Autologous Transfusion - From Euphoria to Reason: Clinical Practice Based on scientific knowledge - Nottwil, January 16-17, 2004: Proceedings

Vol. 31, Issue 4, August 2004 Concept Autologe Transfusion 3rd International Symposium Autologous Transfusion – from...

Author: G. Singbartl | W. Schleinzer

88 downloads 693 Views 1MB Size Report

This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form

DOWNLOAD

Vol. 31, Issue 4, August 2004

Concept Autologe Transfusion

3rd International Symposium

Autologous Transfusion – from Euphoria to Reason: Clinical Practice Based on Scientific Knowledge Nottwil, January 16–17, 2004

Proceedings Guest Editors G. Singbartl, Soltau W. Schleinzer, Nottwil

Basel · Freiburg · Paris · London · New York · Bangalore · Bangkok · Singapore · Tokyo · Sydney

Contents · Inhalt

Vol. 31, Issue 4, August 2004 Editorials

Band 31, Heft 4, August 2004 Editorials

199 Autologous Transfusion – from Euphoria to Reason: Clinical Practice Based on Scientific Knowledge Singbartl, G. (Soltau); Schleinzer, W. (Nottwil)

201 Amendment of the Transfusion Act: Easing of Legal Provisions on Salvaged Blood Preparations

199 Autologe Transfusion – von der Euphorie zur Ratio: Praktisches Handeln aus wissenschaftlicher Erkenntnis Singbartl, G. (Soltau); Schleinzer, W. (Nottwil)

201 Änderung des Transfusionsgesetzes: Lockerung der gesetzlichen Bestimmungen zu Eigenblutspenden

von Auer, F. (Bonn)

von Auer, F. (Bonn)

Original Articles

Originalarbeiten

204 Particle Contamination of Salvage Blood Plasma in Cardiac Surgery Engström, K.G. (Umeå)

214 Outcome Quality in Terms of Product Safety in Preoperative Autologous Blood Donation

204 Partikelverunreinigungen im Plasma aus wiedergewonnenem Blut in der Herzchirurgie Engström, K.G. (Umeå)

214 Ergebnisqualität im Sinne der Produktsicherheit bei der präoperativen Eigenblutspende

Walther-Wenke, G.; Pollmeier, A.; Horstmann, E.; Böcker, W. (Münster)

Walther-Wenke, G.; Pollmeier, A.; Horstmann, E.; Böcker, W. (Münster)

Review Articles

Übersichtsarbeiten

221 Quality Management in Blood Salvage: Implementation of Quality Assurance and Variables Affecting Product Quality Hansen, E.; Bechmann, V.; Altmeppen, J.; Last, M.; Roth, G. (Regensburg)

228 Product Quality in Preoperative Autologous Blood Donation – Determinants of Erythropoiesis Weisbach, V.; Eckstein, R. (Erlangen)

221 Qualitätsmanagement bei der maschinellen Autotransfusion: Qualitätskontrollergebnisse und qualitätsbestimmende Einflussfaktoren Hansen, E.; Bechmann, V.; Altmeppen, J.; Last, M.; Roth, G. (Regensburg)

228 Ergebnisqualität im Sinne der Produktwirkung bei der präoperativen Eigenbluspende – Determinanten für die Erythrozytenregeneration Weisbach, V.; Eckstein, R. (Erlangen)

232 Autologous Direct Re-Transfusion – Contra Rosolski, T.; Mauermann, K. (Wismar)

237 Dilutional Coagulopathy, an Underestimated Problem? Fries, D.; Streif, W.; Haas, T.; Kühbacher, G. (Innsbruck)

244 Monitoring of Perioperative Dilutional Coagulopathy Using the ROTEM® Analyzer: Basic Principles and Clinical Examples Innerhofer, P.; Streif, W.; Kühbacher, G.; Fries, D. (Innsbruck)

© 2004 S. Karger GmbH, Freiburg Fax +49 761 4 52 07 14 E-mail [email protected] www.karger.com

Artikel (Volltext) und Inhaltsverzeichnisse sowie das vorläufige Inhaltsverzeichnis des nächsten Heftes: www.karger.com/tmh

232 Autologe direkte Retransfusion – Contra Rosolski, T.; Mauermann, K. (Wismar)

237 Die Dilutionskoagulopathie, ein unterschätztes Problem? Fries, D.; Streif, W.; Haas, T.; Kühbacher, G. (Innsbruck)

244 Monitoring der perioperativen Dilutionskoagulopathie mittels ROTEM® Analyzer: Grundlagen und klinische Beispiele Innerhofer, P.; Streif, W.; Kühbacher, G.; Fries, D. (Innsbruck)

Contents · Inhalt

Vol. 31, Issue 4, August 2004 251 Do We Have enough Evidence to Know when to Transfuse Erythrocytes? Weiskopf, R.B. (San Francisco, CA)

257 Autologous Transfusion in Children: Blood-Saving Techniques Kühbacher, G.; Innerhofer, P. (Innsbruck)

262 Artificial Oxygen Carriers: Hemoglobin-Based Oxygen Carriers – Current Status 2004 Standl, T. (Hamburg)

Band 31, Heft 4, August 2004 251 Verfügen wir ausreichend Evidenz, um zu wissen, wann Erythrozyten zu transfundieren sind? Weiskopf, R.B. (San Francisco, CA)

257 Autologe Transfusion bei Kindern: Fremdblutsparende Maßnahmen Kühbacher, G.; Innerhofer, P. (Innsbruck)

262 Künstliche Sauerstoffträger: Hämoglobinlösungen – Stand 2004 Standl, T. (Hamburg)

269 Haemoglobin Hyperpolymers, a New Type of Artificial Oxygen Carrier – Concept and Current State of Development Barnikol, W.K.R.; Poetzschke, H. (Witten)

282 Blood Irradiation for Intraoperative Autotransfusion in Cancer Surgery – the View of Transfusion Medicine Weisbach, V.; Eckstein, R. (Erlangen)

269 Hämoglobin-Hyperpolymere, künstliche Sauerstoffträger eines neuen Typs – Konzept und aktueller Stand der Entwicklung Barnikol, W.K.R.; Poetzschke, H. (Witten)

282 Bestrahlung und Retransfusion von maschinell aufebereitetem Wund- und Drainageblut in der Tumorchirurgie aus klinisch transfusionsmedizinischer Sicht Weisbach, V.; Eckstein, R. (Erlangen)

286 Advantages of Intraoperative Blood Salvage with Blood Irradiation in Cancer Surgery Hansen, E.; Pawlik, M.; Altmeppen, J.; Bechmann, V. (Regensburg)

286 Vorteile der maschinellen Autotransfusion mit Blutbestrahlung bei Tumoroperationen Hansen, E.; Pawlik, M.; Altmeppen, J.; Bechmann, V. (Regensburg)

293 Meetings and Conferences

293 Tagungen und Kongresse

194 Imprint 294 Guidelines for Authors

194 Impressum 295 Hinweise für Autoren

Prospects for contents of the next issues are given on page 296.


Artikel (Volltext) und Inhaltsverzeichnisse sowie das vorläufige Inhaltsverzeichnis des nächsten Heftes: www.karger.com/tmh

Einen Ausblick auf den Inhalt der kommenden Hefte finden Sie auf Seite 296.

Editorial Transfus Med Hemother 2004;31:199–200

Autologous Transfusion – from Euphoria to Reason: Clinical Practice Based on Scientific Knowledge

‘Your own blood is the safest blood’ – this is the statement of the American Association of Blood Banks (AABB) on its official ‘Autologous Blood Poster’ [1]. It is made for lays in order to promote preoperative autologous blood donation (PABD) and to demonstrate its potential advantages. Autologous transfusion has become an established component of the patient’s blood supply. However, after having done the emotional difficult step ‘from euphoria to reason’ the next logical step of autologous blood management should be done, namely ‘from reason to quality management and quality assurance’. In Germany, quality management concerning both autologous and allogeneic blood transfusion has been given by the German Transfusion Act (Transfusionsgesetz – TFG) from 1998 [2]. Concerning autologous transfusion, quality management was of minor importance to the clinician so far, and stood far behind routine autologous clinical practice. Due to the continuous increase in safety of allogeneic blood, the total risk of autologous blood donation/intraoperative blood salvage plus retransfusion of the appropriate autologous blood products should be smaller than with allogeneic blood transfusion [3]. This, however, means that autologous blood conservation measures have to achieve given quality standards. Therefore, the 3rd International CAT Symposium that was held on January 16–17, 2004 in Nottwil, Switzerland, intensively dealt with quality and safety perspectives of autologous transfusion. Under legal perspectives, PABD which describes both preoperative blood collection and its hemoseparation into autologous packed red blood cells (RBCs) and fresh frozen plasma represents a pharmaceutical procedure comparable to manufacturing of a drug. It has to be made known to the appropriate official local authority and needs its approval for production [4]. German guidelines [5] by the Bundesärztekammer and the Paul-Ehrlich-Institut describe both the appropriate minimal prerequisites with respect to PABD and the corresponding autologous products and the transfusion-specific ed-


Accessible online at: www.karger.com/tmh

ucation for physicians in charge of an autologous blood bank and autologous manufacturing. These guidelines primarily aim at quality measures concerning the safety of the blood products. However, concerning the patient’s benefit from autologous transfusion, it is important to consider efficacy of autologous transfusion measures, too, i.e., increase in total RBC mass is an additional and clinical important quality parameter. Since PABD per se represents nothing else but the transfer of patient’s RBCs from his/her body into a plastic bag, availability of a given number of autologous units does not represent an appropriate or even identical increase in total RBC mass. Only, if the decisive determinants of the efficacy of PABD are considered, PABD will result in an additional increase in RBC mass, and may be associated with a decreased need for allogeneic blood. Under legal perspectives, and in contrast to PABD, neither perioperative blood salvage (PBS) and its mechanical processing by sophisticated technology nor autologous direct retransfusion of salvaged but unwashed wound blood (ADR) nor acute normovolemic hemodilution (ANH) do represent manufacturing of a drug. Due to their timely close relation to the surgical intervention, these measures are considered an integrated part of a patient’s medical treatment, and, therefore, do not need the approval by the local health authority [4]. However, with respect to these autologous measures, quality perspectives referring to product safety and efficacy have to be considered and established, too [6]. In addition, concerning PBS in patients undergoing tumor surgery, medical and legal perspectives are to be discussed with respect to irradiation of the processed blood product before its retransfusion. Intense discussion on this topic at this symposium resulted in partial amendment of the appropriate part of the TFG to legally enable application of PBS, irradiation of the processed blood product, and its retransfusion to this special group of patients [7].

Prof. Dr. med. Günter Singbartl Tannenweg 15, D-29614 Soltau E-mail [email protected]

ADR is mainly applied in cardiac and orthopedic surgery patients. However, this measure is discussed controversially with respect to its efficacy, effectiveness, cost-effectiveness, and potential risks of adverse events. It seems impossible to establish given quality standards for this measure comparable to those of PABD, PBS, and allogeneic RBC units regarding hematocrit, plasma hemoglobin or unwanted by-products being released either from platelets and white blood cells, or resulting from blood coagulation/fibrinolysis. Measures to improve the quality of this procedure and the respective blood product may be of scientific relevance to the researcher interested; however, a reasonable relation between efficacy, effectiveness, cost-effectiveness and clinical benefit to the patient is to be questioned. The discussion on the effectiveness of ANH as a blood/RBC conservation measure is as old as this measure is clinically applied, i.e. older than 20 years. However, this discussion does not consider the so-called dilution coagulopathy that represents both a quantitative problem of volume substitution applied and a qualitative problem of the type of volume substitute administered. The lower the ‘transfusion trigger’ accepted, the more important becomes the topic of dilution coagulopathy. With respect to blood coagulation neither cristalloids nor colloids, neither HES nor gelatine solution are ‘inert’. Another very often neglected perspective is application of autologous transfusion measures in children. Although these techniques might be limited to special centers, both anesthesiologists and clinically interested transfusion specialists should be aware of the basics of this special topic. Artificial oxygen carriers of different chemical nature are under discussion and undergo clinical trials for years. Despite broad clinical interest and expectations, they are still not approved by the governmental authorities and thus are not available for routine clinical usage, at least in Western Europe. While this makes to critically question whether or not artificial oxygen carriers can fulfil what they are promising and we are expecting from them, another group of hemoglobin-based oxygen carriers, the so-called hemoglobin polymers, has come into focus; however, their investigation is in its very early experimental beginnings, and they have still a long way to go until tested in clinical trials. An ‘autologous symposium’ without a fundamental discussion on the rational procedure and individual clinical management and acceptance of the ‘transfusion trigger’ could be consid-

200

Transfus Med Hemother 2004;31:199–200

ered ‘incomplete’ – especially against the background of clinical studies dealing with tolerance of ‘low’ anemic hemoglobin levels by healthy volunteers, various groups of patients, and the clinical experience gathered from acceptance of dilution anemia during the last decade. Last but not least, this topic – like the preceding ones – also stands for the basic idea of this symposium – ‘clinical practice based on scientific knowledge’. By the agreement of the editors and publishers of journals mentioned below, it is possible to publish the lectures presented on this international/interdisciplinary symposium both in German in Anästhesiologie, Intensivmedizin, Notfallmedizin, Schmerztherapie (Thieme, Stuttgart) and in English in TRANSFUSION MEDICINE AND HEMOTHERAPY (Karger, Freiburg i.Br.). The topic of autologous transfusion covers the interest of both anesthesiologists and transfusion specialists. Therefore, this kind of publication in two different languages and journals representing two different medical specialties was chosen by the guest-editors, in order to enable the presentation of the papers to the broad public of the two different clinical specialties interested and concerned. As the papers appear in two different languages, they are not considered a socalled ‘double-publication’ on the corresponding topic. The organizers of this symposium / guest-editors are very grateful both to the editors of these scientific journals for the opportunity to present the papers of this meeting and to the authors for their highly appreciated contributions. G. Singbartl, Soltau, W. Schleinzer, Nottwil

References 1 Autologous Blood Poster: Bethesda, American Association of Blood Banks, 2003. [email protected]. 2 Gesetz zur Regelung des Transfusionswesens (Transfusionsgesetz) vom 1. Juli 1998. Bundesgesetzblatt I, 1998, p 1752, 3 Karger R, Weippert-Kretschmer M, Kretschmer V: Preoperative autologous and plasma donation and retransfusion; in Kretschmer V, Blauhut B (eds): Blood, Blood Products, Blood Saving Techniques. Bailliere’s Clin Anaesthesiol. London, Bailliere Tindall, 1997, vol 11, pp 319–333. 4 Deutsch E, Bender AW, Eckstein R, Zimmermann R: Transfusionsrecht. Stuttgart, Wissenschaftliche Verlagsgesellschaft, 2001. 5 Wissenschaftlicher Beirat der Bundesärztekammer und Paul-Ehrlich-Institut: Richtlinien zur Gewinnung von Blut und Blutbestandteilen und zur Anwendung von Blutprodukten (Hämotherapie). Köln, Deutscher Ärzte-Verlag, 2000. 6 Hansen E, Dietrich G, Kasper SM, Leidinger W, Singbartl G, Wollinsky KH: Vorschläge zum internen Qualitätsmanagement bei der Retransfusion von intraoder postoperativ gewonnenem Wund-/Drainageblut. Anästhesiol Intensivmed 2002;43:81–84.

Singbartl

Editorial Transfus Med Hemother 2004;31:201–202

Amendment of the Transfusion Act: Easing of Legal Provisions on Salvaged Blood Preparations

Washed wound blood used in automated autotransfusion (AAT) is taking on an ever greater importance in autologous blood donation practice. AAT involves, inter alia, that, after washing, the red cells are collected in a bag or other recipient and labelled. This constitutes the manufacture of products that are blood products as defined in section 2 no. 3 of the Transfusion Act (Transfusionsgesetz – TFG) [1, 2]. However, the Federal Constitutional Court ruled that (so-called Frischzellen-Urteil (live cell ruling) of February 16, 2000 – BVerFGE 102,26), as long as the production of such salvaged blood preparations and their operating room use are performed by one and the same person or under his or her instruction, as would usually be the case, the Federal Drug Law (Arzneimittelgesetz – AMG) [3] does not apply (cf. also section 4a sentence 1 No. 3 AMG). The situation changes, however, from both the factual and legal perspectives in the case of wound blood preparations that must be irradiated and, for this purpose, need to be transferred from the operating room to another hospital ward or outside facility so that the producing/using physician temporarily forfeits the power of disposition over the drug. Under the law governing drugs, this act of supply amounts to marketing as specified in section 4 subs. 17 of the AMG. Consequently, the provisions of the AMG apply, especially those governing the manufacturing authorization pursuant to section 13 AMG and the provisions on the ordinance on radiopharmaceuticals and drugs treated with ionizing radiation of 1987 (Verordnung über radioaktive oder mit ionisierenden Strahlen behandelte Arzneimittel – AMRadV) [4]. The AMRadV also stipulates exemptions from the basic marketing ban on radioactive drugs or drugs treated with ionizing radiation as specified in section 7 subs. 1 AMG for bloodbased drugs. According to section 1 subs. 2 AMRadV, bloodbased drugs produced using electron, gamma or X-rays to inactivate blood components may be placed on the market, i.e. supplied to third parties, if certain prerequisites are met. One



of these is for the irradiated blood-based drug to have undergone a marketing authorization procedure. This is logical in the case of similar blood-based drugs which are irradiated according to a standardized procedure because irradiation basically produces a different drug that is tested as part of the marketing authorization procedure and can then be reproduced according to the authorized specimen. However, irradiated wound blood preparations are blood drugs that are produced on a customized basis and, due to the production process and their characteristic features, elude standardized reproducibility. A model marketing authorization, which is an option for other irradiated blood products, such as preoperative autologous blood donations, does not seem feasible for irradiated salvaged blood preparations. This is the technical reason why the Amendment of the TFG, by amending section 1 subs. 2 and 3 of the AMRadV exempts irradiated wound blood preparations from the obligation to obtain a marketing authorization. Thanks to this exemption, these drugs are marketable even without a regulatory marketing authorization. The basic requirement for such products to be covered by a manufacturing authorization is further eased by providing for the so-called ‘minor’ manufacturing authorization. The amendment of section 14 subs. 2 AMG now permits the combination of the functions of production and control manager in one person, not only for autologous blood donations but also for irradiated salvaged blood. This, too, is subject to the prerequisite that the production (specifically irradiation), testing, and use are done within the sphere of responsibility of a hospital department or other medical facility. Hence, the essential criterion is that the producing and testing department or facility has responsibility for and influence on the use of the irradiated salvaged blood. This can be achieved, for instance, if the producing and testing facility designates a correspondingly qualified person to use the irradiated blood product in the operating room or at the ward, or if the producing

Ministerialrat Friedger von Auer Bundesministerium für Gesundheit und Soziale Sicherung Am Propsthof 78a, D-53121 Bonn Tel. +49 228 941-11 50, Fax -49 27 E-mail [email protected]

and testing facility issues instructions, for which the facility assumes responsibility, to the department using the blood. If the product is used in conformity with these instructions by the surgeon or anesthetist, use will be deemed to take place within the sphere of responsibility of the producing and testing department, which implies that the latter is liable both for the accuracy of the instructions and the use of the product according to these instructions. F. von Auer, Bonn

202

Transfus Med Hemother 2004;31;201–202

References 1 Transfusionsgesetz (Transfusion Act), July 1, 1998, Bundesgesetzblatt I, p 1752, at last amended by article 20 of the ordinance of November 25, 2003 Bundesgesetzblatt I, p 2304. 2 von Auer F: Autologous transfusion – pros and cons from the health policy perspective. Infus Ther Transfus Med 2002;29:110–116. 3 Arzneimittelgesetz (Drug Law), December 11, 1998, Bundesgesetzblatt I, p 3585, ammended by law of August 21, 2002, Bundesgesetzblatt I, p 3352. 4 Ordinance of November 28, 1987, Bundesgesetzblatt I, p 502, at last amended by article 10 of the ordinance of July 20, 2001 Bundesgesetzblatt I, p 1714.

von Auer

Original Article · Originalarbeit Transfus Med Hemother 2004;31:204–212

Received: April 7, 2004 Acccepted: May 25, 2004

Particle Contamination of Salvage Blood Plasma in Cardiac Surgery K.G. Engström Heart Center, Cardiothoracic Surgery Division, University Hospital of Umeå, Sweden

Key Words Salvage blood · Cardiac surgery · Microembolic particles · Lipids · Plasma · Capillary flow resistance

Schlüsselwörter Wiedergewonnenes Blut · Herzchirurgie · Mikroembolisierende Partikel · Lipide · Plasma · Kapillärer Fließwiderstand

Summary Background: Blood autotransfusion is performed in cardiac surgery also postoperatively, but more common is the intra-operative return of pericardial suction blood to the aorta. The latter scenario is often discussed in relation to the occurrence of diffuse brain damage due to lipid microembolization from recycled blood contaminated with liquid wound fat. The present study focuses on microembolization from another possible source of lipids in salvaged blood, namely hemolytic debris of membrane phospholipids. Patients and Methods: The capillary pore flow resistance of plasma was measured in vitro and compared with microscopy particle analysis of density-fractionated plasma. A model of controlled hemolysis was employed. Plasma from pericardial suction, venous reservoir, and postoperative mediastinal drains were retrieved from 23 routine cardiac surgery patients. Results: The plasma of pericardial suction blood appeared more deteriorated with respect to capillary-occluding properties than mediastinal blood plasma (p = 0.001). Both types of salvage blood differed from the less affected venous blood plasma. When mediastinal drain blood plasma was density-separated the bottom fraction showed severely deteriorated flow properties (p < 0.001). The finding was partly due to the presence of traumatized erythrocytes that contaminated the plasma. However, the top fraction of plasma demonstrated significantly higher flow resistance than did venous blood plasma (p = 0.003), which point to additional factors of importance. After induced hemolysis apparent erythrocytes disappeared but hemolytic membrane fragments remained in the plasma bottom fraction and affected negatively the capillary flow resistance. Conclusions: It is suggested that hemolytic membrane fragments in salvage blood are an additional potential source of microembolic lipids being different from wound fat-derived contaminants.

Zusammenfassung Hintergrund: Die Eigenbluttransfusion wird im Rahmen der Herzchirurgie angewandt, und zwar sowohl postoperativ als auch – noch häufiger – als intraoperative Rückgabe perikardial gesammelten Bluts in die Aorta. Das letztgenannte Szenario wird häufig im Zusammenhang mit dem Auftreten diffuser Hirnschäden diskutiert, und zwar infolge von Lipidmikroembolien durch wiedergewonnenes Blut, das mit flüssigem Wundfett verunreinigt ist. Diese Untersuchung befasst sich jedoch mit Mikroembolisierungen, die durch in wiedergewonnenem Blut vorkommende Lipide anderen Ursprungs, nämlich, durch hämolysebedingten Debri von Membranphospholipiden, verursacht werden. Patienten und Methoden: Der kapillarporöse Fließwiderstand von Plasma wurde in vitro gemessen und mit der mikroskopischen Partikelanalyse dichtefraktionierten Plasmas verglichen. Es wurde ein Modell kontrollierter Hämolyse eingesetzt. Dazu wurde Plasma bei 23 Patienten mit herzchirurgischen Routineeingriffen aus dem Perikardialbereich, aus dem venösem Reservoir und aus postoperativen mediastinalen Drainagen gewonnen. Ergebnisse: Das perikardial mittels Sauger gewonnene Blut schien – hinsichtlich kapillärer Verschlusseigenschaften – mehr zersetzt zu sein als das aus mediastinalem Blut gewonnene Plasma (p = 0,001). Diese beiden Arten wiedergewonnenen Bluts unterschieden sich vom weniger beeinträchtigten, aus venösem Blut stammenden Plasma. Nach Dichtetrennung von Plasma aus mediastinalem Drainageblut zeigte die Bodenfraktion deutliche reduzierte Fließeigenschaften (p < 0,001). Dieser Befund wurde zum Teil durch die Anwesenheit geschädigter Erythrozyten hervorgerufen, die das Plasma verunreinigten. Aber auch die Spitzenfraktion des Plasmas zeigte signifikant höhere Fließwiderstände als Plasma, das aus venösem Blut stammte (p = 0,003), was auf weitere wichtige Einflussgrößen hinweist. Nach induzierter Hämolyse verschwanden diese Erythrozyten, aber die in der Bodenfraktion des Plasmas verbliebenen hämolytischen Membranfragmente beeinträchtigten den kapillären Fließwiderstand weiterhin negativ. Schlussfolgerungen: Es ist zu vermuten, dass hämolytische Membranfragmente in wiedergewonnenem Blut eine mögliche zusätzliche Quelle mikroembolisierender Lipide darstellen, die sich von den aus Wundfett stammenden Verunreinigungen unterscheiden.



