skipping 155 sorbents 256, 257 Hemodynamic stability 207, 231, 246 sepsis syndrome 231 thermal balance 245 Hemoglobin cerebrovascular function 171 concentration 158, 160, 164 European Best Practice Guidelines 157, 158 normal 169 normalization 165, 166, 169 target 160 Hemoperfusion 257, 260–263 Heparin 164, 242, 261 High efficiency dialysis 255 High-flux dialysis 219, 262 Home dialysis 177, 178, 186 ‘Homeostasis’ 190 Hospitalization rates 147 anemia 162 daily hemodialysis 207, 208 hemoglobin-related in end-stage renal disease 169 Hyaluronan, serum 147 Hydrogen peroxide 201 Hyperparathyroidism 255 Hypertension 195, 197, 219 decreased arterial distensibility 141 occlusive arterial lesions 140 Hyperthermia 245 Hypoalbuminemia 145 Hypocalcemia 252 Hypochlorous (HOCl) acid 200–203 Hyponatremia 195 Hypotension 182, 187 acute renal failure 239 intermittent hemodialysis for acute renal failure 239 intradialytic 192, 197, 217 multiple organ dysfunction 229 reducing 219 Hypoxia, chronic 160

ICAM-1 148 Immune suppression, dialysis-related 255 Implantable artificial kidney 186 Incremental dialysis 176, 177, 181 Infection, bacterial 182, 251 acute dialysis method choice 231 C-reactive protein elevation 147 dialysis-related 255 periodontal 147 pocket 253, 254 prophylaxis 181, 191, 192 sepsis 229, 253 Inflammation, chronic 143–149, 187 continuous renal replacement therapy 239 phagocyte activation 201, 202 Insulin resistance 148, 149 ‘Integrated care’ approach 179, 180 Intensive care unit dialysis 227–231 Intermittent dialysis kinetics 212, 213 Intermittent hemodialysis 185, 186, 229, 239 International Dialysis Outcomes and Practices Study 155, 156 Interstitial water 221 Intradialytic complications 197, 198, 217 Iron supplementation, intravenous 163, 164, 207

Malnutrition 143–149, 263 Membrane, dialyzer, see Dialyzer membrane Mental health, patient 182 ß2-Microglobulin 122–125, 255, 258 dialysis treatment 262 Microinflammation 193 Middle molecule removal 256–258 daily and nocturnal hemodialysis 207 ß2-microglobulin 260–263 Moens-Korteweg equation 141 Morbidity, see also Co-morbidity, Morbidity/mortality, specific conditions hematocrit-associated 172 Morbidity/mortality 219, see also specific conditions cardiovascular disease 139–141 hemodialysis, by country 143–149 ‘unphysiology’ as cause 208

Subject Index

Blood Purif Vol. 19, No. 2, 2001

Japanese Society of Dialysis Therapy Registry 153 Kt/V, see Dialysis dose Left ventricle diastolic dysfunction 219 function 183 Left ventricular hypertrophy 141, 170–172 anemia 165, 166 hematocrit normalization 161, 162 hemoglobin levels 169 residual renal function 180 short daily hemodialysis 206 Leukocytes 203, 258, 261 transient leukopenia 257 Lipids 147–149, 255 Lombardy Dialysis and Transplant Registry 153–155 Low-flux dialysis 262

267

Mortality, see also specific conditions acute renal failure 229, 234, 238, 239 anemia 157, 172 central venous catheter 252 end-stage renal disease patients, by country 152, 154, 155 hemodialysis 176, 177, 208 vs peritoneal dialysis 180, 182, 183 hypoalbuminemia 146 ICAM-1 148 ß2-microglobulin 262 peritoneal dialysis 176, 177, 179, 180 relative risk 154 sleep apnea 207 sudden death 182, 183 Multiple organ failure 229–231 Myeloperoxidase 201

Platelets 257, 258, 261 Polashegg ionic dialysance equation 198 Polycythemia, ‘relative’ 160 Postdialysis urea rebound 242, 243 Potassium modeling 198 Pravastatin 149 Predialysis mean arterial pressure 218 Profiling, on-line electrolyte 191 Progressive kidney disease 168, 171, 172 Pulse wave velocity 140, 141