Dr. Karl Gunnar Engström Heart Center, Cardiothoracic Surgery Division University Hospital of Umeå S-901 85 Umeå Tel. +46 90 7858491 E-mai: [email protected]

Introduction The clinical routines of autotransfusion differ between various surgical and anesthetic techniques. In cardiac surgery two principally different conditions of salvage blood use can be identified. During the intra-operative period pericardial suction blood (PSB) is routed back into the arterial line and aorta via the heart-lung machine. This occurs most commonly via a screen filter, but at some centers via selective filters or cell saver washing. The PSB routine is associated with risks in terms of microembolic brain damage [1]. The other procedure is the post-operative return of mediastinal drain blood (MDB). When returned without cell saver washing, MDB has been questioned for causing various complications such as mediastinitis [2], aggravated inflammation, coagulopathy [3, 4], and increased cardiac enzymes [5]. However, several publications report positive outcome and reduced need for homologous blood [6]. Blood-borne particles in relation to the use of cardiopulmonary bypass (CPB) are clinical issues in current scientific focus [7]. Mechanical handling of the aorta is linked with the incidence of stroke [8–10], a risk that is increased if the aorta has atherosclerotic disease. The stroke incidence after cardiac surgery is about 2%, depending on study design and type of surgery [7]. More common than stroke is the diffuse spectrum of cognitive dysfunction which is reported to occur in about 50% of patients undergoing coronary bypass surgery [11]. The mechanisms seem multifactorial. The complex definition of diffuse brain damage, which require extensive psychological testing and consideration of temporal patterns of the symptoms [12, 13], is a scientific challenge. Another factor to deal with is the co-existence of delirium or confusion after cardiac surgery, which is seen in about 20% of the patients [14]. It seems plausible to assume that microembolization contributes to diffuse brain damage although other mechanisms should not be excluded, e.g. inflammation [15–17], cellular rheology [18, 19], or anesthetic and/or CPB routines [7, 20]. The embolus may be gas [21] or may result from technical particles within the CPB circuit [22]. The particles causing embolization may also be of biological origin. Recently it was demonstrated that aortic clamping not only produces macroembolic material but also significant amounts of microscopic debris [10]. However, research in cardiac surgery has mainly focused on the possible role of contaminating wound fat lipids from recycled PSB to cause microembolization. This issue was discussed in the early development of CPB [23] but was neglected during the rapid progress in technology and surgery. The salvaging of PSB must also be seen with respect to its life-saving properties during critical complications in surgery and is also used as an intentional bleeding into the pericardium for surgical de-airing of the heart. The issue of lipid embolism was recently re-addressed [24] with the definition of small capillary and arteriolar dilations (SCADs) in the brain microcirculation. The link between contaminating lipids

Particle Contamination of Salvage Blood Plasma in Cardiac Surgery

from wound fat tissue and SCADs was proven in that cell saver washing of PSB reduced the embolic occurrence [1]. The biophysical consequences of lipid contamination on PSB plasma capillary flow resistance are profound [25]. Nevertheless, the multifactorial nature of diffuse brain damage is an intriguing challenge in the scientific search to identify side effects of PSB recycling. Whereas the intra-operative recycling of PSB is commonly used in cardiac surgery, the post-operative recycling of MDB, in particular without cell saver washing, seems to be more questionable [2–5]. In this respect the salvage blood use in cardiac surgery is similar to many other procedures applying salvaged blood. The simplest way to avoid the problem is by discarding salvage blood. PSB, like many other types of wound blood, is characterized by profound hemolysis but also inflammation [17]. In routine coronary bypass surgery the PSB bleeding volume amounted to about 600 ml [17], a volume loss that is most often tolerated by the patient. Due to the intentional hemodilution during CPB and the even lower hematocrit of PSB, the amount of salvaged blood was found to correspond to about 300 ml blood loss when the volume was recalculated to correct for hemodilution [17]. However, bleeding patterns vary [26], and donor blood transfusion is associated with risk of its own, e.g. immunomodulation [27]. In principle terms, contaminating lipids can be removed by means of density separation (e.g. centrifugal cell saver washing [1]) or absorption (e.g. filtration [28]). The cell saver is considered to be very effective in removing fat [29], whereas filter methods are questioned for their efficacy [28]. The filter efficacy increased significantly when the biophysical lipid properties were modified by cooling although filter clogging and hemolysis were amplified [30]. These methods can be combined which is exemplified in simple terms by a spontaneous density separation column for retained PSB in which lipids also interact with the lipophilic plastic/polyester material [31]. Deteriorated inflammatory activation and coagulation are important questions to address in conjunction with salvage blood use [4]. Cell savers handle this issue effectively, whereas filter techniques are more complex. When a leukocyte-reducing filter was connected to the drain outlet of knee arthroplasty, the frequency of drain blood leukocytes decreased, including some pro-inflammatory cytokines [32, 33]. On the other hand, a complement reaction was triggered, most likely due to a complement-to-surface activation in the filter medium [33]. With respect to lipid embolization and its removal, scientific methodology varies markedly, a fact that hampers interpretation and comparison between studies. The measurement of contaminating lipids at low concentration is fairly complex due to rapid density separation and surface adhesion [31]. A common experimental method is to use soya bean oil [29] in replacement for human wound fat lipids. However, soya oil differs substantially from human lipids in both density separation and surface adhesion properties [34]. The most obvious difference is the temperature-dependent viscosity of human


205

Fig. 1. a Device to measure capillary-pore flow resistance. Medium loaded into a horizontal 10-ml syringe (A) reaches a flow chamber (B) in direct contact with a capillary membrane (C). Medium passes through the device via a pneumatically operated sliding valve (D). Medium emerges from the device and connects to a stop-cock (E). The syringe (A) is modified to have a vent (F) open to air. The flow chamber (B) has a vent (G) for loading and flushing which is occluded during the perfusion experiment. For a similar purpose has the flow channel a vent (H) for air removal and flushing in between experiments. The device is in fluid connection (I) with a beaker (J) on top of a high-precision digital balance (K). The balance measures the accumulated weight of medium that passed the capillary membrane by the force of a constant hydrostatic pressure of 600 Pa. The balance (K) communicates (L) with a computer (M). The experimental design is programmed into the computer (M), that receives data for evaluation and storage, as well as operates the flow device by digital control (N) of the sliding valve (D). b Experimental design to density separate drain-blood plasma into fractions. A 8-mL sample is aspirated from the top and mid layer of the plasma column. The remaining plasma is removed except for the last 10-ml bottom portion. High-density particles collected at the bottom are resuspended in the 10-ml sample. See text for further details.

wound lipids which solidifies at about 10 °C, whereas soya oil is far less temperature-sensitive, a phenomenon that affects also its surface property [34]. The aim of the present study was to analyze the plasma capillary flow resistance of post-operative MDB in comparison to that of intra-operative PSB and to that of venous blood. Moreover, image analysis of MDB plasma density fractions was used to analyze particle contamination and to investigate how hemolysis interferes with the results.

Patients and Methods Patients and Blood Collection Blood was collected from 23 routine cardiac surgery patients (average age of 64.8 ± 1.4 years). Females accounted for 26% of the population. PSB was collected during CPB in an emptied standard infusion 2-liter bag (No 9315, Fresenius Kabi AB, Uppsala, Sweden). Re-transfusion occurred at weaning from CPB. Prior to recycling, a 15-ml PSB sample was collected under vigorous PSB mixing. A corresponding blood sample was taken from the venous reservoir. The patient was fully heparinized with an activated clotting time kept above 480 s. Shed MDB was collected the day after surgery. All patients in whom PSB sampling was performed were operated on for coronary bypass, which also accounted for the great majority of patients in whom mediastinal drains were retrieved. PSB was recycled via a screen filter in all patients, whereas post-operative MDB autotransfusion did not occur. The study design was approved by the Ethics

206


Committee at the Umeå University Hospital (Dnr 01/379, 02/068.02/166, and 03/067). Clinical Details Anesthesia, surgery, and CPB followed standard methods. The CPB prime solution consisted of 1,400 ml Ringer acetate, 60 g mannitol, 160 mmol NaCl, and 7,500 IU heparin. The CPB included a membrane oxygenator with open-circuit venous hard-shell reservoir and a roller pump (Terumo-Sarns 9000 Perfusion system; Terumo Cardiovascular Systems, Ann Arbor, MI, USA). The pump run in non-pulsatile mode and had non-occlusive roller setting. Cold antegrade crystalloid cardioplegia was used. Moderate hypothermia was applied, the acid base management following the α-stat regime, and pump flow was regulated to yield a venous saturation above 70%. Blood Handling Discarded mediastinal drains were emptied via a 200 µm screen filter (Mediplast AB, Malmö, Sweden) to remove blood clots. Heparin 50 IE/ml was added to the defibrinated drain blood to ensure anticoagulation. Plasma fractions (MDB, PSB, and venous blood) were gently collected after 10-min centrifugation at 1,500 g. All experiments were conducted at room temperature 22 °C. Analysis of Capillary Pore Flow Resistance of Plasma The capillary pore flow properties were evaluated using polycarbonate membranes with 5 µm diameter (pore density 4.0 × 105/cm2; Nuclepore Corp., Pleasanton, CA, USA). The filters were mounted in a KGE perfusion device (ESTRAB, Umeå, Sweden, fig 1a). A constant hydrostatic driving pressure of 600 Pa was applied under control of a computer and

Engström

200

150

Blank NaCl

125

Vein blood plasma

a

175 Flow rate, µl/s

Flow rate, µl/s

175

100 MDB plasma

75 50

150 125 100 Top and Mid plasma fractions

75 50

PSB plasma

25

25

0 0

0.5

1

1.5

2

Blank NaCl

2.5

Bottom plasma fraction

0

3

0

0.5

1

Time, s

1.5

2

2.5

3

Time, s

Fig. 2. Capillary-pore flow rate as a function of time, for indicated types of plasma, and with reference to blank-flow of NaCl. Mediastinal drain blood, MDB, pericardial suction blood, PSB, and venous blood plasma. See table 2 for numeric curve analysis. Mean values ± SEM, n = 6 for each plasma group.

200

b

Blank NaCl

175

a pneumatic sliding valve. The accumulated weight was continuously recorded by a digital balance (Mettler PM 480 Deltarange; Mettler Instrumente AG, Greifensee, Switzerland) and online transferred to the computer for analysis. The perfusion duration was 3.15 s, with temporal and weight resolution of 7.5 Hz and 1 mg (~1 µl), respectively. Each filter was calibrated with a blank run prior to the plasma sample (see flow curves in fig. 2 and 3). At flow onset an acceleration occurred before steady state flow was reached (see blank curves). For data evaluation the flow was therefore subdivided into an acceleration (0–0.5 s window) and a steady state phase (0.5–3.0 s window) [35]. At steady state the flow of plasma declined with time due to a gradual occlusion of capillary pores by particles, lipid droplets, or contaminating cells. The negative slope is an indicator of capillary occluding properties. To compensate for filter variations, the relative flow versus blank medium was also calculated (see tables). Further, flow resistance was evaluated as an integrated ‘area under the curve’ (AUC) of each time window. An average flow curve was calculated from triplicate recordings of each plasma sample and blank perfusion. The study was based on 306 individual perfusion experiments. Experimental Design of Plasma Density Separation and Induction of Hemolysis The capillary obstructing properties of MDB plasma were tested in an experimental model. Plasma exposed to 10-min centrifugation at 1,500 g contained contaminating erythrocytes which interacted with the capillary pores. In a second step of plasma centrifugation (5,000 g for 20 min) particles were density separated. The design is illustrated in figure 1b. The effects of hemolytic products on the capillary flow resistance were tested by induced hemolysis. Erythrocytes that contaminated MDB plasma were lyzed by freezing (–18 °C) and re-thawing. Results were compared before and after induced hemolysis. Experimental Design to Measure Particles in Plasma The plasma samples of MDB, before and after hemolysis, were analyzed by computerized image processing. Disposable microscopy chambers were designed having a bottom glass slide and a top 18 × 18 mm cover slip separated by a 0.10 mm thick Parafilm™ (type M; American National Can, Menasha, WI, USA). The plasma sample was vigorously shaken


Flow rate, µl/s

150 Top and Mid plasma fractions

125 100 75 50

Bottom plasma fraction

25 0 0

0.5

1

1.5

2

2.5

3

Time, s Fig. 3. Capillary-pore flow rate as a function of time, for density-fractionated plasma samples of MDB (Top,.Mid, and Bottom). The reference blank-flow is indicated in each panel. a Flow prior to induced hemolysis. See table 3 for numeric data evaluation. n = 6. b Flow after induced hemolysis. n = 5. See table 4 for numeric curve analysis. Mean values ± SEM.

prior to loading. The chamber was mounted on to the stage of an inverted microscope (Olympus CK40-F200; Olympus Optical Company Ltd, Tokyo, Japan) equipped with a CDPlan x40-FPL objective lens. The image was recorded using a black/white camera (C5405–01; Hamamatsu Photonics, Hamamatsu City, Japan). Five microscopic views were measured at random from each plasma sample. The focus depth within the chamber was to gain maximum number of particles. The plasma top fraction had a maximum at top with lipid droplets, whereas in the plasma bottom fraction remaining erythrocytes and membrane fragments accumulated at the chamber bottom. A 2.5-min sedimentation was allowed before measurements were commenced. The microscopic inputs were processed by a computerized image analyzer (Zeiss KS 300, version 3.0; Carl Zeiss Vision GmbH, Hallbergmoss, Germany). A digital light adjustment routine was employed to ensure repro-


207

Table. 1. Curve analysis of plasma capillary-flow resistance in MDB and PSB versus venous blooda

Fraction  A AB B BC C CA (n = 6) post-hoc (n = 6) post-hoc (n = 6) post-hoc Acceleration phase 0.0 to >0.5 s window Slope of the curve – Plasma flow, µl/s 161 ± 22 Relative plasma flow, –/s 0.49 ± 0.10 AUC Plasma flow, µl Relative plasma flow, s

38 ± 2 0.29 ± 0.02

188 ± 22 –0.87 ± 0.31

p value A-B-C ANOVA

197 ± 10 –0.25 ± 0.19

0.018

40 ± 2 0.37 ± 0.02

0.387

41 ± 2 0.35 ± 0.02

0.068

0.039

Steady state 0.5 to >3.0 s window Slope of the curve Plasma flow, µl/s –15.5 ± 1.7 Relative plasma flow, –/s –0.09 ± 0.01

0.001 0.000

–31.1 ± 3.9 –0.20 ± 0.02

0.000 0.000

–8.4 ± 1.9 –0.05 ± 0.01

0.060 0.056

0.000 0.000

AUC Plasma flow, µl Relative plasma flow, s

0.458 0.711

176 ± 19 1.13 ± 0.10

0.003 0.000

293 ± 14 1.90 ± 0.04

0.008 0.000

0.005 0.000

198 ± 24 1.18 ± 0.12

A = MDB plasma; B = PSB plasma; C = venous blood plasma a Curve analysis with reference to figure 2. The flow curves are separated into acceleration phase 0–0.5 s and steady-state phase 0.5–3.0 s. The slope and area-under-curve are calculated for absolute flow and for relative flow, respectively. The relative flow curve (not shown in figures) is the ratio between absolute flow divided with blank flow. The within-group differences are indicated to the right (ANOVA). Post-hoc between-group difference (Duncan) are shown between columns, as indicated in the table head. A p value above 0.05 is considered nonsignificant. Data are given as mean values ± SEM.

Table 2. Curve analysis of plasma capillary-flow resistance in density fractions of MDB, prehemolysisa

Fraction  Top Mid Bottom

p value ANOVA

Number

Acceleration phase 0.0 to >0.5 s window Slope of the curve Plasma flow, µl/s 127 ± 16 Relative plasma flow, –/s –0.84 ± 0.17

225 ± 10 –0.79 ± 0.12

53 ± 16 –1.06 ± 0.18

0.000

6 6


30 ± 2 0.27 ± 0.01

30 ± 1 0.26 ± 0.01

21 ± 2 0.19 ± 0.02

0.000 0.000

6 6

Steady state 0.5 to >3.0 s window Slope of the curve Plasma flow, µl/s Relative plasma flow, –/s

–15.3 ± 0.9 –0.10 ± 0.01

–15.8 ± 0.4 –0.10 ± 0.00

–8.2 ± 1.9 –0.06 ± 0.01

0.001 0.004

6 6


112 ± 20 0.71 ± 0.12

112 ± 15 0.72 ± 0.11

32 ± 6 0.22 ± 0.05

0.000 0.000

6 6

a Curve analysis with reference to figure 3a of density-fractionated MDB plasma before induced hemolysis. The flow curves are separated into acceleration phase 0–0.5 s and steady-state phase 0.5–3.0 s. The slope and area-under-curve are calculated for absolute flow and for relative flow, respectively. The relative flow curve (not shown in figures) is the ratio between absolute flow divided with blank flow. The within-group differences are indicated to the right (ANOVA). Post-hoc data are not indicated due to the obvious group results presented in figure 3a. A p value above 0.05 is considered nonsignificant. Data are given as mean values ± SEM.

208


Engström

Table 3. Curve analysis of plasma capillary-flow resistance in density fractions of MDB, posthemolysisa

Fraction  Top Mid Bottom

p value ANOVA

Number

Acceleration phase 0.0 to >0.5 s window Slope of the curve Plasma flow, µl/s 180 ± 14 Relative plasma flow, –/s –0.51 ± 0.16

192 ± 8 –0.75 ± 0.11

155 ± 16 –0.39 ± 0.31


37 ± 1 0.30 ± 0.01

37 ± 1 0.31 ± 0.01

31 ± 2 0.25 ± 0.02

0.011 0.028

5 5

Steady state 0.5 to >3.0 s window Slope of the curve Plasma flow, µl/s Relative plasma flow, –/s

–9.4 ± 0.9 –0.05 ± 0.01

–8.4 ± 0.8 –0.04 ± 0.00

–13.4 ± 1.2 –0.08 ± 0.01

0.000 0.000

5 5


238 ± 12 1.42 ± 0.05

243 ± 9 1.45 ± 0.04

166 ± 25 1.00 ± 0.15

0.002 0.002

5 5

5 5

a Curve analysis with reference to figure 3b of density-fractionated MDB plasma after freeze-induced hemolysis. The flow curves are separated into acceleration phase 0–0.5 s and steady-state phase 0.5–3.0 s. The slope and area-under-curve are calculated for absolute flow and for relative flow, respectively. The relative flow curve (not shown in figures) is the ratio between absolute flow divided with blank flow. The within-group differences are indicated to the right (ANOVA). Post-hoc data are not indicated due to the obvious group results presented in figure 3b. Data are given as mean values ± SEM.

ducible illumination. The analyzing sequence used one-step pixel dilation and erosion followed by geometric measurement from a preset gray threshold level. Bright field illumination was used. However, in addition phase contrast microscopy was applied to detect particles of hemolysis which did not give contrast in bright field microscopy. Maximum diameter, shape factor, and projected area were recorded. A cut-off filter removed small-size debris corresponding to 0.77 µm2. The data were transferred to an Excel spreadsheet programmed to separate particles into various types. The projected area was re-calculated to its corresponding circular diameter. Expected cellular particles were sub-sorted within a 5–15 µm diameter window of circular shape (shape factor 0.7–1, 1 = circular). Platelet-type circular particles below 5 µm were separated from similar-size particles of irregular shape (shape factor < 0.7). Irregularily shaped large particles (>5 µm) constituted the last group. Statistical Methods All data are given as mean values ± SEM. Continuous variables were analyzed using one-way or two-way ANOVA and, if significant, Duncan’s multiple range test for post-hoc comparison. In addition paired and unpaired Student’s t tests were performed when appropriate. A p value above 0.05 was considered non-significant. Statistica version 6.1 (StatSoft, Tulsa, OK, USA) was used.

Results Capillary Flow Resistance of MDB and PSB Plasma versus Venous Blood Plasma The capillary flow properties of MDB, PSB, and venous blood plasma differed significantly. This is illustrated in figure 2 and by curve analyses in table 1. In brief, the NaCl blank (0.5–3.0 s window) showed a horizontal slope with time (p = 0.707), indicating no capillary obstruction. All plasma components were


characterized by a decline in flow rate with time, being highest for PSB followed by MDB (both p < 0.001 versus horizontal plane). Venous blood plasma also showed a decline due to particle clogging (p = 0.008) although it was less steep than that found with salvaged blood (table 1). PSB differed from MDB and venous blood plasma (table 1). There was a borderline significance for slope values when comparing MDB and venous plasma. AUC gave high statistical output for group difference, whereas PSB and MDB did not differ (table 1). Notably the slope is sensitive to particle-to-capillary interaction, whereas the AUC integrates also peak flow rate and plasma viscosity. Capillary Flow Resistance of Density-Fractionated MDB Plasma versus State of Hemolysis When MDB plasma was density separated, the bottom fraction significantly differed from the top and mid fractions with respect to capillary flow resistance (fig. 3a, table 2). The top and mid fractions had identical flow patterns. When comparing these two fractions with venous blood, the flow patterns also differed (ANOVA p = 0.003), indicating that the deviant capillary flow resistance of MDB is not only related to particles of the bottom fraction of MDB. If frozen/re-thawed MDB plasma was density fractionated the capillary flow results were improved when compared with those of pre-hemolytic unfrozen plasma (fig. 3b, table 3). There was an upward shift in flow curves (reduced capillary flow resistance) being significant for the plasma top and mid fractions (AUC, 0.5–3 s time window, both p = 0.001), but not for the plasma bottom fraction (p = 0.128).


209

Number of particles, n/view/ml

60

a

Circular 5–15 µm 50 40

Circular 5 µm Non-circular 30%, anti-D is by far the most frequent alloantibody followed by anti-Kell, anti-CD, antiE, anti-LE(a), anti-Fy(a), anti-c, anti-Jk(a), and anti-Cw. Quality Controls on Autologous Products The blood bag systems, stabilizers and production processes used at the ITM MS for preservation are identical for autologous and allogeneic blood. The following erythrocyte products have been manufactured since 1988: – January 1988 to November 1989: CPDA-1 EC, buffy coatfree, shelf life 35 days for allogeneic and autologous EC; – December 1989 to April 1995: SAGM-EC, buffy coat-free, shelf life 35 and 42 days for allogeneic and autologous EC, respectively; – May 1995 to August 2000: PAGGS-M-EC, buffy coat-free, shelf life 42 and 49 days for allogeneic and autologous EC, respectively; – from September 2000: PAGGS-M-EC, leukocyte-depleted, shelf life 42 and 49 days for allogeneic and autologous EC, respectively. The 7-day longer shelf life was chosen in order to have an interval as long as possible between the last autologous blood sample and the operation to allow for a maximal erythrocyte gain through patient’s hematopoiesis. Since June 2000, the hemotherapy guidelines [4] define the hemolysis rate of erythrocytes and sterility testing on 1% of products as parameters for the quality control of autologous products. Since then, EC and plasma not required for transfusion have been returned to the

Product Safety in Preoperative Autologous Blood Donation

Table 3. Hemolysis rate in autologous and allogeneic (shelf life 42 days) EC in PAGGS-M, leukocyte-depleteda Hemolysis rate in % of erythrocyte mass   autologous EC allogeneic EC (n = 74) (n = 4,819) Mean Standard deviation

0.28 0.15

0.19 0.10

Median Minimum Maximum

0.24 0.03 0.8

0.17 0.40 0.73

a Evaluation

of study protocols from January 1, 2002 to June 30, 2003.

ITM MS by hospitals. Due to the 7-day longer shelf life, the hemolysis rate of autologous EC in the products tested is slightly higher than that in allogeneic EC, (table 3). The limit value of the hemolysis rate is defined as 0.8% of the erythrocyte mass of an EC. As the sterility of preserved autologous blood is classified as a particularly critical parameter, sterility testing has been performed systematically since 1992. Initially almost all autologous blood samples were tested, but since 1996 only a randomly selected unit of autologous blood per collection date has been tested. At present, the spot check is 1% of EC and plasma. In total, 11,415 autologous units have been found to be sterile (table 4). Self-Inspection and Official Inspections The industrial regulation for pharmaceutical companies [5] requires the systematic self-inspection of drug manufacture. Self-inspection by the cooperating hospitals in all areas involved with autologous blood sampling and storage occurs at 1- to 2-year intervals, based on a standardized inspection form. The findings obtained provided an opportunity for the completion of documentation, for the clear definition of responsibilities, for systematic equipment tests, for the optimization of hygiene instructions, and for the development of personnel training programs. Official inspections of the hospital departments involved in au-


217

Table 4. Results of sterility testinga

1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 (until June) Autologous blood samples tested First culture positive Second culture prepared Second culture positivec

327 0

958 1,655 1,251 1,332 1,545 1,569 1,527 7 16 8 0 0 0 0 4b 16 8 0 0 0

891 0

227 0

105 0

28 0

a Methods:

1992–1995 DAB direct inoculation aerobically and anaerobically; 1996 to June 2003 automatic procedure BacT/Alert. Material: 1992 to June 2002 buffy coat; July 2002 to June 2003 nontransfused products (EC and GFP). b No repetition possible in 3 cases. c Overall, none of the 11,415 autologous blood samples was tested confirmed positive.

Table 5. Reports on transfusion reactions from 1988 to June 2003 (overall 36,000 transfused autologous EC) Transfusion reactions Clinical features

n

Rise in temperature to 38.6 C 3 h after transfusion Chills, hypotension Chills Dizziness 30 min after transfusion Erythema, facial swelling, rise in temperature at end of transfusion Skin marbling, labial cyanosis, labial swelling 1 h after transfusion Severe hypotension, bronchospasm, decrease in 02 saturation, nausea, vomiting, immediate improvement after discontinuation of transfusion

1 1 2 1

Total

1 1

1  8 (0.023%)

tologous blood collection are performed by inspectors from the district authorities, also at intervals of 1–2 years. The focuses of these inspections are typically SOPs, forms, documentation, monitoring of hygiene plans, training documentation, temperature records, and apparatus logbooks. The strategy of increasingly detailed requirements and tests is clearly apparent. Adverse Events in Autologous Blood Transfusion Over the study period, 8 transfusion reactions related to the transfusion of autologous EC were reported to the ITM MS (table 5). The symptoms consisted predominantly of chills and a rise in temperature, in some cases accompanied by allergic reactions. The laboratory tests to establish the etiology, including bacteriological tests, yielded negative results. With a consumption rate of 74%, the rate of reaction for 36,000 transfused autologous blood EC is calculated as 1:4,500 (0,023%). In 1 case, there was a mix-up of an autologous fresh plasma which was administered to another patient with the same ABO blood group. Serious complications requiring long-term dialysis were observed in 1 female patient whose 2 erythrocyte concentrates were incorrectly stored deep-frozen and as a result were completely hemolytic on transfusion.

218


Discussion This article presents data from a preoperative autologous blood program in which the product-related operating steps in donation, processing, and testing are under the responsibility of transfusion physicians while patient-related tasks lie predominantly in the hands of clinicians. One fundamental benefit of this cooperation is that the extensive requirements of the AMG [6] can be met much more readily. The good manufacturing practice(GMP)-compliant manufacture and testing of drugs from blood requires special knowledge and GMP-compliant premises and equipment [7] which are normally not found in hospitals. The initial testing of the autologous blood and accompanying documents on receipt for processing has been shown to be an essential step in guaranteeing product safety. In 6 of 100 autologous blood units abnormalities are found, which must be assessed to determine whether they might entail a risk in transfusion. Clot formation and underfilling are the consequence of problems in collection which result essentially from difficult venous conditions in patients. Deviations from the detailed collection operating procedure result in a, to all intents and purposes, problematical destruction of preserved autologous blood [8]. Autologous blood patients benefit from the consistent separation of autologous blood into components and the use of optimal additive solutions for the longest possible preservation of erythrocytes because of the saving in time and in the biochemical quality of the erythrocytes. An additional improvement in storage quality is obtained with leukocyte depletion [9, 10]. In addition, visually undetectable clots in the EC result in blockage of the leukocyte depletion filter and, hence, the exclusion of autologous blood of critical quality. The routine quality control of autologous EC involves the determination of the hemolysis rate and shows values within the specified range for a a shelf life of 49 days. As autologous blood is not a finished medicinal product, a minimum quantity of 100 ml of erythrocytes is tolerated as the lower limit in the patient’s interest. This corresponds to an hemoglobin content of at least 25 g per EC compared to allogeneic blood EC with at least 40 g and a mean of 56 g per EC. As expected, the infection serology test results of autologous

Walther-Wenke/Pollmeier/Horstmann/Böcker

blood patients show a fundamental difference from blood donor test results [11]. In autologous blood patients, the study objective is to establish an infection status and, where necessary, to forgo autologous blood collection. The screening of allogeneic blood donors on the other hand is intended for the safety of transfusion recipients. The introduction of detection of the virus genome of HIV, HCV, and HBV has allowed an improvement in safety which reduces the residual risk of HIV to < 1:11,000,000, of HBV to < 1:500,000, and of HCV to < 1: 13,000,000 [12]. Autologous blood products cannot and should not meet the high safety requirements of allogeneic products. However, this implies particular demands on the prevention of mix-ups in transfusion. After the successful reduction in transfusion-associated risks of viral infection, bacterial contamination of blood components, and the associated risk of sepsis has for some time now taken center stage. A series of case reports describes bacteria-associated transfusion complications through contaminated autologous blood products [13]. The hemotherapy guidelines [4] take account of this risk in that particular attention must be paid to risk factors for bacteremia in the anamnesis taken prior to autologous blood collection. However, skin disinfection and the technique for collection of blood also require particular care since predominantly organisms of the resident and transient skin flora appear as contaminants of blood components. Systematic monitoring of autologous blood for sterility has occurred in the ITM MS since 1992. As a destructive test of the product was not possible, until June 2000 the buffy coat obtained during the production of the components underwent stability testing. In this way, the test was performed shortly after donation. By contrast, sterility testing is currently performed on nontransfused components on expiry of the shelf life. Assessment of own data leads to the conclusion that under the conditions practiced the risk of contamination in autologous components is at least not higher than that in allogeneic components. The ‘Microbiological Tests in Transfusion Medicine’ subgroup of the German Advisory Committee Blood (Arbeitskreis Blut) has retrospectively collected data on contamination rates from medical transfusion establishments in Germany over various periods (1995/1996, 1998, 2001) [personal communication, 14]. The data from 1998 showed contamination rates of 0.3% for autologous EC and of 0.17% for allogeneic EC. The compiled data from 2001 showed a contamination rate of 0.22% for autologous EC, which is significantly higher than that of allogeneic EC amounting to only 0.13% (p = 0.01, Fisher’s test). On interpreting the data, it should be borne in mind that they can be assessed as being an indication of a difference between the two product groups, but not as evidence of a higher rate of contamination of autologous units. In the blood collection and processing steps which involve the risk of bacterial contamination particular care is, in the patient’s interest, essential in both the allogeneic and the autologous areas. Routine self-inspections in cooperating hospitals under the su-

pervision of the production manager responsible under the AMG aim at comparing nominal and actual results [15]. They have resulted in the large-scale elimination of deficiencies, the acceptance of GMP requirements, and constructive cooperation between hospitals and transfusion physicians. To this extent, self-inspections are of great benefit as a quality assurance measure. However, the frequency of regulatory inspections has increased considerably in the last years, coming along with a shift of emphasis to SOPs, forms, documentation and, recently in particular, to the qualification of rooms and equipment as well as to validation processes. The requirements of the regulatory officials are becoming increasingly detailed and but are not always adopted with insight and understanding. In the authors’ view, a constructive dialogue on the purpose and benefit of a series of official requirements and measures is necessary. In order to obtain information about adverse reactions in autologous blood transfusion, users are encouraged to report these to the ITM MS as soon as possible and to forward the remaining product together with a sample of the patient’s blood. The total of 10 reported cases have been classified in relation to whether the products were used as instructed. Transfusion reactions were reported in 8 cases which may clinically be classified predominantly as febrile, nonhemolytic transfusion reactions. Comparable reactions have been described In other reports on autologous transfusion [16, 17] transfusion reactions were described to occur with similar frequency (0.1203–0.043%). Soluble biological response modifiers which accumulate in the product during storage, particularly cytokines, are assumed to be the cause of the symptoms and can induce typical reactions not only after application of allogeneic but also of autologous products [13]. The prolonged storage period of autologous EC, which as a rule are transfused shortly before the expiry of the shelf life, is also a predisposing factor. There were numerous reports [18, 19] on mix-ups of autologous blood in transfusion and on problems of handling errors. Thhe risk of mistransfusion in which a sample is transfused to a wrong recipient is given as 1:16,000 to 1:25,000. Such mixups bear a substantially higher risk of transfusion-associated infection when autologous blood products were transfused. In summary, quality assurance measures, consecutive controls in the various operating stages, and attentiveness during use are of particular importance in a preoperative autologous blood program. Autologous components manufactured and tested by sampling according to the operation modes described here are equivalent in terms of quality to allogeneic products, but exhibit a greater range of variation in terms of their erythrocyte content. Since autologous blood harbors an increased residual risk of transfusion-associated viral infection, mix-ups in use must be considered particularly critical. Incorrect allocation and handling errors continue to be a serious problem in autologous and allogeneic transfusions.

Product Safety in Preoperative Autologous Blood Donation


219

References 1 Paul-Ehrlich-Institut: Bericht zur Meldung nach § 21 TFG für die Jahre 1999 und 2000. Bundesgesundheitsbl Gesundheitsforsch Gesundheitsschutz 2003;46:1016–1032. 2 Pollmeier A: Präoperative Eigenblutgewinnung in einem überregionalen DRK-Blutspendedienst – retrospektive Analyse eines Patientenkollektivs aus 9 Jahren (1988–1996). Inauguraldissertation, Medizinische Fakultät der University DuisburgEssen, 1999. 3 Communication from the Blood Working Party of the Federal Ministry of Health: Mindestanforderungen zur Sterilitätstestung von Blutkomponenten. Bundesgesundheitsbl 1997;11:452–453. 4 Wissenschaftlicher Beirat der Bundesärztekammer, Paul-Ehrlich-Institut: Richtlinien zur Gewinnung von Blut und Blutbestandteilen und zur Anwendung von Blutprodukten (Hämotherapie). Köln, Deutscher Ärzteverlag 2000. 5 Industrial Regulation for Pharmaceutical Companies (PharmBetrV) of 08.03,1985 (Federal Gazette I P. 546) Last Amended by § 35 of the Transfusion Act of July 1998 (Federal Gazette I P. 1752). 6 Gänshirt KH: GMP-GLP-QS-Richtlinien bei der Eigenblutherstellung; in Mempel W, Mempel M, Schwarzfischer G, Endres W (Hrsg): Eigenbluttransfusion aus heutiger Sicht. Hämatologie. München, Sympomed, 1996, vol 5, pp 100–109.

220

7 Eckstein R, Weisbach V: Mindestvoraussetzungen und Qualitätssicherung bei der präoperativen Eigenblutspende; in Mempel W, Mempel M, Schwarzfischer G, Endres W (Hrsg): Eigenbluttransfusion aus heutiger Sicht. Hämatologie. München, Sympomed, 1996, vol 5, pp 155–166. 8 Walther-Wenke G, Böcker W, Horstmann E: Eigenblut – das bessere Blut? Ergebnisse der Konserveneingangskontrolle; in Mempel W, Mempel M, Schwarzfischer G, Endres W (Hrsg): Eigenbluttransfusion aus heutiger Sicht. Hämatologie. München, Sympomed, 1996, vol 5, pp 236–283. 9 Mansouri Taleghani B, Langer R, Grossmann R, Opitz A, Halbsguth U, Buchheisler A, Schuler S, Bachthaler R, Wiebecke D: Verbesserung der biochemischen und rheologischen Qualität von Vollblut und Erythrozytenkonzentraten durch Leukozytendepletion vor Lagerung. AINS 2001;36(suppl 1): 511–519. 10 Walther-Wenke G, Walker WH: Etablierung der leukozytendepletierten autologen Vollblutkonserve. AINS 2002;37:686–689. 11 Robert Koch-Institut: Epidemiologisches Bulletin Nr. 13. March 31, 2000. 12 Seifried E, Findhammer S, Roth WK: Status of NAT Screening for HCV, HIV and HBV – Experiences of the German Red Cross Blood Donation Services; in Brown F, Seitz R (eds): Advances in Transfusion Safety. Dev Biol. Basel, Karger 2002, vol 108, pp 23–27.


13 Yomtovian R, Praprotnik D. Adverse consequences of autologous transfusion practice; in Popovsky MA (ed): Transfusion Reactions. Bethesda, AABB Press, 2001, pp 261–314. 14 Walther-Wenke G, Dörner R, Baumann B, Brandstädter W, Exner M, Heinz H-P, Lange H, MontagLessing T, Trobisch H, Werner E: Methoden und Ergebnisse der Sterilitätstestung von Blutkomponenten in Deutschland 1995 und 1996 Infusionsther Transfusionsmed 1998;25(suppl 1):1, abstract no 1/2m. 15 Böcker W, Horstmann E, Walther-Wenke G: Selbstinspektion – ein wichtiger Bestandteil der Qualitätssicherung; in Mempel W, Mempel M, Schwarzfischer G, Endres W (Hrsg): Eigenbluttransfusion aus heutiger Sicht. Hämatologie. München, Sympomed, 1996, vol 5, pp 110–113. 16 Renner SW, Bracey AW, Yomtovian R: Autologous blood Transfusion Practices. CAPQ-Probe 95–08. Chicago, College of American Pathologists, 1995. 17 Domen RE: Adverse reactions associated with autologous blood transfusion: Evaluation and incidence at a large academic hospital. Transfusion 1998;38:301–306. 18 Shulman I: Your worst transfusion night-mares. Workshop presentation. American Society of Clinical Pathologists. Orlando, FL, October 16,1993. 19 Linden JV: Errors in transfusion medicine. Scope of the problem. Arch Pathol Lab Med 1999;123: 563–565.

Walther-Wenke/Pollmeier/Horstmann/Böcker

Review Article · Übersichtsarbeit Transfus Med Hemother 2004;31:221–227

Received: April 7, 2004 Accepted: May 9, 2004

Quality Management in Blood Salvage: Implementation of Quality Assurance and Variables Affecting Product Quality* E. Hansen

V. Bechmann

J. Altmeppen

M. Last

G. Roth

Klinik für Anästhesiologie, Universität Regensburg, Germany

Key Words Quality management ⋅ Blood salvage ⋅ Intraoperative autotransfusion ⋅ Free hemoglobin ⋅ RBC recovery ⋅ Plasma washout

Schlüsselwörter Qualitätsmanagement ⋅ Intraoperative Autotransfusion ⋅ Freies Hämoglobin ⋅ Erythrozytenausbeute ⋅ Plasmaauswaschrate

Summary Implementation of quality management in intraoperative blood salvage with controls of product and process quality supports early recognition and repair of dysfunction if respective actions are laid down in the quality management handbook. In order to avoid insufficient quality and to improve process quality, a broad understanding of the processes and of the variables affecting quality is needed, which is based on experimental tests. For this purpose the use of fresh whole blood as test blood and of protein as parameter for determination of the plasma elimination rate is favorable over outdated banked blood and free hemoglobin. Tested this way, the process of blood collection by suction is by far not as harmful to RBCs as expected. Partially filled bowls, lower wash volumes, or a fastened filling or washing should be avoided. Plasma washout can be improved by higher wash volumes or by slower filling and washing, which avoids increased loss of RBCs. Quality management which is based on a better understanding of the procedure as well as quality controls can help to supply high-quality blood for optimal hemotherapy by blood salvage.

Zusammenfassung Die Umsetzung eines Qualitätsmanagements bei der maschinellen Autotransfusion mit Kontrollen der Produktund der Prozessqualität kann dazu beitragen, frühzeitig Fehlfunktionen zu erkennen und abzustellen, wenn entsprechende Vorgehensweisen im Qualitätssicherungshandbuch festgelegt worden sind. Das Abstellen von Fehlern und die Verbesserung der Prozessqualität erfordern ein gründliches Verständnis des Prozesses der Blutaufbereitung, das auf experimentellen Testungen von Einflussgrößen basiert. Grundsätzlich ist dafür die Verwendung von frischem Vollblut und von Gesamteiweiß als Parameter an Stelle von abgelaufenen Blutkonserven bzw. freiem Hämoglobin von Vorteil. So getestet, stellt sich der Ansaugvorgang als bei weitem nicht so problematisch wie angenommen heraus. Halbvolle Glocken, ein geringes Waschvolumen oder eine Beschleunigung von Füll- und Waschvorgang sind zu vermeiden. Eine Verbesserung der Auswaschrate ist mit erhöhten Waschvolumina zu erreichen oder durch langsameres Waschen, wobei ein zusätzlicher Erythrozytenverlust vermieden werden kann. Das Qualitätsmanagement mit einem verbesserten Verständnis des Blutaufbereitungsprozesses und mit Qualitätskontrollen trägt wesentlich dazu bei, mit Hilfe der maschinellen Autotransfusion Blut höchster Qualität für eine optimale Hämotherapie bereitzustellen.

* A German version of this article is published in Anästhesiologie, Intensivmedizin, Notfallmedizin, Schmerztherapie (Thieme, Stuttgart).



Introduction As with any treatment and use of blood, a quality management (QM) is also indicated in case of intraoperative blood salvage (IBS). A proposal for such a QM system has been worked out and published [1]. In the following, the implementation of these concepts and their performance in clinical practice are described. In addition, experimental investigations focusing on a better understanding of the method and its variables are presented and discussed with regard to clinical impact. For the structural quality, regulation of the responsibilities as well as periodical education and training of staff is necessary. In addition, there is a need for regional workshops on QM in IBS to support customers in the implementation of QM handbooks and standard operating procedures (SOPs). Structural quality calls for a written description of the procedure and its quality assurance. Links should be made in the QM handbook for hemotherapy of the hospital addressing the indications for IBS together with the indications of type-and-screen and of blood supply in the specific surgical procedures. Nevertheless, with respect to the ongoing discussion whether salvaged blood is a drug and under the respective regulations or is part of the surgical therapy [2], a special QM handbook for IBS seems to be advisable. It should be written by the department or institution that actually performs the IBS. Contra-indications and indications should be listed. It should also be indicated if IBS is an integral part of the surgical procedure or if standby collection is provided and complete setup of the apparatus and processing of blood is performed only after collection of a sufficient amount of wound blood. For quality assurance, parameters and procedures of quality control as well as consequences in the case of aberrant test results must be defined and laid down in the IBS QM handbook. Quality Assurance in Intraoperative Blood Salvage Quality controls together with documentation and a broad understanding of the procedure are essential factors of process quality. For documentation, a machine-readable form was developed at our university hospital. The following data are documented: – personal identification, – surgical procedures with indication for IBS, with specifications of applications in cancer surgery or pediatric surgery, – equipment- and process-related data such as apparatus type, identification of individual apparatus, used disposals (e.g. centrifuge bowl, reservoir), and the applied program (in manual mode also filling rate, washing rate and wash volume), – start and end of collection, – volume of shed blood collected in the reservoir (minus volume of anticoagulant solution), estimated total blood loss and, if applicable, postoperatively collected and processed blood from drainages,

222


– amount of washed RBC concentrate produced and its ‘strength’, i.e. the hematocrit (HCT), derived from product control, – amount of transfused salvaged blood to indicate unambiguously if blood was only collected and processed or also transfused, – if applicable, irradiation of salvaged blood in cancer surgery [3] and irradiation dose, – total number of transfusions, including predonations, to identify the contribution of IBS to total hemotherapy. Further important factors of process quality are the definition and application of quality controls as well as the definition of provisions if the set threshold values are not fulfilled. For assessment of product quality, the parameters volume and HCT have been proposed. The produced blood volume is measured in sufficient accuracy by the autotransfusion apparatus. The HCT can be evaluated with sufficient precision by blood gas analysis once during every application of IBS which is feasible since HCT and hemoglobin testing is routinely done during transfusion-relevant surgery. The control of product quality can serve as an important early indicator of any dysfunction in the process. We have eliminated a specific program mode with automatically elevated washing rate based on such testing after values repeatedly fell below the set limits. The subsequent consequences of a failure to reach the proposed product HCT of 50% as set down in the QM handbook are the repetition to exclude a measuring error and, after confirmation of the failure, a process control. The control parameters of process quality are the RBC recovery and the plasma elimination rate. RBC recovery can be calculated from the volumes and the HCT values in the processed wound blood (sample from reservoir) and in the produced RBC concentrate (sample from transfusion bag). The plasma elimination rate is derived from the respective supernatant volumes and concentrations of an indicator such as total protein, albumin, free hemoglobin (frHb), or potassium ions [1]. Alternatively, assessment of the frHb residual contamination of the product is discussed. Microbiological testing of sterility is not necessary since positive cultures of skin- and air-borne bacteria are found frequently without any clinical relevance or consequences as this blood is not stored and therefore bacterial proliferation does not occur. The proposed limit for RBC recovery and plasma elimination is 80 and 90%, respectively. For control of process quality every autotransfusion device in clinical use should be tested at least every 3 months. For this purpose, a blood sample is drawn from the reservoir after thorough mixing of the content, and the HCT and the total protein concentration in the supernatant is determined. Subsequently the centrifugation is started, the processed and produced blood volume documented, and a sample drawn from the product for measurement of HCT and total protein concentration. After feeding these data in a data sheet, the RBC recovery and plasma washout are calculated automa-

Hansen/Bechmann/Altmeppen/Last/Roth

Table 1. Disadvantages of outdated RBC as test blood

RBC buoyant density RBC cell size Plasma Aggregates RBC cell membrane Hemolysis

Outdated banked RBC

Fresh whole blood donation with 5% hemolysis

altered altered – +++ rigid variable

normal normal + – normal defined

standardized, nonreproducible composition, e.g. highly variable HCT, or varying extent of hemolysis. Most often outdated banked blood is used for its ready availability and low or no costs. But it is well known that prolonged storage at low temperature results in significant changes in buoyant density (ρp) and cell size (diameter dp) of RBCs. The RBC concentrates in additive solution generally used contain almost no plasma; thus density (ρf) and viscosity (ηf) of the medium is significantly reduced compared to plasma. Consequently, 4 of the 5 parameters in Stoke’s law of sedimentation vs=

tically. If the test results fall below the fixed limits, the test is repeated as set down in the QM handbook. With protein as the parameter of elimination the test results are available within a reasonable time, and the control measurement can be performed the same day, often even with the same disposal. In case there is no wound blood left, test blood is used. If the rerun of the test confirms the insufficient quality of the process, an analysis of potential causes and a modification of process variables has to follow. For the latter a thorough understanding of the process and of the variables that affect quality is essential. During 80 routine (monthly) quality controls performed within the last 18 months, only in 4 cases the results did not fulfill the standard requirements, making a control measurement necessary. In two cases, the low values of RBC recovery (and product HCT) were not confirmed and were attributed to sampling during the ‘first run’. In one case the sensor for detection of the buffy coat at the shoulder of the bowl had to be adjusted, in another case a program which automatically elevates the filling rate was eliminated. Therefore, quality controls can facilitate in clinical practice early detection of technical dysfunction and improve process quality.

dp2 × (ρp-ρf) ×a 18 × ηf

(1)

are affected. It is therefore clear that the behavior of these RBCs during centrifugation and cell separation and washing can hardly represent salvaged wound blood. Formation of aggregates and loss of cell membrane flexibility also affects sedimentation of the RBCs. In addition, the extent of hemolysis is very variable in outdated banked blood. For these reasons we use whole blood from fresh donations, and add a defined amount of hemolysed blood (e.g. by freezing and thawing of 5% of the blood volume) (table 1).

Free Hemoglobin

In order to guarantee constant high performance and quality in intraoperative blood salvage, testing of new equipment, programs, and disposals is advisable, as is a thorough understanding of the process. For both purposes, equipment testing and experiments on variables affecting quality the same parameters as for the assessment of process quality, namely RBC recovery and plasma elimination, are especially useful. In the literature often only one of these parameters is reported to be used. The tests are not part of the clinical routine and are not in the responsibility of the customer, but are reserved to institutions with corresponding research interests and facilities. Nevertheless, the customer should have a basic understanding of data acquisition and interpretation. Therefore, two fundamental topics are discussed here as preliminary remarks: ‘the adequate test blood’ and ‘frHb as a quality parameter’. Wound blood is a relatively unsuited source of test blood for experiments with autotransfusion devices, because of non-

frHb is commonly used in quality control for assessment of plasma elimination rate or residual contamination in salvaged blood. This is also true for the proposals concerning QM in IBS [1, 4]. In this regard, it is of note that the frHb elimination rate always is significantly lower than the protein elimination rate [5–7] (e.g. 96 vs. 99%, or 80 vs. 91%). In own investigations on the damaging effects during suction and collection of wound blood the use of fresh blood with no hemolysis resulted in elimination rates for frHb below 50% or yielded even negative elimination rates. This is also reflected by the fact that the concentration of frHb in the salvaged and washed blood leveled at about 0.3g/l, irrespective of whether starting concentration was 3.0 g/l (blood with hemolysis) or 0.03 g/l (blood without hemolysis). In the latter case hemolysis in the product (concentration of frHb, in some cases also the amount of frHb) exceeded that in the starting material. This means that hemolysis was added during processing of salvaged blood, probably during emptying, since usually no frHb is left in the waste at the end of the wash step. Consequentially, frHb is a ‘mixed parameter’, i.e., it is determined by the effectivity of the washout procedure as well as by (little) hemolysis related to the process. Since hemolysis is low, it has no effect on the protein concentration in consideration of very high initial plasma values. Measurements of protein concentrations result in valid and reproducible data on effective plasma elimination (table 2), reflecting the elimination of most of the soluble contaminating agents such as activated factors, mediators, cytokines or enzymes. The process-associated hemolysis of

Quality Management in Blood Salvage


Experimental Testing of Devices and Process Variables

223

Table 2. Disadvantages of free hemoglobin as quality parameter of IBS frHb

Total protein

simple expensive not routine time consuming

cheap routine results rapidly available

Disturbances

hemolysis during/after sampling (e.g. centrifugation)

–

Mixed parameter

dependent on washout and on process associated hemolysis

only dependent on wash-out

Starting value

variable

consistently high (plasma)

Indicator for dangerous constituents

poor correlation

protein elimination representative for almost all soluble constituents

Method of determination complex

6,000 1,951

375 1,586 >6,000 1,866

*Blood loss < 300 ml / 4 h. **Blood loss 300–1,000 ml / 4 h.

Table 4. Comparison of drainage blood before and after filtration with three ADR systems [13]

consumed in the drainage blood. The occurrence of coagulation and fibrinolysis was proven by increased TAT complex and increased D-dimer concentrations (table 2). The quality of the blood depends not only on the site from which it is collected but also on the flow rate. With a blood flow of less than 300 ml in 4 h indicators of cell destruction – e.g. free hemoglobin and elastase – were twice as high compared with those found with a higher blood loss per time (table 3) [12]. Schulze et al. [12] therefore assumed that these enormous deflections of the system were due to more extended contact activation – within the organism and the collecting system – associated with the slower blood loss. The proponents of ADR, however, claim that unwanted substances are sufficiently removed by a filter system. We assessed three filter systems with regard to handling and effectiveness in patients undergoing total knee replacement [13]. Wound blood was examined before passage of the filter system and after passage of the ADR system. We found that hemoglobin and Hct were low and remained low in the wound blood. The load of activated platelets was not substantially reduced by the filter system. The same applies to free hemoglobin. Hemostasis is likewise substantially disturbed. The increased values for TAT complex indicated extensive activation of the coagulating system, and those for prothrombinantiplasmin complex reflected increased fibrinolysis. Fibrin degradation products also were not reduced by the filter system (table 4). Quality management as it is required for homologous transfusion and – increasingly – for intraoperative autotransfusion, is not possible with ADR.

Effects of the Re-Transfusion of Wound and Drainage Blood on the Patient Beyond the in vitro situation it is of utmost importance what re-transfusion does to the patient. This was addressed in a study on patients that received drainage blood after total hip replacement [14]. After transfusion of only 450 ml drainage blood – less than 10% of the blood volume – an increase of

Wound blood  C S H Hemoglobin, mmol/l Hct, % Thrombozyten, Gpt/l Free hemoglobin, mmol/l TAT complex, mg/l Prothrombin-antiplasmin complex, mg/l Fibrin degradation products, mg/ml

5.0 24 34 141 >300 >5,000 20

4.9 23 43 165 >300 >5,000 47

Filtered blood   C S H

6.0 29 65 242 >300 >5,000 80

5.5 25 40 143 >300 >5,000 23

4.7 25 44 106 >300 >5,000 47

5.4 25 57 141 >300 >5,000 80

C = Consta Vac® (Stryker, Kalamazoo, MI, USA); S = Solcotrans® (Smith & Nephew Richards, Memphis, TN, USA), H = Haem-o-Trans® (Haemotrans Blutspende GmbH, Leipzig, Germany)

234


Rosolski/Mauermann

Table 5. Comparison of drainage blood and patient blood 24 h after transfusion of 450 ml drainage blood [14]

Fibrinogen Fibrin degradation productsa, mg/ml D-dimerb, mg/ml After 5 days aNormal bNormal

Drainage blood

Patient blood

not measurable 49 ± 40 62 ± 65 –

normal 18 ± 21 9 ± 22 12 ± 17

value < 8 mg/ml value < 0.5mg/ml

operative collection of wound drainage blood and its re-transfusion did not significantly reduce the need for homologous blood transfusion [19]. This conclusion is confirmed by others [20, 21] The results of a meta-analysis of 6 studies on the effectiveness of re-transfusion of wound blood in patients undergoing cardiac surgery did not show any homologous blood-sparing effect. In a meta-analysis of 4 studies with 200 orthopedic patients, a transfusion-sparing effect was shown. However, this was interpreted very cautiously by the authors and somewhat put in question [22].