National Kidney Foundation-Dialysis Outcome Quality Initiative Guidelines, see Dialysis Outcome Quality Initiative Guidelines Neutrophils 220–202 Nitrendipine 141 Nitric oxide 148 Nocturnal hemodialysis 206–209, 215 Normal Hematocrit Cardiac Trail 170–172 Normalized protein catabolic rate 218, 242 Nutrition 175, 176, 207, 234

Race, patient 154, 219 Reactive oxygen species 255, 256 Recirculation 246, 248, 249 Recombinant human erythropoetin, see Epoetin therapy Relative risk of mortality, see Mortality Remodeling, arterial 140, 141 Renal functional recovery 227–230 Renal replacement 227–231, 238–243, see also Continuous renal replacement therapy Residual renal function 175, 176, 180 automated peritoneal dialysis 186 hemodialysis 178 incremental dialysis 181 Resistivity, post-dialysis 221 Reuse, dialyzer 260

Oliguria 229, see also Residual renal function Outcomes, patient acute renal failure 238, 239 conductivity 197, 198 early initiation of dialysis 175, 176 end-stage renal disease 152–156 peritoneal dialysis 177, 179, 180 Overhydration 195, 217 Oxidant stress, uremia-related 200, 204 Oxidation, plasma protein 201–204 Oxidative stress, inflammation 149 Oxygen transport 169 uptake 171, 172 Paired filtration dialysis 196, 197 Parathyroid hormone 258 Peak predialysis concentration 212, 213 Perfusion, coronary 141 Perindopril 141 Peritoneal dialysis 146, 154, 179–183 automated, residual renal function 175, 176 hematocrit normalization 169 hemoglobin concentration 160 malnutrition 145 sorbent use 256 Phagocyte activation 200–202 Phosphate control 186, 209, 215 nocturnal hemodialysis 207 Physical activity, patient 159, see also Exercise capacity Plasma osmolality 217, 218 Plasma protein oxidation 201–204

268

Blood Purif Vol. 19, No. 2, 2001

Quality of life 155, 182 daily and nocturnal hemodialysis 69, 209, 215 short daily 206 hematocrit normalization 161, 169, 171 mortality 154

Scandinavian Erythropoetin Study 171 Scandinavian Multicentre Study 159, 161, 162, 170, 172 Schedule, dialysis treatment 182, 183, 206–209, 211–216 Sepsis, see Infection, bacterial Serum amyloid A 147 Seven Countries Study 155 Sickle cell disease 159 Sickness impact profile 171 Single-pass systems, hemodialysis 189 Size, patients 208, 209 Sleep apnea 207 Small molecule removal 207, 255, 256, 263 Sodium 195–198, 217–219 Soluble adhesion molecules 148 Solute clearances, acute renal failure 239 control 229 imbalance 227 removal 212 Sorbent augmented dialysis 255–258 Spanish Quality of Life Study 171 Staphylococcus coagulase 251 Sudden death 182, 183 Survival, see Mortality Systemic inflammatory response syndrome 231

Subject Index

Tamm-Horsfall glycoprotein 147 Tassin group study 193, 208, 219 Tattersall algorithms 214 Taurolidine 251–254 Temperature control 245, 246, 248, 249 Thermal energy balance, extracorporeal 245–250 Thermodilution 246 Thirst 217 Time average concentration 207, 213 Time average deviation 208 Transplantation, renal 154, 208, 217 ß2-microglobulin fall 262 pretransplant dialysis modality 181, 182 Ultrafiltration 219, 260 continuous renal replacement therapy 231, 234 Fresenius controlled ultrafiltration dialysis machine 261 profiling 187 Underdialysis 186 United Network of Organ Sharing 182 United States Hematocrit Study 157, 159, 160 cardiac outcome 161

United States Renal Data System 155, 168 Case Mix Severity Study 153, 154 hemodialysis vs peritoneal dialysis 180, 183 ‘Unphysiology’ 207, 208 Urea clearance, equivalent renal 207 concentration, pre/post-dialysis 238 distribution, volume 214 kinetic modeling 211, 238 acute renal failure 240 ‘three-point’ 242 removal rate 238 Uremia ‘new’ toxins 255–257 oxidant stress 200–204 vascular disease 139–141 Urine volume 228 Volume dialysis fluid control 192 overload 227, 229 resuscitation 229, 230 Weight gain 217–221

Subject Index

Blood Purif Vol. 19, No. 2, 2001

269