Autologous Direct Re-Transfusion and Cost fibrin degradation products was observed up to 5 days after ADR. The D-dimer concentration was still increased 24 h after re-transfusion of the wound blood, indicating severe and sustained derangement of the hemostatic system (table 5). Beyond that there is a number of case reports documenting complications of ADR [15–18].

The hope to reduce costs by routine use of ADR systems remains unfulfilled. If the same increase in hemoglobin concentration is to be achieved as obtained by one homologous transfusion, on the average 3.4 samples taken by ADR systems are necessary. The cost of the material alone amounts to the 5-fold of just one homologous unit [23].

Avoiding Homologous Blood by the Application of Autologous Direct Re-Transfusion?

Conclusion

The assumed effect of reducing the need for homologous blood transfusion is disproven by current studies. One prospective study in 161 orthopedic patients showed that post-

ADR is rarely indicated [24], is of low efficiency, supplies blood of poor quality, and causes high costs in routine application.

References 1 Highmore W: Practical remarks on an overlooked source of blood-supply for transfusion in post-partum haemorrhage. Lancet 1874:i:89–92. 2 Ducan J: On re-infusion of blood in primary and other amputation. Br Med J 1886;30:192–194. 3 Kasper SM, Kasper AS: Die Geschichte der autologen Bluttransfusion im 19. Jahrhundert. Zentralbl Chir 1996;121:250–257. 4 Dietrich GV: Identische Transfusionstrigger bei autologen und homologen Konserven? AINS 2002; 37:749–751. 5 Transmed Medizintechnik GmbH: Herstellerinformation. www.transmed-medizintechnik.de. 6 Hansen E: Retransfusion von autologem Blut. Erwiderung auf einen Leserbrief. Anaesthesist 2003; 52:169–172. 7 Wissenschaftlicher Beirat der Bundesärztekammer, Paul-Ehrlich Institut: Richtlinien zur Gewinnung von Blut und Blutbestandteilen und zur Anwendung von Blutprodukten (Hämotherapie). Köln, Deutscher Ärzteverlag, 2000. 8 Vorstand und Wissenschaftlicher Beirat der Bundesärztekammer: Leitlinien zur Therapie mit Blutkomponenten und Plasmaderivaten, 3. erweiterte Auflage. Köln, Deutscher Ärzteverlag, 2003. 9 Page P: Perioperative autotransfusion and its correlation to hemostasis and coagulopathies. J Extra Corpor Technol 1992;23:14–21.

Autologous Direct Re-Transfusion – Contra

10 Imhoff M, Schmidt R, Horsch S: Intraoperative autotransfusion with new disposable systems. Ann Vasc Surg 1986;23:131–133. 11 Singbartl G, Röhrs E, Hook D, Asmussen W, Schleinzer W, Tilsner V: Gerinnungsphysiologische Untersuchungen von nicht gewaschenem autologen Drainageblut. 6. Informationstagung über Eigenbluttransfusion. München, 1992, Abstraktband. 12 Schulze HJ, Wendel HD, Khalighi K, Helle W, Seboldt H: The quality of autotransfused chestdrainage blood after cardiac surgery: A study of coagulation. Thorac Cardiovasc Surg 1996;44:183– 187. 13 Rosolski T, Mauermann K, Frick U, Hergert M: Direkte Autotransfusionssysteme liefern Blut unzureichender Qualität. AINS 2000;35:21–24. 14 Wixson RL, Kwaan HC, Spies SM, Zimmer AM: Reinfusion of postoperative wound drainage in total joint arthroplasty. J Arthroplasty 1994;9:351– 358. 15 Azzopardi, N Yarbi M: A fatal reaction following scavenged autologous blood transfusion. Anaesth Intensive Care 1993;21:335–336. 16 Ouriel K, Shortell CK, Green RM, DeWesse JA: Intraoperative autotransfusion in aortic surgery. J Vasc Surg 1993;18:16–22. 17 Milne AA, Drummond GB, Paterson, DA, Murphy WG, Ruckley CV: Disseminated intravascular coagulation after aortic aneurysm repair, intraoperative salvage autotransfusion and aprotinin. Lancet 1994;344:470–471.

18 Woda R, Tetzlaff JE: Upper airway oedema following autologous blood transfusion from wound drainage system. Can J Anaesth 1992;39:290–292. 19 Bellavita P, Rizzi L, Celega E, Poma R, Pezzoli F: Recupero postoperatorio di sangue chirurgia protesica dànca e di ginocchio. Transfus Sang 1998;43:86–92. 20 Rosencher N, Vassilieff V, Tallet F, Toulon O, Leoni J, Tomeno B, Conseiler C: Comparison of OrthEvac ans Solcotrans Plus devices for the autotransfusion of blood drained after total knee joint arthroplasty. Ann Fr Anesth Reanim 1994;13: 318– 325. 21 Adalberth G, Bystrom S, Kolstad K, Mallmin H, Milbrink J: Postoperative drainage of knee arthroplasty is not necessary: a randomised study of 90 patients. Acta Orthop Scand 1998;69:475–478. 22 Huet C, Salmi LR, Fergusson D, Koopman-van Gemert AWM, Laupacis A: A meta-analysis of the effectiveness of cell salvage to minimize perioperative allogenic blood transfusion in cardiac and orthopedic surgery. Anesth Analg 1999;89:861–869. 23 Rizzi L, Bertacchi P, Ghezzi LM, Bellavita P, Scudeller G: Postoperative blood salvage in hip and knee arthroplasty. A prospective study on cost effectiveness in 161 patients. Acta Orthop Scand 1998;69:31–34. 24 Pitsaer E: Transfusion of recuperated blood in total knee arthroplasty (in French). Rev Chir Orthop Reparatrice Appar Mot 2002;88:777–789


235



Dilutional Coagulopathy, an Underestimated Problem?* D. Friesa aDepartment bDepartment

W Streifb

T. Haasa

G. Kühbachera

of Anaesthesiology and Critical Care Medicine, of Pediatrics, Innsbruck University Hospital, Austria

Key Words Crystalloids ⋅ Colloids ⋅ Dilutional coagulopathy ⋅ Thrombelastography

Schlüsselwörter Kristalloide ⋅ Kolloide ⋅ Dilutionskoagulopathie ⋅ Thrombelastographie

Summary When no fresh frozen plasma is available, acute major blood loss is compensated above all with crystalloids, colloids, and erythrocyte concentrates, meaning that all plasma clotting factors are diluted. Consumption coagulopathy is almost always accompanied by dilutional coagulopathy. Formulas for calculating critical blood loss and standard coagulation tests are often not helpful in the case of massive transfusion. On the other hand, systems suitable for point of care, such as thrombelastography, have important advantages. In the case of consumption and dilutional coagulopathy plasma coagulation is disturbed, and critical values are first seen for fibrinogen. Fibrin polymerization is not only impaired by the bleeding-induced loss and dilution of fibrinogen but also by interaction with artificial colloids, particularly hydroxyethyl starch preparations. Therapy of consumption and dilutional coagulopathy calls for fresh frozen plasma. If this is not available in sufficient quantity or within a reasonable time, coagulation factor concentrates must be used. Neither fresh frozen plasma therapy nor treatment with coagulation factor concentrates has been the subject of detailed clinical study. Further studies are needed to work out guidelines for coagulation management in the case of massive blood loss.

Zusammenfassung Große Blutverluste werden, solange keine Frischplasmen zur Verfügung stehen, akut in erster Linie durch Kristalloide, Kolloide und Erythrozytenkonzentrate ersetzt, was eine Verdünnung aller plasmatischer Gerinnungsfaktoren zur Folge hat. Eine blutungsbedingte Verlustkoagulopathie geht daher nahezu immer mit einer Verdünnungskoagulopathie einher. Formeln zur Berechnung kritischer Blutverluste sowie Standardgerinnungstests sind im Falle einer Massivtransfusion oft nicht hilfreich. Demgegenüber haben «Point-of-care»-taugliche Systeme wie die Thrombelastographie bedeutende Vorteile. Im Rahmen der Verlust- und Dilutionskoagulopathie ist zunächst die plasmatische Gerinnung gestört, wobei Fibrinogen als erstes kritische Werte zu erreichen scheint. Dabei wird die Fibrinpolymerisation nicht nur durch den blutungsbedingten Fibrinogenverlust und Verdünnung beeinträchtigt, sondern auch durch die Interaktion mit künstlichen Kolloiden, insbesondere mit Hydroxyethylstärkepräparaten. Die Therapie der Verlust- und Verdünnungskoagulopathie erfolgt mit Frischplasmen. Sind diese nicht in ausreichender Zahl oder in einem akzeptablen Zeitraum verfügbar, muss auf Faktorenkonzentrate zurückgegriffen werden. Weder die Therapie mit Frischplasmen noch die Behandlung mit Gerinnungsfaktorenkonzentraten wurde in ausreichendem Maße klinisch untersucht. Weitere Studien sind notwendig, um Richtlinien zu erarbeiten, wie das Gerinnungsmanagement im Rahmen großer Blutverluste zu erfolgen hat.




Dr. med. Dietmar Fries Klinik für Anaesthesie und Allgemeine Intensivmedizin Universität Innsbruck Anichstraße 35, A-6020 Innsbruck Tel. +43 512 504-804 55, Fax -27 49 E-mail [email protected]

Introduction Coagulopathy developing in major surgical interventions or in polytrauma patients with massive blood loss is of multifactorial origin: Hypothermia, metabolic disturbances, and hyperfibrinolysis can influence hemostasis, but are not a regular occurrence [1–6]. A combination of consumption and dilutional coagulopathy almost always results in an enhanced tendency to bleeding [7–10]. In the initial phase, as long, as no fresh frozen plasma is available, the blood volume loss is compensated with crystalloids, colloids, and erythrocyte concentrates, which necessarily dilutes all plasma clotting factors. Thus, consumption coagulopathy is almost always accompanied by dilutional coagulopathy. The extent of coagulopathy depends on the quantity and dynamics of blood loss, the quantity and type of volume therapy administered as well as the original concentration of hemostatic factors. If the plasma levels of procoagulatory coagulation factors, thrombocytes, and erythrocytes drop below the critical threshold, adequate hemostasis is no longer ensured. This provokes a tendency to diffuse bleeding, including mucosal bleeding, bleeding next to intravascular catheters, etc., without corresponding surgical bleeding [11]. Several formulas can be used to calculate critical blood loss which can lead to consumption or dilutional coagulopathy (for example, loss of total blood volume in 4 h or of twice the total blood volume in 24 h); these formulas appear very questionable with regard to their clinical suitability, were often developed in the whole blood era, and were never subject to clinical testing. For this reason reliable and immediate monitoring of the actual coagulation status is necessary. Because of the delay of up to 30 min and more standard laboratory analyses are often useless [12]. In this situation a valid point-of-care system is of great practical benefit [13].

Dilutional Coagulopathy: Thrombelastography vs. Standard Coagulation Tests Thrombelastography measures clot formation not only by the time to coagulation, as for example in standard photooptical laboratory techniques, but also functionally. Clot strength is continuously measured in whole blood and consists of coagulation activation, thrombin formation, fibrin formation and polymerization, thrombocyte activation, and thrombocyte-fibrin interaction [14–16]. Thrombelastometry (ROTEM®; Pentapharm, Munich, Germany), a modification of the customary thrombelastography after Hartert that is suitable for point of care, affords new test approaches, permitting differential diagnosis of factor deficiency, heparin effect, hyperfibrinolysis, fibrinogen deficiency as well as thrombocytopenia and thrombocytopathy. The extent of fibrin polymerization disturbance can be reliably measured, for example by addition of a thrombocyte-blocking reagent (FibTEM®, Pentapharm). Activators can be used to drastically reduce measuring time for the

238


ROTEM system [17, 18]. Reliable information on coagulation status is already available after 10 min. ROTEM measurements are performed in whole blood. This also shows the current influence of hematocrit in the analysis, which is not the case for standard plasma tests. Sufficient hematocrit is not only of major importance for tissue oxygenation but also for coagulation. Valeri et al. [19] showed that a decrease of only 15% in hematocrit is associated with a 60% prolonged bleeding time. In addition to rheology, the coagulation-active effect of erythrocytes might also be due to enhanced ADP release, stimulation of thromboxane A2 synthesis as well as enhanced thrombin generation by red blood cells [20–23]. At a hematocrit level of less than 20% critical impairment of hemostasis can be expected, with concomitant thrombocytopenia having a stronger effect on the coagulatory system [24]. An additional disadvantage of standard coagulation tests as compared to thrombelastography is that the administration of synthetic colloids as volume therapy distorts the results of optical coagulation tests (prothrombin time (PT), activated partial thromboplastin time (aPTT), fibrinogen as described by Klaus). Colloids cause plasma to turn cloudy, which in turn results in an erroneous shortening of the measured coagulation times as well as falsely high values when determining the fibrinogen concentration after Klaus [25]. In this respect functional fibrinogen measurement by thrombelastography, as already mentioned, might be advantageous [26]. Volume Therapy and Its Effect on Coagulation The question as to the optimal volume replacement for compensating intravascular blood loss is the subject of ongoing controversy. Ever since the results of the Cochrane Injuries Group Albumin Reviewers [27] were published, the use of albumin has markedly declined. Dextrans are also hardly used anymore for volume therapy. Because of their extremely negative effects on the coagulatory system they should be used as anticoagulants, if at all, and not for volume therapy [28]. The question as to whether colloid infusions are generally superior to crystalloids has also not been conclusively answered. A meta-analysis by Schierhout and Roberts [29] analyzed 26 studies with 1,622 patients. The authors came to the conclusion that the administration of colloids is associated with enhanced mortality. Another meta-analysis by Choi et al. [30] reviewed 17 studies with 814 patients. Only the subgroup of polytrauma patients showed a lower rate of mortality with crystalloids as compared to colloids. Whether or not these meta-analyses are conclusive for European conditions is questionable. The studies involved were primarily American studies and thus employed dextrans and hetastarches (high molecular hydroxyethyl starch (HES) preparations with a high substitution degree), which have not been used for a long time in Central Europe because of their high potential for side effects, particularly involving the coagulatory system.

Fries/Streif/Haas/Kühbacher

Crystalloids impair the coagulatory system primarily by means of their dilutional effect. Several studies surprisingly postulated that low doses of crystalloids as well as colloids cause hypercoagulability. These were primarily in vitro studies using nonactivated thromeleastography measurements. Despite the shortened clot formation times and enhanced clot strengths seen for thrombeleastography, no change in the activated coagulation markers (thrombin-antithrombin complex) was observed [32–35]. It is possible that the measured shortened clot formation times and enhanced clot strengths are the result of an in vitro effect produced by the influence of the sedimentation of red blood cells in diluted specimens subject to long measuring time [36]. Measuring with activated thrombeleastography and markedly shortened measuring time was, however, not able to confirm this phenomenon [37, 38]. Moreover, Petroianu et al. [39] concluded that hemodilution with crystalloids and colloids caused a decrease in the activity of various clotting factors in vitro. Thus, why hemodilution should activate coagulation while activated coagulation markers remain unchanged and the activity of various coagulation factors and the thrombocyte count decrease, is not known. In addition to their dilutional effect gelatin preparations also exert specific effects on the coagulatory system. Above all, they impair fibrin polymerization and disturb the network of the fibrin monomers. Moreover, reduced clot elasticity and clot weight have been reported for gelatin replacement [40, 41]. HES has been reported to be associated with an increased tendency to bleed, above all when using solutions with a high molecular weight and a high replacement degree [42]. HES solutions cause a von Willebrand type 1-like syndrome characterized by diminished factor VIII activity and diminished von Willebrand factor (vWF) plasma levels [43]. In addition, HES also impairs fibrin polymerization [38]. A relatively new preparation is 6% HES 130/0.4 (Voluven®; Fresenius Pharma Austria GmbH, Graz, Austria); it has a medium molecular weight and a low substitution degree and thus probably does not as strongly affect the coagulatory system [44, 45].

Dilutional Coagulopathy – a Plasma or Cellular Coagulation Problem? Transfusion practice has changed vastly over the last 20 years. While formerly whole blood was administered to treat anemia, only erythrocyte concentrates are used today. In contrast to whole blood, erythrocyte concentrates contain only approximately 20 ml plasma per concentrate. Previously, thrombocytopenia was primarily responsible for coagulopathy in the framework of massive transfusions. Today, as long as exclusively crystalloids, colloids, and erythrocyte concentrates are administered, a coagulation factor deficiency is at least initially responsible for increased bleeding tendency [11, 46–50].

Dilutional Coagulopathy, an Underestimated Problem?

McLoughlin et al. [10] investigated the consequences of a considerable normovolemic hemodilution in vivo as well as in an animal model. Despite the fact that 75% of the total blood volume was replaced in 8 patients during major spinal surgery, no clinically relevant thrombocytopenia was detected. While the thrombocyte count was still >150 × 103/µl, plasma coagulation was drastically impaired. Not only did coagulation times (PT, PTT) double, but also a clinically relevant hypofibrinogenemia (50 mg/dl) was measured. In the animal model the authors gradually diluted pigs to death. Here, too, plasma coagulopathy occurred first, followed by clinically relevant thrombocytopenia [10]. In 60 patients Hiippala et al. [51] studied the effects of massive bleeding during surgical intervention on plasma and thrombocyte coagulation, which were compensated with colloids and erythrocyte concentrates, using regression analysis for this purpose. The authors came to the conclusion that hypofibrinogenemia (8

Fig. 2. Ca2+ and phospholipids (INTEM) or tissue factor (EXTEM) are used to activate the coagulation process in the tested blood sample which is placed in the cup. Initiation of coagulation, measured as clotting time (CT, s), depends on concentrations of coagulation factors/inhibitors and shows initial thrombin and fibrin formation. Propagation of clot formation follows when a sufficient thrombin burst has been built up. This propagation depends on concentrations of coagulation factors/inhibitors and also of fibrinogen. It is measured as clot formation time (CFT, s) and is defined as the time needed to reach a clot firmness of 20 mm. The α angle describes the kinetics of clot formation. The final clot strength results from firm aggregation of platelets and formation of a stable fibrin network. Thus, measurement of maximum clot firmness (MCF, mm) depends on counts and function of platelets and fibrinogen concentrations as well as on concentrations of coagulation factor XIII [28, 30].

CT = Clotting time; CFT = clot formation time; MCF = maximum clot firmness.

Thrombelastography Until now, the only available measurement technique providing all of the above-mentioned features is thrombelastography (TEG) which is gaining increased acceptance for perioperative monitoring of hemostasis, but is also criticized as an invalidated and costly method [19, 20]. However, several recent studies have shown that TEG is sensitive in monitoring hemostasis, even in pediatric patients, can predict bleeding and reduce transfusion requirements [21–24]. Moreover, the reproducibility of thrombelastographic measurements has been shown to be superior to that of activated coagulation time (ACT) and improves with the use of activated tests [25–27]. TEG technique monitors initiation, propagation, and resulting quality of clot formation in whole blood. This technique looks behind the endpoints of initial fibrin formation measured by PT and aPTT and also assesses the kinetics of clot formation as well as the stability or lysis of the formed clot [28]. Initially described by Hartert as early as 1948 as a sophisticated method prone to disturbance of measurement by movement and vibration, it largely remained a research method used only in specialized laboratories [29]. However, over the past years several modifications have been introduced. Instruments are now easily transportable, automatic pipetting sys-

246


tems are used, and computerized analysis and storage of results are provided. One of these newly developed devices is the ROTEM analyzer (rotational thrombelastometry; Pentapharm, Munich, Germany; formerly used abbreviation ROTEG) which was put into clinical practice at our institution 5 years ago and allows 4 tests to be performed in parallel.

ROTEM Measurement Technique Principally whole blood with or without an activating reagent (partial thromboplastin and tissue thromboplastin for monitoring the intrinsic and extrinsic activation of blood coagulation) and with Ca2+ (for citrated blood samples) is placed in a disposable cup using the automatic pipetting system. With ongoing coagulation the mechanical coupling between the surface of the pin and cup is continuously displayed on the computer (fig. 1). A typical ROTEM tracing and its interpretation is shown in figure 2. Normal values of measurements (obtained from healthy volunteers) are listed in table 1. Physiologically, clot firmness (maximum clot firmness, MCF) might show some decrease ( 45 min. Presence of clinically relevant fibrinolysis can be detected as increased maximum lysis (ML, >15% of MCF), early-starting reduction in MCF ( 10 mm and thus normal. A intraoperative decrease in MCF < 40 mm is associated with diminished hemostatic competence and increased bleeding tendency while intraoperative MCF measurements < 35 mm are associated with profuse severe bleeding. These


247

Fig. 4. Measurements were obtained from a 16-year-old girl (body weight 28 kg, calculated blood volume 1,850 ml) undergoing scoliosis repair. a Measurement A was obtained after induction of anesthesia for ventral mobilization. Although preoperative global coagulation tests and fibrinogen concentrations were normal, no blood loss had occurred by that time and only 200 ml of crystalloid solution had been administered, MCF readings were borderline. Subsequent measurements of coagulation factor concentrations revealed that FXIII was only 73% (normal range 70–140%), thus probably explaining the slight reduction in MCF. b Measurements obtained 10 days after ventral mobilization for dorsal stabilization. In contrast, MCF readings and fibrinogen concentrations were much higher than 10 days before. Therefore, supplementation with FXIII was initially avoided. c ROTEM tracing after 1,800 ml blood had been lost (about 100% of calculated blood volume). Although the initial fibrinogen concentration was high (590 mg/dl) at that blood loss, the fibrinogen concentration had already dropped to critical values (FIBTEM MCF 8 mm, fibrinogen 118 mg/dl).

INTEM

EXTEM

a

FIBTEM

b

c

ROTEM thresholds are derived from our practice, using the ROTEM analyzer during the past 5 years in nearly 1,000 patients who underwent extensive surgery or who were examined to explain unexpected or even therapy-resistant bleeding. We have already published our results on dilutional coagulopathy derived from an in vitro study [35] and a study in patients exhibiting moderate blood loss and volume demand during knee replacement surgery [7]. In latter study ROTEM tracings were least diminished when crystalloid fluids were ad-

ministered exclusively. As expected, none of those patients needed hemostatic treatment. In order to answer still open questions on the benefit of ROTEM measurements and their correlation to ‘gold standard tests’, we are now conducting a study in patients undergoing spinal surgery and exhibiting larger blood loss and volume demand. The results of this study should be encouraging for the obvious clinical importance of intraoperative thrombelastographic monitoring.

References 1 American Society of Anesthesiologists Task Force on Blood Component Therapy: Practice guidelines for blood component therapy. Anesthesiology 1996; 84:732–747. 2 Horsey PJ: Multiple trauma and massive transfusion (Editorial). Anaesthesia 1997;52:1027–1029. 3 Hellstern P, Haubelt H: Indications for plasma transfusion. Thromb Res 2002;107:19–22. 4 Hippala S: Replacement of massive blood loss. Vox Sang 1998;74(suppl 2):399–407. 5 Valeri CR, Cassidy G, Pivacek LE, Ragno G, Lieberthal W, Crowley JP, Khuri SF, Loscalzo J: Anemia-induced increase in bleeding time: Implications for treatment of nonsurgical blood loss. Transfusion 2001;41:977–983. 6 De Jonge E, Levi M: Effects of different plasma substitutes on blood coagulation: A comparative review. Crit Care Med 2001;29:1261–1267. 7 Innerhofer P, Fries D, Margreiter J, Klingler A, Kuhbacher G, Wachter B, Oswald E, Salner E, Frischhut B, Schobersberger W: The effects of perioperatively administered colloids and crystalloids on primary platelet-mediated hemostasis and clot formation. Anesth Analg 2002;95:858–865.

248

8 Reiss RF: Hemostatic defects in massive transfusion: Rapid diagnosis and management. Am J Crit Care 2000;9:158–165. 9 Hirshberg A, Dugas M, Banez EI, Scott BG, Wall MJ Jr, Mattox KL: Minimizing dilutional coagulopathy in exsanguinating hemorrhage: A computer simulation. J Trauma 2003;54:454–463. 10 Eckman MH, Erban JK, Singh SK, Kao GS: Screening for the risk for bleeding or thrombosis. Ann Int Med 2003;138:15–24. 11 Murray D, Pennell B, Olson J: Variability of prothrombin time and activated thromboplastin time in diagnosis of increased surgical bleeding. Transfusion 1999;39:56–62. 12 Singbartl K, Innerhofer P, Radvan J, Westphalen B, Fries D, Stogbauer R, Van Aken H: Hemostasis and hemodilution: A mathematical guide for clinical practice. Anesth Analg 2003;96:929–935. 13 Hippala ST: Dextran and hydroxyethyl starch interfere with fibrinogen assays. Blood Coag Fibrinol 1995;6:743–746. 14 MacLeod JB, Lynn M, McKenney MG, Cohn SM, Murtha M: Early coagulopathy predicts mortality in trauma. J Trauma 2003;55:39–44.


15 Stein SC, Young GS, Talucci RC, Greenbaum BH, Ross SE: Delayed brain injury after head trauma: Significance of coagulopathy. Neurosurgery 1992; 30:160–165. 16 Malone DL, Dunne J, Tracy JK, Putnam AT, Scalea TM, Napolitano LM: Blood transfusion, independent of shock severity, is associated with worse outcome in trauma. J Trauma 2003;54:898–907. 17 Lucas CE, Ledgerwood AM, Saxe JM, Dombi G, Lucas WF: Plasma supplementation is beneficial for coagulation during severe hemorrhagic shock. Am J Surg 1996;171:399–404. 18 Chau TN, Chan YW, Patch D, Tokunaga S, Greenslade L, Burroughs AK: Thrombelastographic changes and early rebleeding in cirrhotic patients with variceal bleeding. Gut 1998;43:267–271. 19 Samama.CM: Thrombelastography: The next step (Editorial). Anest Analg 2001;92:563–564. 20 Pivalizza EG, Abramson DC: Thrombelastography: Another point of view (Letter). Anest Analg 2001; 93:517–518. 21 Miller BE, Mochizuki T, Levy JH, Bailey JM, Tosone SR, Tam VK, Kanter KR: Predicting and treating coagulopathies after cardiopulmonary bypass in children. Anest Analg 1997;85:1196–1202.

Innerhofer/Streif/Kühbacher/Fries

22 Shore-Lesserson L, Manspeizer HE, DePerio M, Francis S, Vela-Cantos F, Ergin MA: Thrombelastography-guided transfusion algorithm reduces transfusions in complex cardiac surgery. Anesth Analg 1999;88:312–319. 23 Royston D, von Kier S: Reduced haemostatic factor transfusion using heparinase-modified thrombelastography during cardiopulmonary bypass. Br J Anaesth 2001;86:575–578. 24 Cammerer U, Dietrich W, Rampf T, Brain SL, Richter JA: The predictive value of modified computerized thrombelastography and platelet function analysis for postoperative blood loss in routine cardiac surgery. Anesth Analg 2003;96:51–57. 25 Forestier F, Belisle S, Contant C, Harel F, Janvier G, Hardy JF: Reproductibilité et interchangeabilité du Thromboélastographe®, Sonoclot® et du temps de coagulation activé (Hémochron®), en chirurgie cardiaque. Can J Anesth 2001;48:902–910. 26 Vig S, Chitolie A, Bevan DH, Halliday A, Dormandy J: Thrombelastography: A reliable test? Blood Coag Fibrinol 2001;12:555–561. 27 Miller BE, Guzzetta NA, Tosone SR, Levy JH: Rapid evaluation of coagulopathies after cardiopulmonary bypass in children using modified thrombelastography. Anesth Analg 2000;90:1324–1330. 28 Mallett SV, Cox DJ: Thrombelastography. Br J Anaesth 1992;69:307–313. 29 Hartert H: Blutgerinnungsstudien mit der Thrombelastographie, einem neuen Untersuchungsverfahren. Klein Wochenschr 1948;26:577–583. 30 Chandler WL, Patel MA, Gravelle L, Soltow LO, Lewis K, Bishop PD, Spiess BD: Factor XIIIA and clot strength after cardiopulmonary bypass. Blood Coag Fibrinol 2001;12:101–108. 31 Vorweg M, Hartmann B, Knüttgen D, Jahn MC, Doehn M: Management of fulminant fibrinolysis during abdominal aortic surgery. J Cardiothorac Vasc Anesth 2001;6:764–767. 32 Gottumukkala VN, Sharma SK, Philip J: Assessing platelet and fibrinogen contribution to clot strength using modified thrombelastography in pregnant women. Anesth Analg 1999;89:1453–1455. 33 Ti LK, Cheong KF, Chen FG: Prediction of excessive bleeding after coronary artery bypass graft surgery: The influence of timing and heparinase on thrombelastography. J Cardiothorac Vasc Anesth 2002;16:545–550. 34 Frumento RJ, Hirsh AL, Parides MK, BennettGuerrero E: Differences in arterial and venous thrombelastography parameters: Potential role of shear stress and oxygen content. J Cardiothorac Vasc Anesth 2002;16:551–554. 35 Fries D, Innerhofer P, Klingler A, Berresheim U, Mittermayr M, Calatzis A, Schobersberger W: The effect of the combined administration of colloids and lactated Ringer’s solution on the coagulation system: An in vitro study using thrombelastograph coagulation analysis. Anesth Analg 2002;94:1280– 1287.

Monitoring of Perioperative Dilutional Coagulopathy Using the Rotem Analyzer


249



Do We Have Enough Evidence to Know when to Transfuse Erythrocytes? R.B. Weiskopf Departments of Anesthesia and Physiology, and Cardiovascular Research Institute, University of California, San Francisco, CA, USA

Key Words Perioperative period ⋅ Oxygen delivery (primary measure, surrogate markers)

Schlüsselwörter Perioperative Phase ⋅ Sauerstoffversorgung (Primäre Messgröße, Surrogatmarker)

Summary This brief article reviews the evidence regarding the oxygen delivery / oxygen carrying capacity / hemoglobin concentration at which erythrocytes should be transfused in the perioperative period. The outdated threshold of a hemoglobin concentration (Hb) of 10 g/dl or an hematocrit of 30% has been replaced by recommendations of many practice guidelines, based on clinical experience and data. Generally, in humans, these data address surrogate endpoints, rather than the primary ones of intracellular oxygenation, because of the inability to assess the latter in clinical circumstances. In conscious, healthy, resting humans, oxygen delivery does not have a maximum at Hb 10 g/dl, but remains unchanged until Hb falls to 6–7 g/dl. Even at an Hb of 5 g/dl with an oxygen delivery of 10.7 ml O2/kg/min, systemic markers of inadequate oxygen delivery do not indicate inadequate oxygenation. Anesthetized patients do not show altered oxygen consumption in response to acute reduction of Hb to 8 g/dl. However, those who are not able to increase their oxygen delivery in response to anemia should have a higher critical oxygen delivery than do normal humans. Patients aged 65 years or more and even those with coronary artery disease sufficiently severe to require revascularization surgery do not have decreased oxygen consumption when their Hb is reduced to 9 or 10 g/dl when they are anesthetized. Patients in an intensive care unit do not have a higher mortality when transfused to maintain an Hb of 8.5 g/dl rather than 10.7 g/dl. Assessment of critical organ oxygenation in humans has been difficult. Evaluation of transmyocardial lactate flux or maximal exercise capability has not demonstrated the critical value for the heart in elderly patients or in those with coronary artery disease. Maximal exercise capacity, however, would seem to be a poor surrogate for this purpose. Evaluation of mental function suggests that in healthy, conscious humans, inadequate cerebral oxygenation occurs at an Hb of 5–6 g/dl; however, this has not been evaluated in patients in the perioperative period. Until more specific methods and technology applicable to patients are developed to determine whether a specific patient requires augmentation of oxygen delivery, transfusion therapy in the perioperative period will most often be determined by clinical judgment.

Zusammenfassung Dieser kurze Beitrag beschäftigt sich mit der Evidenz bezüglich der Sauerstoffversorgung, Sauerstofftransportkapazität und Hämoglobinkonzentration (Hb), bei der in der perioperativen Phase Erythrozyten transfundiert werden sollten. Die nicht mehr aktuellen Grenzwerte einer Hb von 10 g/dl oder eines Hämatokrits von 30% wurden in vielen praxisorientierten Richtlinien durch Empfehlungen ersetzt, die auf klinischen Erfahrungen und Daten beruhen. Grundsätzlich ist jedoch für den Menschen festzuhalten, dass diese Daten sich nur auf Surrogat-Endpunkte beziehen, und nicht primär auf die intrazelluläre Oxygenierung, weil diese unter klinischen Umständen nicht messbar ist. Beim wachen, gesunden, ruhenden Menschen hat das Sauerstoffangebot sein Maximum nicht bei einem Hb von 10 g/dl, bleibt aber unverändert, bis der Hb auf 6–7 g/dl abfällt. Selbst bei einem Hb von 5 g/dl mit einer Sauerstoffabgabe von 10,7 ml O2/kg/min weisen systemische Zeichen der inadäquaten Sauerstoffversorgung nicht auf eine inadäquate Oxygenierung hin. Anästhesierte Patienten zeigen keinen veränderten Sauerstoffverbrauch bei einer akuten Reduktion der Hb auf 8 g/dl. Trotzdem sollten die Patienten, die nicht in der Lage sind, ihre Sauerstoffabgabe als Antwort auf die Anämie zu erhöhen, eine höhere kritische Sauerstoffversorgung haben als normale Menschen. Anästhesierte Patienten, die 65 Jahre oder älter sind, und besonders jene mit einer Koronarerkrankung, die einer operativen Revaskularisierung bedürfen, haben keinen verringerten Sauerstoffverbrauch, wenn ihre Hb auf 9 bzw. 10 g/dl reduziert ist. Die Untersuchung der kritischen Organoxygenierung beim Menschen war und ist schwierig. Die Bestimmungen des transmyokardialen Laktatflusses oder der maximalen körperlichen Belastbarkeit haben bei älteren Patienten oder solchen mit Koronarerkrankung für das Herz keinen kritischen Wert gezeigt. Maximale körperliche Belastbarkeit wäre auch eine schlechter Surrogatmarker für diesen Zweck. Die Untersuchung mentaler Funktionen legt nahe, dass beim gesunden, wachen Menschen eine inadäquate Oxygenierung bei einem Hb von 5–6 g/dl eintritt. Dies wurde – wohlgemerkt – aber nicht bei Patienten in der perioperativen Phase untersucht. Bis spezifischere Methoden und bei Patienten anwendbare Technologien entwickelt worden sind, um zu bestimmen, ob die Sauerstoffversorgung bei einem spezifischen Patienten verbessert werden muss, hängt die Transfusionstherapie in der perioperativen Phase meistens von der klinischen Einschätzung ab.



Richard B. Weiskopf, M.D. Department of Anesthesia University of California, San Francisco 521 Parnassus Avenue, San Francisco, CA 94143–0648, USA Tel. +1 415-476 21 32, Fax -502 21 32 E-mail [email protected]

There are several unresolved clinical issues regarding the transfusion of blood. Among these are: – Are the immunomodulatory effects of transfusion clinically important? – Do the storage lesions affecting red cells (i.e. decreased membrane deformability and decreased 2,3-DPG concentration) have clinically important effects on erythrocyte off-loading of oxygen? – What is the oxygen-carrying capacity (hemoglobin concentration) at which erythrocytes should be transfused? This brief review examines the latter issue with respect to perioperative transfusion to treat acute anemia. In 1942 Adams and Lundy [1] presented their personal view that the minimum preoperative hemoglobin concentration should be 10 g/dl (100 g/l) or a hematocrit of 30%. This led to the commonly applied clinical ‘10/30 rule’. Although it was known for many years that blood was capable of transmitting many pathogens, it was not until the recognition that the human immunodeficiency virus could also be transmitted by transfusion that the 10/30 rule began to be questioned seriously. A consensus conference sponsored by the U.S. National Institutes of Health in 1988 concluded that there had been no scientific basis for selection of a transfusion trigger of 10 g/dl, and suggested that a perioperative hemoglobin concentration of 7 g/dl was satisfactory for healthy, normovolemic patients [2]. Unfortunately, there were few data obtained from humans at that time to support that conclusion. A few years later, a task force of the American Society of Anesthesiologists (ASA) suggested that healthy humans could tolerate a hemoglobin concentration of 6 g/dl [3]. More than 30 transfusion guidelines have been produced by various professional medical organizations. The general thought expressed in these guidelines is that red cells should be transfused to augment the oxygen-carrying capacity of an anemic patient. More correctly, it is oxygen delivery that is of concern: Augmentation should be accomplished to either treat or prevent inadequate oxygenation. In the 8 years since the ASA practice guidelines have been published, additional data have been published regarding the physiology and effects of acute anemia, and the results of trials of differing transfusion strategies, thus raising the question of whether sufficient data now exist to establish more firm guidelines. This brief review explores that issue, focusing on data obtained in humans. The point at which oxygen delivery falls below that which is necessary to meet oxygen demand is termed ‘critical oxygen delivery’. Ideally, to assess the adequacy of oxygen delivery, data regarding the intracellular environment should be evaluated. The primary ‘endpoints’ would be intracellular PO2, high energy phosphate concentrations, and the redox state. Surrogate measures could include intracellular hydrogen ion and lactate concentrations. However, surrogate measures do not have perfect sensitivity and specificity, making conclusions drawn from their assessment less certain and, thus, less valuable than con-

252


clusions based on primary endpoints. While these primary and surrogate measures can be evaluated in laboratory experiments, and at times in humans, in general they are not available as standard clinical tools. Thus, intracellular assessment of inadequate oxygenation generally is not possible in the usual clinical setting. Consequently, many studies have assessed the adequacy of whole body oxygen delivery. Here, too, one may evaluate the primary measure, oxygen consumption, or a number of surrogates such as blood lactate concentration or blood base deficit. Cain [4] ‘validated’ the relationship between blood lactate concentration and oxygen consumption, thus demonstrating the former to be a good surrogate marker for systemic evidence of inadequate oxygenation. In doing so, in several separate studies, Cain [4, 5] repeatedly established that the critical oxygen delivery in anesthetized dogs is approximately 10 ml O2/kg/min, which occurred in those dogs at an hematocrit of 10–15%. Von Restorff et al. [6] were unable to find systemic evidence of inadequate oxygen delivery in awake dogs at similar hematocrits, likely because they did not study more severe anemia. Furthermore, the value of critical oxygen delivery varies among species [7], making extrapolation of these data to humans not possible. Thus, this review focuses on data derived from humans. In conscious healthy humans aged 19–69 years, acute isovolemic anemia increases heart rate, stroke volume, and cardiac output [8]. Oxygen delivery does not have a maximum at a hemoglobin concentration of 10 g/dl, but remains unchanged until the hemoglobin concentration is reduced to approximately 6–7 g/dl, depending on age and sex [8]. At lower hemoglobin concentrations, oxygen delivery is decreased. However, even at a hemoglobin concentration of 5 g/dl with an oxygen delivery of 10.7 ml O2/kg/min, primary and surrogate systemic markers of inadequate oxygen delivery (oxygen consumption, blood lactate concentration, blood base excess) do not indicate that oxygen delivery has reached the critical level [8]. Even further reduction of DO2 at a hemoglobin concentration of 5 g/dl, to 7.3 ml O2/kg/min does not produce systemic evidence of inadequate oxygenation [9]. Lesser degrees of acute isovolemic anemia in humans also have not resulted in decreased oxygen consumption [10, 11]. Anesthetized patients, in response to acute anemia, do not increase their heart rate [10, 11], as do awake humans, perhaps owing to the effects of administered opioids [12]. Nevertheless, these humans, also do not show altered oxygen consumption in response to acute reduction of hemoglobin concentration to 8 g/dl [10, 11]. However, those who are not able to increase their oxygen delivery in response to anemia should have a higher critical oxygen delivery than do normal humans. Thus, investigators have attempted to determine the total body critical oxygen delivery in elderly patients and in those with diagnosed or presumed coronary artery disease. Patients aged 65 years or more [13] and even those with coronary artery disease sufficiently se-

Weiskopf

vere to require revascularization surgery do not have decreased oxygen consumption when their hemoglobin concentration is reduced to 9 or 10 g/dl, when they are anesthetized [14]. Similarly, patients undergoing vascular surgery [15] and those who have been stabilized after a period of traumatic hemorrhagic shock [16] do not have decreased oxygen consumption at a hemoglobin concentration of 10 g/dl. Other surrogate measures have been used in an effort to assess inadequate systemic oxygenation. Some studies have used mortality as the clinical endpoint. In a small prospective trial of 95 patients undergoing abdominal and peripheral vascular surgery, randomly allocated postoperatively to two different transfusion strategies, a hemoglobin concentration of 9.8 g/dl did not alter mortality compared to a hemoglobin concentration of 11 g/dl [15]. In a large (838 patients) multicenter, prospective, clinical trial, Hébert et al. [17] found no difference in mortality in patients randomly allocated, on admission to intensive care units in Canada, to two differing transfusion strategies that resulted in hemoglobin concentrations of 8.5 and 10.7 g/dl. Thus, neither study identified the hemoglobin concentration that increased the surrogate measure mortality. The finding of Hébert et al. [17] in their prospective study differed from that of their retrospective chart analysis of 4,470 patients in Canadian intensive care units, which had found a highly statistically significant (p < 0.0001) greater hemoglobin concentration in patients who survived compared to those who died [18]. Additionally, the overall mortality of patients in the retrospective analysis was substantially less than that in the subsequent randomized study. These conflicting data from the same investigators and the same clinical sites highlight the peril of attempting to draw conclusions from retrospective analyses regarding transfusion. There are many problems with retrospective analyses, not the least of which is that transfusions are based most often on clinical judgment, and the rationale cannot often be elicited from chart reviews. The failure to establish a systemic critical oxygen delivery in humans has led several investigators to examine organs thought to be most sensitive to oxygen delivery. Again, owing to the uncertainty of the ability to translate findings from other species to humans, this brief review examines the data resulting from studies in humans related to myocardial and brain oxygenation during acute anemia. As with assessment of the whole body, assessments of individual organs or tissues may use oxygen consumption as a primary measure or surrogate measures such as transorgan lactate flux or organ function. Studies assessing the human myocardial response to anemia have also used other surrogates such as electrocardiographic changes, wall motion abnormalities, morbid cardiac events, including death, and exercise capacity. These surrogates have varying, and for some, uncertain sensitivity and specificity relationships to inadequate myocardial oxygenation. Coronary arteries are capable of substantial dilation and increasing myocardial blood flow in response to acute anemia

[19]. The dilating capacity of canine coronary arteries is exhausted at an hematocrit of 10–15% [19], and the critical oxygen delivery to the myocardium occurs at this hematocrit in healthy dogs [20, 21]. The normoxic myocardium uses lactate as a substrate, and thus, extracts and consumes blood lactate. When hypoxic, the myocardium becomes a net producer of lactate. In dogs, decreasing myocardial blood flow by either creation of a stenosis [21] or reduction of perfusion pressure [22] alters the critical myocardial oxygen delivery as assessed by the hematocrit at which the myocardium produces lactate [21], or the hematocrit at which the segmental myocardial shortening decreases [22]. However, as with the case of whole body evaluation of inadequate oxygen delivery, this has been difficult to demonstrate in humans. In a retrospective analysis of transmyocardial lactate flux in anesthetized patients just prior to institution of cardiopulmonary bypass for coronary artery bypass grafting, no statistical relationship was found between myocardial lactate flux and hemoglobin concentration in a range of 8–16 g/dl [23]. However throughout this hemoglobin range the myocardium of many patients showed net lactate production, and no interventions were performed to determine the hemoglobin concentration at which lactate consumption became lactate production in any patient. In examining myocardial oxygen consumption and lactate flux in 8 patients after termination of cardiopulmonary bypass Mathru et al. [24] found that red cell transfusion-induced increase (not randomized) of hematocrit increased myocardial oxygen consumption, thus suggesting that the lowest hematocrit of 17% was inadequate to meet oxygen demand. However, the small sample size and identical order of treatment for each patient (lack of randomization) makes interpretation of their data problematic. Other investigators have used other surrogate measures. Catoire et al. [25] randomly allocated 20 patients undergoing intraabdominal aortic surgery to either an hematocrit of 37% or to a hemodilution-induced hematocrit of 27%. The lower hematocrit preserved myocardial function at the time of aortic cross-clamping, as assessed by myocardial wall motion scores, while the higher hematocrit did not. Using ECG ST changes during prostatectomy as their clinical marker, Hogue et al. [26] found that patients who had intraoperative ST changes during prostatectomy had a very small (1%) but statistically significant lower hematocrit (28 ± 3%; mean ± SD) than those who did not have ST changes (29 ± 4%). Clearly, there was a great deal of overlap of hematocrits among the two populations, the overall incidence of nearly 30% of ST changes in this population is surprising, and it is difficult to implicate a postoperative hematocrit as the cause of intraoperative ST changes. Several studies have used exercise as a surrogate marker for myocardial function. Johnson et al. [27] found no relationship between hematocrit (range 23–38%) and endurance times on a treadmill in patients 5 days after coronary artery bypass graft surgery. However, transfusion increased the endurance

Do We Have Enough Evidence to Know when to Transfuse Erythrocytes?


253

time of the group that had been randomly allocated to a more conservative transfusion strategy. A recent report of a retrospective chart review of nearly 6,000 patients who underwent repair of a fractured hip found a statistical relationship between the median distance the patients were able to walk at discharge, and their average hemoglobin concentration during hospitalization [28]. However, no information is available regarding how the exercise tests were conducted at the many centers, the relationship was not examined for the hemoglobin concentration at the time of the test, and the authors noted that the distance walked appeared to have been recorded in ‘round’ numbers. Furthermore, there does not appear to be a relationship between these two variables where one might expect the relationship to be most strong: at hemoglobin concentrations below 10 g/dl. The finding is contrary to the earlier smaller prospective study of this group in which there was no difference in discharge location (a surrogate for ability to exercise) of two prospectively randomized groups to two different transfusion strategies which resulted in mean hemoglobin concentrations of approximately 9.7 and 10.7 g/dl [29]. Thus, as in the case of intensive care unit mortality, results from a prospective study do not agree with the conclusions drawn from a retrospective chart review. Of substantial importance, however, is the question of whether maximal exercise capability is an appropriate surrogate measure of myocardial function. In normal, healthy, young males without heart disease, acute isovolemic anemia at rest does not alter oxygen consumption [8, 30]. However, during maximal exercise (using either maximal VO2 or endurance times as a measure), increasing hemoglobin concentration increases oxygen consumption and decreasing hemoglobin concentration decreases oxygen consumption [30–32]. Reducing hemoglobin concentration at maximal exercise decreases oxygen delivery, but not cardiac output, suggesting that myocardial function is not decreased [30]. However, during submaximal work in healthy humans, decreasing hemoglobin concentration from 15 to 10 g/dl does not alter the relationship between oxygen consumption and work level [30]. Thus, it would appear that submaximal work is a better surrogate than is maximal work for studies wishing to use myocardial function as an endpoint. Finally, with respect to myocardial effects of anemia in heart disease, several authors have examined the effects of anemia on mortality. Bracey et al. [33] randomly allocated 437 patients after coronary artery bypass graft surgery to one of two differing transfusion strategies. The mortality of the two groups did not differ. However the difference in hemoglobin concentration of the two groups was only 0.6 g/dl. Wu et al. [34] provided a retrospective analysis of patients with myocardial infarction in the Medicare data base, finding a statistical association of mortality with lesser admission hematocrit. However, 64% of the patients were excluded from the analysis; the analysis did not include the course of hematocrit during hospitalization; and lower admission hematocrit was also

254


associated with trauma, bleeding, recent surgery, shock, heart failure, lower blood pressure and higher heart rate, and lesser use of beta-adrenergic antagonists, aspirin and reperfusion therapy. This strongly suggests that those with lesser hematocrits were not only sicker than those with higher hematocrits but that they also received a lesser quality of care. Both of these factors could be the cause of the association of lower admission hematocrit and mortality. These confounders again emphasize the difficulty using retrospective chart analyses to assess issues regarding transfusion. Furthermore, Wu et al. [34] also noted that red cell transfusion was associated with a lower 30-day mortality when the hematocrit was equal to or less than 33% and an increased 30-day mortality when the hematocrit was equal to or more than 36%. This suggests that clinicians recognized the patients who were at greater risk and provided red cell transfusion for those patients. Substantiation or refutation of this hypothesis can not be readily performed from a retrospective analysis of a database. A similar problem appears in another observational report, finding an association of increased incidence of myocardial infarction with higher hematocrits of patients on admission to the intensive care unit following coronary artery surgery [35]. The thought that the clinicians caring for these patients during surgery recognized those at higher risk and transfused them to a higher hematocrit than those at lesser risk is supported by the additional finding that higher intensive care unit admission hematocrit was also associated with a greater incidence of severe left ventricular dysfunction. Finally, Hébert et al. [36] and Hébert [37] undertook retrospective analyses of subgroups of their multicenter study [17]. Of the total 838 patients in the study, 357 had cardiovascular disease. intensive care unit mortality did not differ in this subgroup between the 160 patients with a hemoglobin concentration of 8.5 g/dl and the 197 patients with a hemoglobin concentration of 10.3 g/dl. In a further subgroup analysis of the 158 patients with ischemic heart disease, which included those following coronary artery surgery, there was no difference in mortality between the two transfusion strategy groups; however, the hemoglobin concentrations of these subgroups were not reported [36, 37]. It should be noted that subgroup analyses do not carry the same force as do analyses of the original, primary study endpoint, and such subgroup analyses of the same database can produce misleading conclusions [38]. The brain has received far lesser attention than has the heart in clinical studies of transfusion although it is likely more sensitive to inadequate oxygen delivery, with injury occurring within 2 min of inadequate cerebral oxygenation [39, 40] and progressing with further duration of inadequate cerebral oxygenation [41]. In unmedicated patients, mental function can be used as a measure of cerebral function. In as much as this is not possible in medicated or anesthetized patients, other surrogate measures such as electrical activities of the brain may be used in those circumstances. We have examined sensitive measures of neurocognitive function and found that there are

Weiskopf

subtle, reversible decrements in unmedicated, normal humans at hemoglobin concentrations of 5 and 6 g/dl [42]. These small changes are reversible by increasing oxygen delivery with transfusion to produce a hemoglobin concentration of 7 g/dl [42] or by breathing oxygen [43]. Most of the data summarized above come from studies of various populations of people. Only prospective studies have a real possibility of providing valid information regarding the results of transfusion strategies and thresholds. These studies can tell us, in general, a probability that one group (e.g. with a specific hemoglobin concentration) did better, worse, or not differently than did another group (e.g. with a different hemoglobin concentration). However, such studies provide only a limited guide to the clinician who seeks to answer the question as to when a specific individual patient requires augmentation

of oxygen delivery. We need specific methods and technology that may be applied to each patient to warn us of impending inadequate oxygen delivery and when oxygen delivery is inadequate [44]. We are beginning to develop such measures, but the clinician does not yet have them available for routine use. Thus, ordinarily, we do not have sufficient information to determine whether a patient requires red cell transfusion, and this therapy continues to be decided, most often, by clinical judgment. Acknowledgement Supported, in part, by a Public Health Service Award from the National Heart, Lung and Blood Institute, National Institutes of Health, Grant # 1 P50 HL54476.

References 1 Adams R, Lundy J: Anesthesia in cases of poor surgical risk: Some suggestions for decreasing the risk. Surg Gynecol Obstet 1942;71:1011. 2 Consensus Conference: Perioperative red blood cell transfusion. JAMA 1988;260:2700–2703. 3 American Society of Anesthesiologists Task Force on Blood Component Therapy: Practice guidelines for blood component therapy: A report by the American Society of Anesthesiologists Task Force on Blood Component Therapy. Anesthesiology 1996;84:732–747. 4 Cain SM: Appearance of excess lactate in anesthetized dogs during anemic and hypoxic hypoxia. Am J Physiol 1965;209:604–610. 5 Cain SM: Oxygen delivery and uptake in dogs during anemic and hypoxic hypoxia. J Appl Physiol 1977;42:228–234. 6 Von Restorff W, Höfling B, Holtz J, Bassenge E: Effect of increased blood fluidity through hemodilution on general circulation at rest and during exercise in dogs. Pflügers Arch 1975;357:25–34. 7 Adams RP, Dieleman LA, Cain SM: A critical value for O2 transport in the rat. J Appl Physiol 1982;53:660–664. 8 Weiskopf RB, Viele M, Feiner J, Kelley S, Lieberman J, Noorani M, Leung J, Fisher D, Murray W, Toy P, Moore M: Human cardiovascular and metabolic response to acute, severe isovolemic anemia. JAMA 1998;279:217–221. 9 Lieberman JA, Weiskopf RB, Kelley SD, Feiner J, Noorani M, Leung J, Toy P, Viele M: Critical oxygen delivery in conscious humans is less than 7.3 ml O2 . kg–1 . min–1. Anesthesiology 2000;92:407–413. 10 Van Der Linden P, Wathieu M, Gilbart E, Engelman E, Wautrecht J-C, Lenaers A, Vincent J-L: Cardiovascular effects of moderate normovolaemic haemodilution during enflurane-nitrous oxide anaesthesia in man. Acta Anaesthesiol Scand 1994; 38:490–498. 11 Ickx BE, Rigolet M, Van der Linden PJ: Cardiovascular and metabolic response to acute normovolemic anemia. Anesthesiology 2000;93:1001– 1016. 12 Cahalan MK, Lurz FW, Eger EId, Schwartz LA, Beaupre PN, Smith JS: Narcotics decrease heart rate during inhalational anesthesia. Anesth Analg 1987;66:166–170.

Do We Have Enough Evidence to Know when to Transfuse Erythrocytes?

13 Spahn DR, Zollinger A, Schlumpf RB, Stohr S, Seifert B, Schmid ER, Pasch T: Hemodilution tolerance in elderly patients without known cardiac disease. Anesth Analg 1996;82:681–686. 14 Spahn DR, Schmid ER, Seifert B, Pasch T: Hemodilution tolerance in patients with coronary artery disease who are receiving chronic beta-adrenergic blocker therapy. Anesth Analg 1996;82:687–694. 15 Bush RL, Pevec WC, Holcroft JW: A prospective, randomized trial limiting perioperative red blood cell transfusions in vascular patients. Am J Surg 1997;174:143–148. 16 Fortune JB, Feustel PJ, Saifi J, Stratton HH, Newell JC, Shah DM: Influence of hematocrit on cardiopulmonary function after acute hemorrhage. J Trauma 1987;27:243–249. 17 Hébert PC, Wells G, Blajchman MA, Marshall J, Martin C, Pagliarello G, Tweeddale M, Schweitzer I, Yetisir E: A multicenter, randomized, controlled clinical trial of transfusion requirements in critical care. N Engl J Med 1999;340:409–417. 18 Hébert PC, Wells G, Tweeddale M, Martin C, Marshall J, Pham B, Blajchman M, Schweitzer I, Pagliarello G: Does transfusion practice affect mortality in critically ill patients? Transfusion Requirements in Critical Care (TRICC) Investigators and the Canadian Critical Care Trials Group. Am J Respir Crit Care Med 1997;155:1618–1623. 19 Von Restorff W, Höfling B, Holtz J, Bassenge E: Effect of increased blood fluidity through hemodilution on coronary circulation at rest and during exercise in dogs. Pflügers Arch 1975;357:15–24. 20 Jan KM, Chien S: Effect of hematocrit variations on coronary hemodynamics and oxygen utilization. Am J Physiol 1977;233:H106–H113. 21 Levy P, Kim S, Eckel P, Chavez R, Ezz F, Gould S, Salem M, Crystal G: Limit to cardiac compensation during acute isovolemic hemodilution: Influence of coronary stenosis. Am J Physiol 1993;265:H340– H349. 22 Crystal GJ, Salem MR: Myocardial oxygen consumption and segmental shortening during selective coronary hemodilution in dogs. Anesth Analg 1988;67:500– 508. 23 Doak GJ, Hall RI: Does hemoglobin concentration affect perioperative myocardial lactate flux in patients undergoing coronary artery bypass surgery? Anesth Analg 1995;80:910–916.

24 Mathru M, Kleinman B, Blakeman B, Sulllivan H, Kumar P, Dries D: Myocardial metabolism and adaptation during extreme hemodilution in humans after coronary revascularization. Crit Care Med 1992;20:1420–1425. 25 Catoire P, Saada M, Ngai L, Delaunay L, Rauss A, Bonnet F: Effect of preoperative normovolemic hemodilution on left ventricular segmental wall motion during abdominal aortic surgery. Anesth Analg 1992;75:654–659. 26 Hogue CWJ, Goodnough LT, Monk TG: Perioperative myocardial ischemic episodes are related to hematocrit level in patients undergoing radical prostatectomy. Transfusion 1998;38:924–931. 27 Johnson R, Thurer R, Kruskall M, Sirois C, Gervino E, Critchlow J, Weintraub R: Comparison of two transfusion strategies after elective operations for myocardial revascularization. J Thorac Cardiovasc Surg 1992;104:307–314. 28 Lawrence V, Silverstein J, Cornell J, Pederson T, Noveck H, Carson JL: Higher Hb level is associated with better early functional recovery after hip fracture repair. Transfusion 2003;43:1717–1722. 29 Carson J, Terrin M, Barton F, Aaron R, Greenburg A, Heck D, Magaziner J, Merlino F, Bunce G, McClelland B, Duff A, Noveck H: A pilot randomized trial comparing symptomatic vs. hemoglobin-leveldriven red blood cell transfusions following hip fracture. Transfusion 1998;38:522–529. 30 Woodson R, Wills R, Lenfant C: Effect of acute and established anemia on O2 transport at rest, submaximal and maximal work. J Appl Physiol 1978;44:36–43. 31 Horstman D, Weiskopf R, Jackson R: Work capacity during 3-wk sojourn at 4,300 m: Effects of relative polycythemia. J Appl Physiol 1980;49:311–318. 32 Ekblom B, Wilson G, Åstrand P-O: Central circulation during exercise after venesection and reinfusion of red blood cells. J Appl Physiol 1976;40: 379–383. 33 Bracey AW, Radovancevic R, Riggs SA, Houston S, Cozart H, Vaughn WK, Radovancevic B, McAllister HA Jr, Cooley DA: Lowering the hemoglobin threshold for transfusion in coronary artery bypass procedures: Effect on patient outcome. Transfusion 1999;39:1070–1077.


255

34 Wu W, Rathore S, Wang Y, Radford M, Krumholz H: Blood transfusion in elderly patients with acute myocardial infarction. N Engl J Med 2001;345: 1230–1236. 35 Spiess BD, Ley C, Body SC, Siegel LC, Stover EP, Maddi R, D’Ambra M, Jain U, Liu F, Hershkowitz A, Mangano DT, Levin J,: Hematocrit value on intensive care unit entry influences the frequency of q-wave myocardial infarction after coronary artery bypass grafting. The Institutions of the Multicenter Study of Perioperative Ischemia (McSPI) Research Group. J Thorac Cardiovasc Surg 1998;116:460– 467. 36 Hébert PC, Yetisir E, Martin C, Blajchman MA, Wells G, Marshall J, Tweeddale M, Pagliarello G, Schweitzer I: Is a low transfusion threshold safe in critically ill patients with cardiovascular diseases? Crit Care Med 2001;29:227–234.

256

37 Hébert PC: The TRICC trial: A focus on the subgroup analysis. Vox Sang 2002;83(suppl 1):387–396. 38 Lee K, McNeer J, Starmer C, Harris P, Rosati R: Clinical judgment and statistics. Lessons from a simulated randomized trial in coronary artery disease. Circulation 1980;61:508–515. 39 Dreher W, Kuhn B, Gyngell M, Busch E, Niendorf T, Hossmann K, Leibfritz D: Temporal and regional changes during focal ischemia in rat brain studied by proton spectroscopic imaging and quantitative diffusion NMR imaging. Magn Reson Med 1998;39:878–888. 40 Basu S, Liu X, NozarI A, Rubertsson S, Miclescu A, Wiklund L: Evidence for time-dependent maximum increase of free radical damage and eicosanoid formation in the brain as related to duration of cardiac arrest and cardio-pulmonary resuscitation. Free RadicRes 2003;37:251–256.


41 Hossmann K: Ischemia-mediated neuronal injury. Resuscitation 1993;26:225–235. 42 Weiskopf RB, Kramer JH, Viele M, Neumann M, Feiner J, Watson JJ, Hopf H, Toy P: Acute severe isovolemic anemia impairs cognitive function and memory and humans. Anesthesiology 2000;92: 1646–1652. 43 Weiskopf R, Feiner J, Hopf HW, Viele M, Watson J, Kramer JH, Ho R, Toy P: Oxygen reverses deficits of cognitive function and memory and increased heart rate induced by acute severe isovolemic anemia. Anesthesiology 2002;96:871–877. 44 Weiskopf RB: Do we know when to transfuse red cells to treat acute anemia? Transfusion 1998;38: 517–521.

Weiskopf


Received: April 7, 2004 Acccepted: May 25, 2004

Autologous Transfusion in Children: Blood-Saving Techniques* G. Kühbacher

P. Innerhofer

Department of Anaesthesiology and Critical Care Medicine, University of Innsbruck, Austria

Key Words Pediatric anesthesia · Autologous blood · Hemodilution · Autotransfusion

Schlüsselwörter Kinderanästhesie · Autologe Bluttransfusion · Hämodilution · Autotransfusion

Summary It is uncontested that blood-saving strategies should also be applied in children. However, in the past, blood-saving techniques saw limited use in children although they are wellestablished in adults. This is due to technical and methodical problems but also to the physiologically limited compensatory mechanisms for diminished oxygen delivery in this age group. For this reason, the various blood-saving techniques cannot be universally applied to all age groups. During the first year of life most perioperative techniques are of only limited benefit: Even a small amount of blood loss in relation to total blood volume makes a transfusion of allogeneic blood necessary, which is far before enough shed blood can be recovered for re-transfusion. Preoperative autologous donation of blood calls for a high degree of cooperation by the child and the parents and is equally demanding in terms of organization and skills. Therefore, this procedure is used mainly in school-age children and in adolescents. Intraoperative blood salvage, by contrast, is already worthwhile in children of at least 1 year with an expected blood loss of 20% of blood volume. Acute normovolemic hemodilution and deliberate hypotension should be recommended only in an age group where cardiovascular compensatory mechanisms are given. Supporting procedures aimed at avoiding the need for allogeneic blood include perioperative administration of erythropoietin, iron supplementation, blood-saving surgical techniques and careful hemostasis. These prerequisites, the combination of various techniques as well as the definition of an age-specific low transfusion trigger have contributed to a marked decrease in the need for allogeneic blood products in children over the recent years.

Zusammenfassung Es ist unbestritten, dass auch für Kinder Strategien zum Einsparen allogener Blutprodukte zur Anwendung kommen sollten. Allerdings sind fremdblutsparende Verfahren in der Vergangenheit, obwohl schon seit langem bei Erwachsenen etabliert, bei Kindern eher restriktiv gehandhabt worden. Einerseits ist dies bedingt durch technische Probleme der Verfahren, andererseits auch durch die in dieser Altersgruppe physiologisch limitierte Kompensationsmöglichkeit eines eingeschränkten Sauerstoffangebots. Demgemäß sind die verschiedenen Blutsparverfahren nicht universell auf alle Altersgruppen anwendbar. Bei einem Lebensalter unter 1 Jahr sind die meisten perioperativen Blutsparmaßnahmen nur begrenzt rentabel. Ein geringer Blutverlust in Relation zum Blutvolumen macht bereits die Transfusion von allogenen Blutprodukten erforderlich, noch bevor das Wundblut wiederaufbereitet werden kann. Die präoperative autologe Blutabnahme erfordert ein hohes Maß an Kooperation des Kindes und der Eltern sowie einen beträchtlichen organisatorischen und technischen Aufwand. Daher ist dieses Verfahren hauptsächlich Kindern ab dem Schulalter und Jugendlichen vorbehalten. Die intraoperative maschinelle Autotransfusion hingegen kann schon ab dem 1. Lebensjahr und einem erwarteten Blutverlust von 20% des Blutvolumens rentabel eingesetzt werden. Die akute normovolämische Hämodilution und die kontrollierte Hypotension sollte erst ab einem Alter angewandt werden, ab welchem die kardiovaskulären Kompensationsmechanismen gewährleistet sind. Unterstützende Maßnahmen zur Einsparung von Fremdblut bestehen im perioperativen Einsatz von Erythropoetin, einer ausreichenden Eisensubstitution, blutsparender Operationsverfahren und einer sorgfältigen Blutstillung. Diese Voraussetzungen und die Kombination der verschiedenen Techniken sowie eine altersbezogene Festsetzung niedriger Transfusionstrigger hat in den letzten Jahren auch bei Kindern zu einer deutlichen Einsparung von Fremdblut geführt.




Univ. Doz. Dr. Gabriele Kühbacher Klinik für Anaesthesie und Allgemeine Intensivmedizin Universität Innsbruck Anichstraße 35, A-6020 Innsbruck Tel. +43 512 504-802 71, Fax -24 50 E-mail [email protected]

Introduction Since the early 1980s the increasing use of blood-saving techniques in children has been described in the literature [1, 2]. In addition to the risk of infection and the danger of immunomodulation, the potential risk of alloimmunization against blood products is of special importance in view of the long life expectancy of these patients. Blood saving for elective surgery starts with careful perioperative planning, continues with the intraoperative implementation of technical devices and methods, and terminates with the definition of a transfusion trigger threshold for the particular patient. The following methods can be applied in children: – preoperative autologous blood donation (PAD), – intra- and postoperative autotransfusion (IAT), – intraoperative blood-saving methods such as acute normovolemic hemodilution (ANH) and deliberate hypotension, – supporting procedures (erythropoietin, coagulation therapy). In the following these methods are described, and their practicability for children is discussed.

Preoperative Autologous Blood Donation As we know from adults, PAD is an effective means in elective surgery to reduce exposure to allogeneic blood. In 1974 Cowell and Swickard [1] were the first to report on this method in children as young as 7 years of age. A total of 193 patients predisposed 1–2 autologous blood units 67% of whom underwent the surgical procedure without needing allogeneic blood. However, 30% of the donations were accompanied by technical difficulties, and 352 attempted venous punctures were necessary. In order to perform PAD, organizational, infrastructural and patient-specific prerequisites must be met. In recent reports the efficacy of this procedure in children having a body weight of less than 25 kg and an age below 7 years has been doubted [3]. In fact, many institutions prefer not to do autologous donations in infants and young children, because such young patients mostly need sedation or general anesthesia for the procedure, and venous access can be very difficult. Nevertheless, the literature contains some reports of successful cases of PAD in infants and children as young as 3 months of age and a success rate of 80–95% [4–7]. Prior to cardiac surgery in children aged between 3 and 15 years up to 5 PADs were conducted using cryotechnology [8]. Thereafter, 94% of these patients, and even all of the children younger than 5 years, were operated without allogeneic blood. Another study reported that after erythropoietin (100 U/kg s.c.) was administered 3 times a week 6 ml/kg of autologous blood was drawn and erythropoiesis increased by 11% preoperatively [9]. In contrast to the control group without PAD, 61.5% of which needed allogeneic blood units, only 7.7% of the study group received allogeneic blood.

258


Hemoglobin before donation should be minimum 11 g/dl or not below 10 g/dl. The generally accepted amount of donation per procedure is 10–12% of the estimated blood volume (EBV). The amount of anticoagulant must be adjusted to the amount of blood drawn, namely in a ratio of 1:7. In practice, this means that approximately 30 ml anticoagulant would be appropriate for 250 ml blood. At the present time this procedure still poses technical problems because most of the available bag systems are designed for a blood volume of at least 450 ml with 60 ml anticoagulant. Attempts are being made to develop smaller units with variable addition of anticoagulant. Depending on the estimated blood requirement, the donation program should run for 3–4 weeks preoperatively with 1 donation per week. In children the incidence of adverse reactions such as nausea, dizziness and hemodynamic disturbances during PAD is reported to be up to 27% [2]. It can be reduced to 3% by providing simultaneous volume replacement (crystalloids, colloids). In view of the facts that 0.5 mg of iron is lost per milliliter of blood drawn and that children are known to have limited iron reserves, iron supplementation during the entire course of donation is mandatory. Ideally, oral iron substitution (10 mg/kg/day) should already start 1 week before the 1st donation procedure and continue perioperatively. In order to obtain an even higher collection rate, erythropoietin (400 U/kg/week s.c.) can be administered [9]. This has been shown to increase the total volume of collected autologous blood by 41%. In the case of insufficient erythropoiesis or if the preoperative time period is too short, the child risks being anemic at the date of planned surgery. This relativizes the benefit of the entire donation program. An additional risk is posed by the possible bacterial contamination of the autologous erythrocyte concentrate. Moreover, in PAD and re-transfusion of autologous blood the possibility of human error (confusion of blood units) cannot be ruled out. For this reason the indication for transfusion of autologous blood is subject to the same restrictions as for allogeneic blood.

Intraoperative Autotransfusion Collection and recovery of intraoperative blood loss can be very difficult in children. Despite careful suctioning and rinsing of sponges, in children younger than 1 year about 50% of total blood loss can be expected to be lost for recovery. In contrast, IAT in adults is an established and efficient blood-saving technique. The cell saver devices available until recently needed a minimum blood loss of 1,000 ml in order to be efficient. Along with the development of smaller bell systems, a blood loss of 300 ml in children and a body weight of 20 kg were considered the lower limits for IAT [10, 11]. Meanwhile, new devices can also process very small amounts of blood [2, 12, 13] (table 1). Bell systems fill the bell, wash and concentrate the shed blood in separate steps (Haemonetics® and Or-

Kühbacher/Innerhofer

thopat®; Haemonetics, Munich, Germany; Dideco®; Sorin Biomedica, Puchheim, Germany). By contrast, the CATS® device (Fresenius AG, Bad Homburg, Germany) performs these steps simultaneously in a separation chamber and can thus produce an erythrocyte concentrate of only 15 ml with an hematocrit (Hct) of 50–60%. An in vitro study compared the Haemonetics, Dideco und CATS devices [13] and showed that both the Dideco and CATS devices produce a highly concentrated erythrocyte concentrate (Hct 50–60%), irrespective of blood loss; and the CATS device was the fastest. By contrast, the Haemonetics device produced an erythrocyte concentrate with Hct values of 25–48%, depending on the blood volume processed. The development of these new cell savers means that IAT can also be effectively performed in small children having at least 10 kg body weight and a minimum blood loss of 20% of their blood volume. The literature presently contains no recommendations for postoperative re-transfusion of unwashed blood in children because the safety of this autologous re-transfusion is still very controversial [14]. However, there is no reason why postoperative blood loss in children cannot be salvaged using a cell saver.

Acute Normovolemic Hemodilution and Deliberate Hypotension ANH is an intraoperative blood-saving procedure that in combination with IAT effectively reduces (75%) the need for allogeneic blood products. The aim of ANH is to induce anemia under normovolemic conditions in order to minimize intraoperative erythrocyte loss and provide whole blood with intact coagulation properties for retransfusion. Depending on the desired minimal Hct (Hctmin), ANH is termed mild (Hct 25–30%), moderate (Hct 20–25%), or severe (Hct < 20%). Immediately after inducing anesthesia, a defined amount of autologous whole blood is drawn into citrate bags. Simultaneously, this blood is normovolemically replaced with lactated Ringer’s solution in a 1:3 ratio or with colloids 1:1. The volume of blood to be drawn is calculated as follows [15]: V = TBV × (Hct0 – Hct1) / Hct0/1

Table 2. Examples of studies reporting on profound ANH in infants and children between 1 month and 19.9 years of agea

(1)

whereby V is the blood volume to be drawn, TBV the total blood volume, Hct0 the starting Hct, Hct1 the desired Hct (25–30%), and Hct0/1 the mean Hct. Until the minimal tolerable Hct is reached (Hctmin 15–20%), the ongoing blood loss is normovolemically compensated with colloids and lactated Ringer’s solution, and the erythrocyte concentrate regained with the cell saver is re-transfused. Only then, ideally at the end of the surgical procedure, is the previously drawn autologous whole blood re-transfused. Under these conditions of restricted arterial oxygen content, the following compensatory mechanisms are triggered under normovolemia: besides a decrease in viscosity which causes increased perfusion ANH is also accompanied by a decrease in systemic vascular resistance (afterload reduction) and an increase in venous backflow (preload increase). The increase in cardiac output is the most important as well as the limiting factor for compensation of ANH-induced reduction in oxygen supply. Under advanced ANH the oxygen extraction rate also increases. Moreover, the shift to the right in the oxygen dissociation curve eases oxygen supply to the tissues. The calculations for oxygen content (CaO2) and oxygen supply (DO2) show that anemia can be compensated to a certain degree by increasing the physically dissolved oxygen (high FiO2) but after all by an increase in cardiac output. On the other hand, calculations show that only a pro-

Table 1. Comparison of presently available cell saver devices with regard to required filling volumea Cell saver

Bell (chamber), ml

Reservoir, ml

Haemonetics Haemonetics Orthopat Dideco CATS

250 125 100 55 30

1,000 300 200 100 60

a Comparison of EBV in reservoir approximately needed to fill the bell system or separation chamber (CATS®). For a process volume of 55 ml and 30 ml with an estimated Hct of 30% an erythrocyte concentrate of approximately 30 ml and 15 ml, respectively, with an Hct of 50–60% can be expected.

Studies

Number of patients

Age, years

Volume substitute for ANH

Hctmin, %

SvO2, %

Schaller et al., 1984 [20] Haberkern and Dangel, 1991 [21] Fontana et al., 1995 [22] AlyHassan et al., 1997 [23]

8 30 8 16

0.083–12 3.4–19.9 12.5 ± 1.5 1–8

lactated Ringer’s solution lactated Ringer’s solution human albumin 5% HES/dextran

9 10 6.3 14

58 55 60 68

aVolume replacement during ANH was established using Lactated Ringer’s solution, human albumin, HES, and dextran. The data given for Hct and SvO2 are the absolutely lowest reported values for these patient collectives. SvO2 corresponds to the mixed-venous saturation that is recommended as the indicator for the critical threshold for oxygen supply in those studies (in Haberkern and Dangel [21] the value represents the central venous saturation [17]).

Blood-Saving Techniques in Children


259

nounced ANH can ensure effective saving of allogeneic blood products [16]. CaO2 = (Hb × SO2 × 1.38) + (PO2 × 0.003) DO2 = CO × CaO2

(2) (3)

Severe ANH entails the risk that the critical threshold for oxygen supply (critical DO2) is reached where oxygen consumption becomes dependent on oxygen supply [17]. This raises the question as to the critical threshold for oxygen supply in children. Infants in the first months of life are hardly able to increase their stroke volume in order to adjust their cardiac output to this critical situation. Moreover, at this age the proportion of fetal hemoglobin with a left-shift in the oxygen dissociation curve is variable. For this reason, ANH in children younger than 1 year is considered dangerous and is not recommended [12]. The critical DO2 below which oxygen consumption is dependent on oxygen supply and below which lactacidosis as defined by a systemic lack of oxygen manifests itself, is not clearly defined for children. In the literature the critical thresholds for DO2 vary between 184 and 330 ml/min/m2 [18, 19]. Nevertheless, massive forms of ANH have been performed in children of various ages [20–23]. Table 2 shows the age distribution, the volume substitution, and the lowest measured values for Hct and mixed-venous saturation under severe hemodilution. These studies report no serious complications despite massive decrease in oxygen supply. One study reported that a patient temporarily experienced an S-T segment drop at a hemoglobin concentration of 2.1 g/dl, which was quickly compensated by re-transfusing autologous blood [22]. At a minimal hemoglobin value of 3 g/dl, DO2 dropped on average from 532 to 260 ml/min/m2. By contrast, a study comparing dextran with hydroxyethyl starch (HES) as colloid substitute reported a minimal DO2 of 311 ml/min/m2 at 17% Hctmin [23]. In terms of blood-saving techniques these reports are very positive. Nevertheless, lactacidosis occurred in isolated cases, which leads us to assume that the absolutely critical threshold was reached. Animal experiments showed that below 10% Hct systemic and myocardial lactate production commence as a manifestation of oxygen deficiency [24, 25]. Deliberate hypotension to a mean arterial pressure of 40–50 mm Hg also contributes to a decrease in intraoperative blood loss. In addition to the hypotensive effect of inhalation anesthetics such as isoflurane and sevoflurane, vasodilators (nitroglycerin, labetalol, prostaglandin) have been shown to decrease the mean arterial pressure in children [26]. The administration of central regional blocks (spinal, epidural anesthesia), particularly involving intrathecal administration of opioids, can cause a marked drop in intraoperative blood loss. This technique is mainly applied for spinal corrections. Immediately before the operation is commenced, sufentanil and morphine are injected intrathecally in a single-shot technique. Several studies showed that this technique produces significantly less blood loss than does systemic analgesia (27 vs. 53%) [27, 28].

260


A combination of deliberate hypotension and ANH can be justified only when ANH is mild to moderate [29]. In the case of severe ANH, the cardiovascular compensation mechanisms can be negatively influenced by the additionally induced hypotension so that cardiac performance and microvascular oxygen supply are severely strained. In an infant aged 7 months, undergoing severe ANH (Hct 9%) and halothane-induced hypotension and having a mean arterial pressure of only 38 mm Hg the cardiac index dropped from 3.9 to 0.95 l/min/m2 [20]. Massive lactacidosis occurred simultaneously. Following retransfusion of autologous blood the child’s condition improved by the end of surgery. This report leads us to conclude that, given the massive decrease in oxygen supply, the cardiodepressive effect of halothane caused, albeit reversible, cardiac decompensation and oxygen deficit.

Coagulation Therapy ANH or interventions involving massive volume exchange can cause a disturbance in homeostasis by diluting the coagulation factors. This is called dilutional coagulopathy [30, 31]. Dilutional coagulopathy is aggravated when the patient is hypothermic, which means that special care is to be taken to ensure that children are normothermic. In the case of increasing hemodilution, close intraoperative coagulation monitoring by means of thrombelastography (ROTEM®, Pentapharm, Munich, Germany) will first show a drop in fibrinogen concentration and subsequently in the extrinsic and intrinsic coagulation factors. Below 150 mg/dl hypofibrinogenemia clinically also causes a marked bleeding tendency. As a fast alternative to administration of fresh frozen plasma (FFP), early fibrinogen substitution (Haemocomplettan®, 20–30 mg/kg; Aventis Behring GmbH, Marburg, Germany) can slow progressive intraoperative blood loss and reduce the need for allogeneic blood. Continued loss of coagulation factors can be compensated with prothrombin concentrate and ATIII (10 IU/kg each). In order to adequately compensate coagulopathy a minimum FFP of 20 ml/kg would be necessary by contrast.

Conclusions Blood-saving techniques that are well-established for adults can also be used for children. The efficacy and benefit of these methods is, however, dependent on the child’s age and in children under the age of 1 year very limited. From the age of 1 year a combination of various techniques, and above all IAT, produces good results. Administration of erythropoietin, iron substitution with or without PAD as well as ANH are mainly for older children and adolescents. All in all, blood-saving techniques can only be successful in the face of good surgical planning and close cooperation between the attending physicians (anesthesiologist, surgeon, transfusion specialist).

Kühbacher/Innerhofer

References 1 Cowell HR, Swickard JW: Autotransfusion in children’s orthopedics. J Bone Joint Surg 1974;56:908– 912. 2 Murto KTT, Splinter WM: Perioperative autologous blood donation in children. Transfus Sci 1999;21:41–62. 3 British Committee for Standards in Haematology, Blood Transfusion Taskforce: Guidelines for autologous transfusion. 1. Preoperative autologous donation. Transfus Med 1993;3:307–316. 4 Longatti PL, Paccagnella F, Agostini S, Nieri A, Carteri A: Autologous hemodonation in the corrective surgery of craniostenosis. Childs Nerv Syst 1991;7:40–42. 5 Yamazaki Y, Mizuno R, Yuno S: Predeposited autologous blood transfusion for pediatric surgery. J Jpn Soc Pediatr Surg 1992;28:978–982. 6 Kemmotsu H, Kazuya J, Nakamura H, Yamashita M: Predeposited autologous blood transfusion for surgery in infants and children. J Pediatr Surg 1995; 30:659–661. 7 Taguchi T, Suita S, Nakao M, Yamanouchi T, Inaba S: The efficacy of predeposited autologous blood transfusions in general pediatric surgery. Surg Today 2000;30:773–777. 8 Masuda M, Kawachi Y, Inaba S, Matsuzaki K, Fukumura F, Morita S, Tominaga R, Yasui H: Preoperative autologous blood donations in pediatric cardiac surgery. Ann Thorac Surg 1995;60:1694– 1697. 9 Sonzogni V, Crupi G, Poma R, Annechino F, Ferri F, Filisetti P, Bellavita P: Erythropoietin therapy and preoperative autologous blood donation in children undergoing open heart surgery. Br J Anaesth 2001;87:429–434. 10 Mazzarello G, Lampugnani E, Carbone M, Rivabella L, Ivani G: Blood saving in children. Anaesthesia 1998;53(suppl 2):30–32.

Blood-Saving Techniques in Children

11 DePalma L, Luban NLC: Autologous blood transfusion in pediatrics. Pediatrics 1990;85:125–128. 12 Weber TP, Große Hartlage MA, VanAken H, Booke M: Anaesthetic strategies to reduce perioperative blood loss in paediatric surgery. Eur J Anaesth 2003;20:175–181. 13 Booke M, Hagemann O, VanAken H, Erren M, Wullenweber J, Bone HG: Intraoperative autotransfusion in small children: An in vitro investigation to study its feasibility. Anesth Analg 1999;88: 763–765. 14 Krohn CD, Reikeras O, Bjornsen S, Brosstad F: Fibrinogen, fibrin and its degradation products in drained blood after major orthopaedic surgery. Blood Coagul Fibrinolysis 1999;10:167–171. 15 Gross JB: Estimating allowable blood loss: Correction for dilution. Anesthesiology 1983;58:277–280. 16 Feldman JM, Roth JV, Bjoraker DG: Maximum blood savings by acute normovolemic hemodilution. Anesth Analg 1995;80:108–113. 17 VanWoerkens EC, Trouwborst A, vanLanschot JJ: Profound hemodilution: What is the critical level of hemodilution at which oxygen delivery-dependent oxygen consumption starts in an anesthetized human? Anesth Analg 1992;75:818–821. 18 Leone BJ, Spahn DR: Anemia, hemodilution and oxygen delivery. Anesth Analg 1992;75:651–653. 19 Singler RC, Furman EB: Hemodilution, how low a minimum hematocrit? Anesthesiology 1987;53:72. 20 Schaller RT, Schaller J, Furman EB: The advantages of hemodilution anesthesia for major liver resection in children. J Pediatr Surg 1984;19:705–710. 21 Haberkern M, Dangel P: Normovolaemic haemodilution and intraoperative autotransfusion in children: Experience with 30 cases of spinal fusion. Eur J Pediatr Surg 1991;1:30–35.

22 Fontana JL, Welborn L, Mongan PD, Sturm P, Martin G, Bunger R: Oxygen consumption and cardiovascular function in children during profound intraoperative normovolemic hemodilution. Anesth Analg 1995;80:219–225. 23 AlyHassan A, Lochbuehler H, Frey L, Messmer K: Global tissue oxygenation during normovolaemic haemodilution in young children. Paediatr Anaesth 1997;7:197–204. 24 Wilkerson DK, Rosen AL, Gould SA, Sehgal LR, Sehgal HL, Moss GS: Oxygen extraction ratio: A valid indicator of myocardial metabolism in anemia. J Surg Res 1987;42:629–634. 25 Haisjackl M, Luz G, Sparr H, Germann R, Salak N, Friesenecker B, Deusch E, Meusburger S, Hasibeder W: The effects of progressive anemia on jejunal mucosal and serosal tissue oxygenation in pigs. Anesth Analg 1997;84:538–544. 26 Tobias JD: Controlled hypotension in children. Paediatr Drugs 2002;4:439–453. 27 Goodarzi M: The advantages of intrathecal opioids for spinal fusion in children. Paediatr Anaesth. 1998;8:131–134. 28 Gall O, Aubineau JV, Berniere J, Desjeux L, Murat I: Analgesic effect of low-dose intrathecal morphine after spinal fusion in children. Anesthesiology 2001;94:447–452. 29 Copley LA, Richards BS, Safavi FZ, Newton PO: Hemodilution as a method to reduce transfusion requirements in adolescent spine fusion surgery. Spine 1999;24:219–222. 30 Hiippala ST, Myllylä GJ, Vahtera EM: Hemostatic factors and replacement of major blood loss with plasma-poor red cell concentrates. Anesth Analg 1995;81:360–365. 31 Singbartl K, Innerhofer P, Radvan J, Westphalen B, Fries D, Stogbauer R, Van Aken H: Hemostasis and hemodilution: A quantitative mathematical guide for clinical practice. Anesth Analg 2003;96: 929–935.


261



Artificial Oxygen Carriers: Hemoglobin-Based Oxygen Carriers – Current Status 2004* T. Standl Klinik und Poliklinik für Anästhesiologie, Universitätsklinikum Hamburg-Eppendorf, Germany

Key Words Blood substitute ⋅ Cell-free hemoglobin ⋅ Hemoglobin solution ⋅ Artificial oxygen carriers ⋅ Tissue oxygenation ⋅ Volume substitute

Schlüsselwörter Blutersatz ⋅ Gewebeoxygenierung ⋅ Hämoglobinlösung ⋅ Künstliche Sauerstoffträger ⋅ Volumenersatz ⋅ Zellfreies Hämoglobin

Summary Because of an impending shortfall of allogeneic blood products within the next decades and ongoing problems associated with relevant costs for testing and storage of banked red blood cell (RBC) units, the development of alternatives has been intensified during the last 15 years. Modern chemically modified hemoglobin-based oxygen carriers (HBOC) are free of RBC membrane remnants, renal toxicity, and ABO antigens which allows transfusion without knowledge of the respective blood group of a patient. Bovine polymerized cell-free hemoglobin can be stored at room temperature for 3 years. In contrast to the perfluorocarbon solutions, HBOC can be applied at room air oxygen concentrations. Animal experiments have shown that HBOC can compensate for intravascular volume deficits in hemorrhagic shock, including restoration of colloid osmotic pressure and organ perfusion, and deliver oxygen to organs and tissues during nearly complete blood exchange. Chemical modifications of HBOC are able to reduce the vasoconstrictive side effect of HBOC which is caused by NO scavenging. In spite of vasoconstriction, the increased oxygen extraction in presence of HBOC in combination with the plasmatic oxygen transport provides enhanced tissue oxygenation even in post-stenotic tissues. HBOC seem to improve the diffusive oxygen transport at the microcirculatory site, thus decreasing tissue damage in acute pancreatitis and tissue injury in the heart and brain after ischemia/reperfusion. Clinical studies showed that the perioperative use of different HBOC (Hemopure®, PolyHeme®, Hemolink®, HemAssist®) can reduce the number of allogeneic RBC units and increase the avoidance rate of allogeneic transfusion in emergency bleeding as well as in vascular, cardiac and noncardiac surgery. Polymerized HBOC appear to have a lower potential of side effects in comparison to intramolecularly crosslinked preparations. However, HBOC-201 (Hemopure) is the only substance which has been licensed for the treatment of patients until now.

Zusammenfassung Die Erforschung und Erprobung von künstlichen Sauerstoffträgern wie den zellfreien Hämoglobinlösungen (hemoglobinbased oxygen carriers = HBOC) wurde in den letzten 15 Jahren auf Grund einer möglichen künftigen Verknappung an homologen Blutprodukten gefördert. Auch bestehen weiterhin Probleme in der Anwendung von Fremdblut sowie steigende Kosten für Gewinnung, Testung und Lagerung von Erythrozytenkonzentraten (EKs). Moderne chemisch modifizierte Hämoglobinlösungen weisen keine Reste von Erythrozytenmembranen, keine nephrotoxischen Nebenwirkungen und keine AB0-Antigenität auf. Dies erlaubt die Transfusion von HBOC ohne Kenntnis der Blutgruppe des Patienten. Polymerisierte Rinder-HBOC können 3 Jahre bei Raumtemperatur gelagert werden und benötigen nicht wie Perfluorocarbone eine erhöhte inspiratorische O2Konzentration. In Tierexperimenten konnte gezeigt werden, dass mit HBOC ein suffizienter Volumenersatz sowie eine adäquate oder sogar verbesserte Gewebeoxygenierung nach hämorrhagischem Schock und während eines Blutaustauschs möglich sind. Durch chemische Modifikation von HBOC kann die Intensität der vasokonstriktorischen Nebeneffekte, deren Ursache die NO-Bindung ist, reduziert werden. Die erhöhte O2Extraktion in Anwesenheit von HBOC in Kombination mit dem plasmatischen O2-Transport bewirkt trotz Vasokonstriktion eine gesteigerte Gewebeoxygenierung sogar im poststenotischen Gewebe. Mit HBOC wird die Perfusion auf mikrozirkulatorischer Ebene verbessert, so dass Gewebeschäden nach akuter Pankreatitis sowie Schäden am Herzen und am Hirn nach Ischämie/ Reperfusion signifikant reduziert werden konnten. Klinische Studien haben gezeigt, dass der perioperative Einsatz von unterschiedlichen Hämoglobinlösungen (Hemopure®, PolyHeme®, Hemolink®, HemAssist®) die Anzahl benötigter homologer EKs reduzieren und die Anzahl der Patienten ohne Fremdbluttransfusion erhöhen kann. Dies wurde bei Notfallpatienten sowie in der Gefäß-, Herz- und Abdominalchirurgie gezeigt. Polymerisierte HBOC scheinen eine geringere Inzidenz von Nebenwirkungen zu haben als intramolekular vernetzte Hämoglobinpräparate. Bislang ist HBOC-201 (Hemopure) die einzige Substanz mit Zulassung für den Einsatz am Patienten.




Prof. Dr. med. Thomas Standl Klinik und Poliklinik für Anästhesiologie Universitätsklinikum Hamburg-Eppendorf Martinistraße 52, D-20246 Hamburg E-mail [email protected]

24

Introduction

Artificial Oxygen Carriers: Hemoglobin-Based Oxygen Carriers

20

total >65 years

16

Units

Several reasons have encouraged the numerous experimental and clinical studies on artificial oxygen carriers during the last two decades. One of the main arguments to enforce the development of alternatives to allogeneic blood products is a worldwide aging population with an increasing need for red blood cell (RBC) substitutes because of the increased morbidity of elderly people. In addition, increasing numbers of elderly patients undergo highly invasive and major surgical procedures, e.g. radical tumor resection, which are associated with significant intraand postoperative blood loss and transfusion requirements. Moreover, elderly patients are often not allowed to predonate autologous blood prior to elective surgery because of cardiovascular or pulmonary diseases. Thus, the percentage of allogeneic bood donors will decrease continuously and inversively to the increasing percentage of elderly people in the industrialized countries. Calculations which are based on the numbers needed for allogeneic transfusion in the 1990s anticipate a shortfall of millions of packed RBC units within the next decades (fig. 1) [1]. Secondly, the transfusion of allogeneic blood, e.g. in case of intraoperative bleeding or trauma, is associated with a low but still existing risk of viral transmission, e.g. HBV, HCV or HIV [2], and of transfusion reactions caused by bacterial contamination, allergic reactions, and blood bank or user errors (clerical error, mislabeling, misidentification). Concerns about the safety of human blood products exist also because of the unclear risk of transmission of Creutzfeldt-Jacob disease and other prions, and the reports on rare but severe parvovirus B19 infections in immunocompromized patients. Another complication of blood transfusion is the transfusion-related acute lung injury which is characterized by severe bilateral lung edema, hypoxemia, tachycardia, fever, and hypotension occurring within 1–6 h after transfusion and is associated with a high mortality [3]. There is also evidence in the literature that allogeneic blood transfusion may increase the risk of infectious complications [4] and of cancer recurrence after major tumor surgery because of immunosuppressive side effects [5]. Moreover, the collection and cross-matching of allogeneic blood is expensive and time consuming, and the storage of RBCs is limited by a shelf life of 35–42 days. Depending on the time of storage, RBCs suffer from 2,3-diphosphoglycerate (2,3-DPG) depletion which increases their oxygen affinity because of a shift of the oxygen dissociation curve to the left, thus decreasing their ability to release oxygen to the tissues. Since old RBCs are less deformable – because of ATP depletion – they can clog capillaries, which can deteriorate microcirculation and tissue oxygenation, especially in patients with hemorrhagic or septic shock. As a consequence, the need for a safe and effective alternative to allogeneic packed RBC units has intensified the develop-

Fig. 1. Incidence of blood component transfusion (blood units) in the USA in 1989 and calculation of blood units required in 2030. The dark grey columns indicate the number needed for patients older than 65 years. Data from [1].

12 8 4 0

1989

2030

ment of two major groups of artificial RBC substitutes, the perfluorocarbon-based and the cell-free hemoglobin-based oxygen carriers (HBOC).

Characteristics of Hemoglobin-Based Oxygen Carriers HBOC are normally produced from outdated human packed RBC units or from bovine blood. In addition, recombinant and transgenic HBOC have been produced which are derived from yeast colonies (Saccharomyces cerevisiae) and tobacco plants or from different animal species. Modern HBOC are deprived of RBC membrane remnants by ultrapurification processes, thus eliminating their toxicity on kidneys and liver and their side effects on the coagulation and complement system [6–8]. HBOC do not possess ABO antigens which allows transfusion without knowledge of the respective blood group of a patient. As a consequence, immediate enhancement of the oxygen-carrying capacity without prior blood group testing and cross-matching will be possible in case of emergency bleeding. In addition, HBOC can also replace allogeneic RBC transfusion in patients with rare blood groups and irregular antibodies [9]. Chemical modifications have decreased the side effects and increased the effectiveness of HBOC. The intravascular halflife has been prolonged by intramolecular cross-linking, e.g. diaspirin connection between the two α-α chains of the hemoglobin (Hb) molecule, and conjugation of macromolecules such as polyethylene glycol to the Hb tetramer (fig. 2). In contrast to these modifications which have no significant influence on the increased colloid oncotic pressure (COP) of HBOC solutions, encapsulation of cell-free Hb into liposomes and intermolecular cross-linking, e.g. polymerization with glutaraldehyde, can reduce the COP, especially in comparison to the unmodified tetrameric Hb molecule. As a consequence, encapsulated or polymerized HBOC solutions can be applied in clinically relevant Hb concentrations of 12–13 g/dl. In case of liposome encapsulation, reduction enzymes and 2,3-DPG can be added to the cell-free Hb molecules thus delaying the oxidation of Hb to metHb and decreasing their oxygen affinity. The increased oxygen affinity (p50 = 13 mm Hg) of cell-


263

Conjugation

polyethylene glycol: pegylated Hb

Encapsulation

liposomes

Intramolecular cross-linking

diaspirin cross-linked (DD) genetic (108 Lysin)

RBC HBOC Intermolecular cross-linking

RBC

O-raffinose (ȕ-ȕ): Hb-raffimer glutaraldehyde polymer

Fig. 2. Chemical modifications of cell-free Hb.

free Hb in comparison to intracellular Hb (p50 = 26 mm Hg) can also be reduced by chemical modifications such as pyridoxalation or by genetic manipulation, which restore the oxygen affinity to physiological levels. Bovine Hb (e.g. HBOC201) has a low oxygen affinity (p50 = 34 mm Hg) per se and a reduced 2,3-DPG dependency since its oxygen affinity normally is regulated by chloride ions rather than by the 2,3-DPG concentration. Bovine Hb also shows a pronounced Bohr effect which should make this Hb an ideal RBC substitute in terms of tissue oxygenation. The bovine glutaraldeyde-polymerized products HBOC-301 and HBOC-201 (Oxyglobin® and Hemopure®, Biopure, Cambridge, MA, USA) can be stored at room temperature for almost 3 years. In contrast to the perfluorocarbon solutions, HBOC can be applied at normal inspiratory oxyen concentrations (FIO2 = 0.21) and as repeated infusions, e.g. during the postoperative period in the ICU or on the ward. The intravascular half-life of HBOC depends on the applied dose and ranges between 8 and 24 h for the majority of intraor intermolecurlarly cross-linked HBOC [10–13]. Encapsulation of HBOC and the formation of HBOC molecules with high molecular weight (millions of Daltons) can enhance the intravascular presence but may also increase the uptake and storage of these compounds in the RES.

Animal Experiments with Hemoglobin-Based Oxygen Carriers Animal experiments have shown that HBOC can compensate for intravascular volume deficits in hemorrhagic or burn shock, including restoration of colloid osmotic pressure and organ perfusion, and deliver oxygen to organs and tissues in animals undergoing extended or nearly complete blood exchange [10, 14–18]. In an animal study which mimics preoperative blood donation substitution with bovine Hb an increased harvest during autologous blood sampling was demonstrated, resulting in decreased need for allogeneic blood transfusion

264


Fig. 3. Two erythrocytes (RBC) in the center of a capillary are surrounded by cell-free hemoglobin (HBOC) which is homogeneously distributed in the plasma (from Standl T, Lipfert B, Reeker W, Schulte am Esch J, Lorke DE: Acute effects of a complete blood exchange with ultrapurified haemoglobin solution or hydroxyethyl starch on liver and kidney in an animal model. Anästhesiol Intensivmed Notfallmed Schmerzther 1996;31:354–361. With kind approval of Thieme Verlag, Stuttgart).

on the day of surgery [19]. When these results are transferred to the clinical setting, autologous blood donation can be performed within a very short time before operation. As a consequence there is no need to suspend surgical procedures after predonation until the bone marrow has compensated blood loss. In addition, HBOC may also serve as an oxygen-carrying bridge and allow predonation in patients with coronary artery disease, helping to overcome the nadir of RBC count which may be associated with cardiac events due to anemic hypoxia. In terms of cardiovascular changes, an increase in mean arterial pressure and systemic vascular resistance caused by NO scavenging is associated with the application of HBOC. It has been recently shown that cell-free Hb molecules can be uptaken by vascular endothelial cells and are transferred to the media by transcytosis, where they can bind to NO, thus creating vasonconstriction [20]. Chemical modifications of HBOC can help to reduce this vasoconstrictive side effect [21–23]. However, systemic vasoconstriction is not paralleled by a reduced regional organ blood flow or impaired microcirculation [22, 23]. In contrast, animal studies have shown improved tissue oxygenation during blood exchange, restoration of the pancreatic microcirculation and tissue oxygen tensions, and reduced acidosis with better mucosal integrity of the intestine after resuscitation from hemorrhagic shock with different HBOC [16, 24, 25]. In experimental acute pancreatitis bovine HBOC-301 was able to significantly reduce histologic tissue damage in rats [26]. In acute porcine pancreatitis hemodilution with HES plus HBOC-301 improved the pancreatic microcirculation and tissue oxygenation and increased the 6-day survival rate [27].

Standl

Table 1. Synopsis of the main HBOC products Product

Source

Chemical modification

Company

Comments on status

Hemopure®

bovine

glutaraldehyde polymerized

Biopure, Cambridge, MA, USA

approval 2001 in SA for treating perioperative anemia FDA licence application

Oxyglobin®

bovine


Biopure, Cambridge, MA, USA

approval 1999 for treating anemia in dogs

PolyHeme®

human


Northfield, Evanston, IL, USA

FDA licence application

Hemolink®

human

O-raffinose intra- and intermolecular cross-linking (raffimer)

Hemosol, Mississauga, ON, Canada

clinical studies suspended

HemAssist®

human

intramolecularly diaspirin α-α cross-linked teramer

Baxter, Deerfield, IL, USA

clinical and experimental studies stopped

Optro®

recombinant

β-chain mutation (108 Lys), intra-molecular cross-linking

Somatogen, Boulder, CA, USA

clinical and experimental studies stopped

PHP

human

surface modified polyoxyethylene pyridoxalated polymer

Apex, Plainfield, IL, USA

under investigation

PEG-Hb®

bovine

polyethylene glycol conjugated (pegylated)

Enzon, Piscataway, NJ, USA

under investigation

There is evidence from these animal studies that the vasoconstrictive effect of HBOC on the major arteries is mitigated in the smaller vessels and even neutralized at the microcirculatory site. In addition, vasoconstriction at the convective site of oxygen transport seems to be more than compensated by the enhanced oxygen off-loading capacity of HBOC which represents the more important diffusive part of oxygen delivery that is responsible for tissue oxygenation. This assumption is confirmed by studies which showed that HBOC-201 has a higher oxygen diffusion coefficient than whole blood [28] and provides also increased oxygen off-load from the RBCs [29]. The augmented oxygen extraction in presence of HBOC seems to be a characteristic essential of modern cell-free Hb which provides enhanced oxygen supply at the tissue site. Moreover, HBOC seem to have a higher oxygenation potency when compared to RBCs [30] because of their plasmatic oxygen transport and enhanced oxygen release from HBOC themselves and from adjacent erythrocytes (fig. 3) [31]. In this context the poorly oxygenated tissues seem to have the highest benefit from the differential oxygenation with HBOC [32]. The increased oxygen extraction by HBOC in combination with a significant plasmatic oxygen transport provides enhanced tissue oxygenation even in poststenotic tissues [33]. In a dog model with acute anemia and 90% stenosis of the left anterior descendent coronary artery, prophylactic application of HBOC-201 was able to prevent acute and severe decrease of myocardial tissue oxygen tensions and of left ventricular contractility in the dependent myocardial territory [34]. In a pig model with critical coronary stenosis and acute hemorrhage diaspirin cross-linked Hb provided a higher coronary perfusion and subendocardial oxygen delivery associated with a lower mortality in comparison with albumin [35]. In a rat


model of myocardial ischemia and reperfusion, infusion of HBOC-301 before coronary occlusion significantly decreased the incidence of severe arrythmias and the degree of histological damage within the respective myocardial area at risk [36, 37]. Hemodilution with diaspirin cross-linked Hb resulted in reduced brain damage after focal cerebral ischemia and reperfusion in rats when compared to albumin [38]. In conclusion, HBOC provide sufficient tissue oxygenation during resuscitation from hemorrhage or nearly complete blood exchange, improve poststenotic tissue oxygenation, and seem to decrease tissue damage in acute pancreatitis and tissue injury in heart and brain after ischemia/reperfusion.

Clinical Studies with Hemoglobin-Based Oxygen Carriers Several different HBOC are actually examined in phase I, II and III studies in volunteers and patients (table 1). However, bovine HBOC-201 is the only substance which has been approved for the treatment of patients with acute perioperative anemia in South Africa in the year 2001. Preceding phase I and II studies in patients undergoing acute normovolemic hemodilution (ANH) before surgery have shown a moderate vasoconstrictive side effect of this Hb preparation, resulting in an increase of the mean arterial pressure by 10–20% [11, 39]. However, no cerebral vasoconstriction could be detected in another group of these patients in whom cerebral blood flow was measured with transcranial doppler sonography [40]. No significant increases in liver and pancreas enzymes were seen after infusion of 0.4 g/dl HBOC-201 in patients undergoing liver resection [41]. The perioperative infusion of 150 g


265

Table 2. Main results of phase III-studies with bovine HBOC-201 Patients

Cardiac surgery [43]

Vascular surgery [42]

Abdominal surgery, orthopedics (submitted)

Avoidance of allogeneic RBC transfusion, %

34

27

43

Reduction of allogeneic RBC units HBOC-201 RBCs

1.7 mean, p = 0.05 2.2

2.0 median, p = 0.9 2.5

2.0 median, p = 0.004 3.0

Mortality, % HBOC-201 Control (RBC)

2 0

6 8

7 7

Incidence of SAEs, % HBOC-201 Control (RBC)

30 31

38 38

29 26

SAEs = Serious adverse events.

HBOC-201 within 96 h or 120 g HBOC-201 within 72 h reduced the amount of transfused units of allogeneic packed RBCs and eliminated the need for allogeneic RBC transfusion in 27 and 34% of the patients undergoing abdominal aneurysm repair and cardiac surgery, respectively [42, 43]. The data of an actually submitted clinical study with HBOC-201 in patients undergoing noncardiac surgery showed that repetitive infusions of relevant doses of HBOC-201 (maximum 210 g per patient) within 6 perioperative days were able to avoid allogeneic transfusion of RBCs in 43% of the patients and to significantly reduce the amount of transfused RBC units (table 2). The incidence of adverse and serious adverse events and the mortality rate were comparable between groups. No major impact on the coagulation system was seen in the clinical trials with HBOC-201 [12, 41]. In a study with diaspirin cross-linked Hb (DCLHb, HemAssist®, Baxter, Deerfield, IL, USA) in noncardiac surgical patients who received a maximum of 75 g Hb in three fractions within 36 h perioperatively, the percentage of nontransfused patients was 48% on postoperative day 1 and 23% on postoperative day 7. The median number of transfused allogeneic RBC units was 2 in the DCLHb and 3 in the control group [44]. However, this study was terminated early because of safety concerns which were based on the higher incidence of side effects such as jaundice, urinary side effects and pancreatitis in the DCLHb group. In patients with traumatic hemorrhagic shock mortality after treatment with 500–1,000 ml DCLHb was significantly higher (38 vs. 15% at 48 h after injury) when compared to the control group which received saline infusions [45]. At 28 days after trauma the mortality was 46% in the DCLHb vs. 17% in the saline group, and the 28-day morbidity was 72% higher in the DCLHb group. A subsequent study with moderate doses of 0.025–0.1 g/kg DCLHb in patients with acute stroke revealed a higher incidence of severe adverse events and deaths compared with patients receiving saline [46]. In the light of the re-

266


sults of these studies the company decided to stop further experimental and clinical trials with DCLHb and recombinant human Hb in 1999. The only successful application of HBOC in trauma patients was performed with human polymerized Hb (PolyHeme®, Northfield, Evanston, IL, USA) until now. These patients received a maximum dose of 300 g of PolyHeme (up to 6 units) on the first decision to transfuse. The total number of allogeneic RBC units transfused on day 1 could be significantly reduced to 6.8 ± 3.9 vs. 10.4 ± 4.2 in the control group [47]. No major side effects were associated with the administration of PolyHeme. In conclusion, clinical studies have shown that the perioperative use of different HBOC can reduce the number of allogeneic RBC units and increase the avoidance rate of allogeneic transfusion. Polymerized HBOC appear to have a lower potential of side effects than intramolecurlarly cross-linked preparations. As a consequence, Hemopure and PolyHeme have applied for licence at the FDA last year.

Problems with the Use of HBOC Potential problems with the clinical use of Hb solutions are oxidation of plasmatic Hb, immunogenicity or immunosuppression, and disturbance of photometric measurements. Oxidation leads to an increasing metHb fraction of the plasmatic Hb which does not transport oxygen. However, total metHb concentrations have not reached pathologic concentrations in animals [16] and patients [41] following blood exchange or ANH with HBOC-201. Chemical binding of redox enzymes to cell-free Hb or addition of these enzymes to encapsulated Hb can possibly delay oxidation of plasmatic Hb [48]. Some animal studies showed immune reactions with increased antibody titers after infusion of heterogeneous Hb in different species [49]. Changes of the steric configuration of the Hb

Standl

molecule intend to move the antigen-presenting sites away from the more reactive surface to the less reactive center. As a consequence, no immunoreaction or specific IgG and IgM production was seen after five repetitive injections of human α-α cross-linked Hb in monkeys [50]. HBOC are cleared by the RES in the liver and spleen. In this context, concerns have been raised with regard to possible immunosuppressive side effects of cell-free Hb [51]. However, animal experiments indicated no adverse effects on the course of sepsis and even improved wound healing after treatment with different HBOC [52, 53]. Increased serum iron and ferritin and reticulocytosis were seen in volunteers and patients after infusions with HBOC-201, indicating that cleared cellfree Hb might serve as a source for enhanced erythropoiesis [12, 41]. However, no final conclusions on the immunomodulating side effects of HBOC can actually be drawn. While disturbance of pulse oximetry by intravascular HBOC is within a clinically tolerable range [13], HBOC cause profound interferences with photometric measurements in clinical chemistry. Hematology, blood gases, electrolyte and coagulation measurements are normally not influenced by plasmatic Hb. The BM/Hitachi systems are able to deliver exact values of plasma parameters even in the presence of plasmatic Hb in concentrations up to 2–3 g/dl. However, some enzymes, e.g. amylase and creatine kinase-MB, still remain highly sensitive for interferences with only low plasma Hb concentrations (e.g. 0.2 g/dl).

Possible Indications of HBOC in the Future In spite of some ongoing questions and problems with the use of HBOC, especially systemic vasoconstriction, these products can possibly help to overcome future shortfalls of banked RBCs. HBOC can be used as a bridge during augmented preoperative hemodilution or in case of emergency or intraoperative bleeding until allogeneic RBC units are available. In many developing countries where blood banks do not yet exist HBOC products, which can be stored for years at room temperature, offer a unique opportunity to provide wide-spread supply of oxygen carriers.

The enhanced oxygenation potential in comparison to RBCs, which is provided by the plasmatic oxygen transport of HBOC, can be used for tissues and organs with critically restricted perfusion, e.g. in sickle cell anemia, or in situations where enhancement of a significantly decreased tumor oxygenation is warranted to improve chemo- and radiosensitivity. Poststenotic tissue oxygenation can easily be improved with HBOC which bypass the stenosis in contrast to erythrocytes, e.g. in patients with acute myocardial ischemia or stroke. There is evidence that HBOC-201 can preserve myocardial energy reserve during ischemia for organ preservation [54]. Pulmonary gas exchange can possibly be improved with HBOC entering compressed capillaries in zone I of the lung in critically ill patients [55]. Because of the vasopressor properties (NO scavenging) in combination with improved perfusion and oxygen release on the microcirculatory site, HBOC may particularly be useful in the treatment of sepsis and acute pancreatitis [26, 27]. Animal experiments of septic shock demonstrated hemodynamic stabilization without significant adverse effects on regional organ blood flow using cell-free Hb [56].

Conclusion HBOC have been designed as RBC substitute in emergency situations as well as in clinical settings. Modern chemically modified HBOC are free of toxic side effects on kidneys and coagulation and can replace RBCs in experimental hemorrhagic shock and during complete blood exchange. Because of the plasmatic oxygen transport and increased oxygen off-loading capacity, HBOC improve the diffusive oxygen delivery and appear to possess a higher potential for tissue oxygenation than RBCs. The main side effect of HBOC is NO binding with reactive increase in systemic vascular resistance. In clinical trials HBOC were able to reduce the number of transfused allogeneic RBC units as well as the number of transfused patients. Only bovine polymerized HBOC-201 has been approved for the treatment of anemia in dogs and patients until today.

References 1 Vamvakas EC, Taswell HF: Epidemiology of blood transfusion. Transfusion 1994;34:464–470. 2 Dodd RY: Emerging infections, transfusion safety, and epidemiology. N Engl J Med 2003;349:1205– 1206. 3 Popovsky MA: Transfusion and lung injury. Transfus Clin Biol 2001;8:272–277. 4 Heiss MM, Mempel W, Jauch KW, Delanoff C, Mayer G, Mempel M, Eissner HJ, Schildberg FW: Beneficial effect of autologous blood transfusion on infectious complications after colorectal cancer surgery. Lancet 1993;342:1328–1333. 5 Landers DF, Hill GE, Wong KC, Fox IJ: Blood transfusion-induced immunomodulation. Anesth Analg 1996;82:187–204.


6 Feola M, Simoni J, Dobke M, Canizaro PC: Complement activation and the toxicity of stroma-free hemoglobin solutions in primates. Circ Shock 1988; 25:275–290. 7 Lee R, Atsumi N, Jacobs EE, Austen WG, Vlahakes GJ: Ultrapure, stroma-free, polymerized bovine hemoglobin solution: Evaluation of renal toxicity. J Surg Res 1989;47:407–411. 8 Lipfert B, Standl T, Dullmann J, Helmchen U, Schulte am Esch J, Lorke DE: Histology and ultrastructure of liver and kidney following blood exchange with ultrapurified, polymerised bovine hemoglobin in comparison with hydroxyethyl starch. Lab Invest 1999;79:573–582.

9 Mullon J, Giacoppe G, Clagett C, McCune D, Dillard T: Transfusions of polymerized bovine hemoglobin in a patient with severe autoimmune hemolytic anemia. N Engl J Med 2000;342:1638–1643. 10 Vlahakes GJ, Lee R, Jacobs EE, LaRaia PJ, Austen WG: Hemodynamic effects and oxygen transport properties of a new blood substitute in a model of massive blood replacement. J Thorac Cardiovasc Surg 1990;100:379–388. 11 Standl T, Wilhelm S, Horn EP, Burmeister M, Gundlach M, Schulte am Esch J: Preoperative hemodilution with bovine hemoglobin. Acute hemodynamic effects in liver surgery patients. Anaesthesist 1997;46:763–770.


267

12 Hughes GS, Francom SF, Antal EJ, Adams WJ, Locker PK, Yancey EP, Jacobs EE: Hematologic effects of a novel hemoglobin-based oxygen carrier in normal male and female subjects. J Lab Clin Med 1995;126:444–451. 13 Hughes GS, Antal EJ, Locker PK, Francom SF, Adams WJ, Jacobs EE: Physiology and pharmocokinetics of a novel hemoglobin-based oxygen carrier in human. Crit Care Med 1996;24:756–764. 14 Harringer W, Hodakowski GT, Svizzero T, Jacobs EE, Vlahakes G: Acute effects of massive transfusion of a bovine hemoglobin blood substitute in a canine model of hemorrhagic shock. Eur J Cardiothorac Surg 1992;6:649–654. 15 Siegel JH, Fabian M, Smith JA, Costantino D: Use of recombinant hemoglobin solution in reversing lethal hemorrhagic hypovolemic oxygen debt shock. J Trauma 1997;42:199–212. 16 Standl TG, Reeker W, Redmann G, Kochs E, Werner C, Schulte am Esch J: Haemodynamic changes and skeletal muscle oxygen tension during complete blood exchange with ultrapurified polymerized bovine haemoglobin. Intensive Care Med 1997;23:865–872. 17 Kerger H, Tsai AG, Satzman DJ, Winslow RM, Intaglietta M: Fluid resuscitation with O2 vs. non-O2 carriers after 2 h of hemorrhagic shock in conscious hamsters. Am J Physiol 1997;272:H525–537. 18 Glasgow SC, Shah AS, Noone RB, Gottfried MR, Eachempati SR, Talarico TL, Vaslef SN: Comparison of various hemoglobin polyoxyethylene conjugate solutions as resuscitative fluids after hemorrhagic shock. J Trauma 2000;48:884–892. 19 Slanetz PJ, Lee R, Page R, Jacobs EE, LaRaia PJ, Vlahakes G: Hemoglobin blood substitutes in extended preoperative autologous blood donation: an experimental study. Surgery 1994;115:246–524. 20 Faivre-Fiorina B, Caron A, Fassot C, Fries I, Menu P, Labrude P, Vigneron C: Presence of hemoglobin inside aortic endothelial cells after cell-free hemoglobin administration in guinea pigs. Am J Physiol 1999;276:H766–770. 21 Buehler PW, Mehendale S, Wang H, Xie J, Li M, Trimble CE, Hsia CJC, Gulati A: Resuscitative effects of polynitroxylated αα-cross-linked hemoglobin following severe hemorrhage in the rat. Free Radic Biol Med 2000;29:764–774. 22 Sherman IA, Dlugosz JA, Perelman V, Hsia CJ, Wong LT, Condie RM: Systemic hemodynamic and hepatic microvascular responses to a 33% blood volume exchange with whole blood, stroma-free hemoglobin, and oxypolyhemoglobin solutions. Biomater Artif Cells Immobilization Biotechnol 1993;21:537–551. 23 Gulati A, Sharma AC, Burhop KE: Effect of stroma-free hemoglobin and diaspirin cross-linked hemoglobin on the regional circulation and systemic hemodynamics. Life Sci 1994;55:827–837. 24 Von Dobschuetz E, Hoffmann T, Messmer K: Diaspirin cross-linked hemoglobin effectively restores pancreatic microcirculatory failure in hemorrhagic shock. Anesthesiology 1999;91:1754–1762. 25 Driessen B, Jahr JS, Lurie F, Griffey SM, Gunther RA: Effects of haemoglobin-based oxygen carrier Hemoglobin Glutamer-200 (bovine) on intestinal perfusion and oxygenation in a canine hypovolaemia model. Br J Anaesth 2001;86:683–692. 26 Strate T, Mann O, Kleinhans H, Schneider C, Knoefel WT, Yekebas E, Standl T, Bloechle C, Izbicki JR: Systemic intravenous infusion of bovine hemoglobin significantly reduces microcirculatory dysfunction in experimentally induced pancreatitis in the rat. Ann Surg 2003;238:765–771. 27 Freitag M, Strate T, Schneider C, Knoefel WT, Izbicki J, Standl T, Schulte am Esch J: Isovolemic hemodilution with HES plus bovine hemoglobin HBOC-200 improves tissue oxygen tensions of the pancreas in comparison with HES or Ringer alone in acute severe porcine pancreatitis. AINS 2002;43: 663.

268

28 Budhiraja V, Hellums JD: Effect of hemoglobin polymerization on oxygen transport in hemoglobin solutions. Microvasc Res 2002;64:220–233. 29 Page TC, Light WR, Hellums JD: Prediction of microcirculatory oxygen transport by erythrocyte/hemoglobin solution mixtures. Microvasc Res 1998; 56:113–126. 30 Standl T, Horn P, Wilhelm S, Greim C, Freitag M, Freitag U, Sputtek A, Jacobs EE, Schulte am Esch J: Bovine haemoglobin is more potent than autologous red blood cells in restoring muscular tissue oxygenation after profound isovolaemic haemodilution in dogs. Can J Anaesth 1996;43:714–723. 31 Standl T, Lipfert, Reeker W, Schulte am Esch J, Lorke DE: Acute effects of a complete blood exchange with ultrapurified haemoglobin solution or hydroxyethyl starch on liver and kidney in an animal model. AINS 1996;31:354–361. 32 Standl T, Freitag M, Burmeister MA, Horn EP, Wilhelm S, Schulte am Esch J: Hemoglobin-based oxygen carrier HBOC-201 provides higher and faster increase in oxygen tension in skeletal muscle of anemic dogs than do stored red blood cells. J Vasc Surg 2003;37:859–865. 33 Horn EP, Standl T, Wilhelm S, Jacobs EE, Freitag U, Freitag M, Schulte am Esch J: Bovine hemoglobin increases skeletal muscle oxygenation during 95% artificial arterial stenosis. Surgery 1997;121: 411–418. 34 Standl T, Horn EP, Burmeister MA, Bangert K, Reiter B, Jacobs EE, Schulte am Esch J: Bovine hemoglobin restores myocardial tissue tension in dogs with acute critical coronary stenosis and extended hemodilution. Anesthesiology 1999;90:A697. 35 Habler O, Kleen M, Pape A, Meisner F, Kemming G, Messmer K: Diaspirin-crosslinked hemoglobin reduces mortality of severe hemorrhagic shock in pigs with critical coronary stenosis. Crit Care Med 2000;28:1889–1898. 36 Rempf C, Burmeister MA, Bartsch-Zwemke S, Rehberg S, Standl T, Schulte am Esch J: Effect of prophylactic or therapeutic application of cell-free hemoglobin HBOC-200 on the severity of cardiac arrhythmias during myocardial ischemia-reperfusion in rats. Eur J Anaesthesiol 2003;20(suppl):A9. 37 Burmeister MA, Rempf C, Krause T, Tuszynski S, Standl T, Schulte am Esch J: Effect of prophylactic and therapeutic application of the hemoglobin solution HBOC-200 on the infarct size after myocardial ischemia and reperfusion in rats. AINS 2003; 44:744. 38 Cole DJ, Schell RM, Drummond JC, Reynolds L: Focal cerebral ischemia in rats. Effect of hypervolemic hemodilution with diaspirin cross-linked hemoglobin versus albumin on brain injury and edema. Anesthesiology 1993;78:335–342. 39 Kasper SM, Walter M, Grune F, Bischoff A, Erasmi H, Buzello W: Effects of a hemoglobin-based oxygen carrier (HBOC-201) on hemodynamics and oxygen transport in patients undergoing preoperative hemodilution for elective abdominal aortic surgery. Anesth Analg 1996;83:921–927. 40 Brauer P, Standl T, Wilhelm S, Burmeister MA, Schulte am Esch J: Transcranial Doppler sonography mean flow velocity during infusion of ultrapurified bovine hemoglobin. J Neurosurg Anesthesiol 1998;10:146–152. 41 Standl T, Burmeister MA, Horn EP, Wilhelm S, Knoefel WT, Schulte am Esch J: Bovine haemoglobin-based oxygen carrier for patients undergoing haemodilution before liver resection. Br J Anaesth 1998;80:189–194. 42 LaMuraglia GM, O’Hara PJ, Baker WH, Naslund TC, Norris EJ, Li J, Vandermeersch E: The reduction of the allogeneic transfusion requirement in aortic surgery with a hemoglobin-based solution. J Vasc Surg 2000;31:299–308.


43 Levy JH, Goodnough LT, Greilich PE, Parr GVS, Stewart RW, Gratz I, Wahr J, Williams J, Comunale ME, Doblar D, Silvay G, Cohen M, Jahr J, Vlahakes G: Polymerized bovine hemoglobin solution as a replacement for allogeneic red blood cell transfusion after cardiac surgery: Results of a randomized, double-blind trial. J Thorac Cardiovasc Surg 2002;124:35–42. 44 Schubert A, Przybelski RJ, Eidt JF, Lasky LC, Marks KE, Karafa M, Novick AC, O’Hara JF; Saunders ME, Blue JW, Tetzlaff JE, Mascha E: Diaspirin-crosslinked hemoglobin reduces blood transfusion in noncardiac surgery: A multicenter, randomized, controlled, double-blinded trial. Anesth Analg 2003;97:323–332. 45 Sloan EP, Koenigsberg M, Gens D, Cipolle M, Runge J, Mallory MN, Rodman G Jr: Diaspirin cross-linked hemoglobin (DCLHb) in the treatment of severe traumatic hemorrhagic shock: A randomized controlled efficacy trial. JAMA 1999; 282:1857–1864. 46 Saxena R, Wijnhoud AD, Carton H, Hacke W, Kaste M, Przybelski RJ, Stern KN, Koustaal PJ: Controlled safety study of a hemoglobin-based oxygen carrier, DCLHb, in acute ischemic stroke. Stroke 1999;30:993–996. 47 Gould SA, Moore EE, Hoyt DB, Burch JM, Haenel JB, Garcia J, DeWoskin R, Moss GS: The first randomized trial of human polymerized hemoglobin as a blood substitute in acute trauma and emergent surgery. J Am Coll Surg 1998;187:113– 122. 48 D’Agnillo F, Chang TM: Polyhemoglobin-superoxide dismutase-catalase as a blood substitute with antioxidant properties. Nat Biotechnol 1998;16: 667–671. 49 Hertzman CM, Keipert PE, Chang TMS: Serum antibody titers in rats receiving repeated small subcutaneous injections of hemoglobin or polyhemoglobin: A preliminary report. Int J Artif Organs 1986;9:179–182. 50 Estep TN, Gonder J, Bornstein I, Aono F: Immunogenicity of diaspirin cross-linked human hemoglobin solutions. Biomater Artif Cells Immobilization Biotechnol 1992;20:603–609. 51 White CT, Murray LJ, Smith DJ, Greene JR, Bolin RB: Synergistic toxicity of endotoxin and hemoglobin. J Lab Clin Med 1986;108:132–137. 52 Langermans JAM, Vuren-van der Hulst MEB, Bleeker WK: Safety evaluation of a modified hemoglobin solution (PolyHbXL) in a murine infection model. Artif Cells Blood Substit Immobil Biotechnol 1996;24:376. 53 Soltero R, Leppaniemi A, Pikoulis E, Ratigan J, Waasdorp C, Malcolm D: Diaspirin crosslinked hemoglobin (DCLHb) increases rat skin wound breaking strength after hemorrhage. Artif Cells Blood Substit Immobil Biotechnol 1996;24:433. 54 Carlucci F, Miraldi F, Barretta A, Marullo AGM, Marinello E, Tabucchi A: Preservation of myocardial energy status by bovine hemoglobin solutions during ischemia. Biomed Pharmacother 2002;56: 247–253. 55 Conhaim RL, Rodenkirch LA, Watson KE, Harms BA: Acellular hemoglobin solution enters compressed lung capillaries more readily than red blood cells. J Appl Physiol 2000;89:1198–1204. 56 Bone HG, Schenarts PJ, Fischer SR, McGuire R, Traber LD, Traber DL: Pyridoxalated hemoglobin polyoxyethylene conjugate reverses hyperdynamic circulation in septic sheep. J Appl Physiol 1998;84: 1991–1999.

Standl



Haemoglobin Hyperpolymers, a New Type of Artificial Oxygen Carrier – Concept and Current State of Development* W.K.R. Barnikol

H. Poetzschke

Bereich Klinische Physiologie, Universität Witten/Herdecke, Witten (Ruhr), Germany

Key Words Haemoglobin hyperpolymers ⋅ Artificial oxygen carrier ⋅ Blood additive ⋅ Oxygen transport ⋅ Drug development ⋅ Plasmatic

Schlüsselwörter Hämoglobin-Hyperpolymere ⋅ Künstliche Sauerstoffträger ⋅ Blut-Additiv ⋅ Sauerstofftransport ⋅ Arzneimittelentwicklung ⋅ Plasmatisch

Summary In the clinical setting, artificial oxygen carriers are needed when a patient has a tissue oxygen deficiency which he/she cannot automatically compensate. There are two quite different situations where this might occur: i) Heavy blood loss (e.g. following an accident) and ii) insufficient perfusion (e.g. as a result of arteriosclerosis or myocardial infarction) or anaemia without blood loss. In the first instance, an iso-oncotic oxygentransporting plasma expander is required, whereas in the second instance a (hypo-oncotic) so-called blood additive is needed. This 2nd type of situation also presents the greater range of very important indications. Experimental work has shown that, in comparison to erythrocytes, dissolved haemoglobin is able to release oxygen more rapidly (effective plasmatic transport) while at the same time also facilitating oxygen release from erythrocytes (mediator function). Blood additives occur naturally in lower forms of life (e.g. earthworm) where they can be found in the form of giant oxygen-carrying molecules. Using these natural forms as a basis, new oxygen-transporting blood additives were designed and developed (so-called haemoglobin hyperpolymers) which exhibit a strong oxygen affinity (half saturation partial pressure p50 = 16 mm Hg) and high co-operativity (n50 = 2.1). One product has, up to now, been produced aseptically on a small technical scale and consists of highly purified, polymerised and pegylated porcine haemoglobin which is free of monomers and oligomers (mean molecular weight approximately 800 kDa). It is sufficiently low in endotoxin (< 0.029 EU/ml) and blood plasma compatible and – at an effective concentration of 3 g/dl in blood plasma – causes only minor increases in oncotic pressure or viscosity. The product has a shelf life of up to 2 years and is administered as a carbonyl derivative. Its intravascular half-life in the conscious rat is 30 h. This product was found to prevent death in rats in wchich acute lung injury was induced using oleic acid. In human self-experiments this product was repeatedly administered: No effects on blood pressure and heart rate, no increase in blood transaminase concentration and no immunological reaction were seen; the latter was also true for selected sensitive mice. Furthermore, the blood additive is universally applicable as an oxygen transporter since, when mixed with a conventional plasma expander, it can also be used to treat an oxygen deficiency occurring together with blood loss.

Zusammenfassung Künstliche Sauerstoffträger werden klinisch benötigt, wenn ein Patient in einen geweblichen Sauerstoffmangel gerät, den er nicht kompensieren kann. Es sind zwei Fälle zu unterscheiden: 1) der starke Blutverlust (z.B. nach Unfall) und 2) die Minderperfusion (z.B. bei Arteriosklerose oder Herzinfarkt) oder Anämie ohne Blutverluste. Im ersten Fall wird ein isonkotischer Sauerstoff-transportierender Plasma-Expander gebraucht, im zweiten ein (hyponkotisches) so genanntes Blut-Additiv. Der zweite Fall stellt den größeren Indikationsbereich. Gewisse Experimente haben gezeigt, dass gelöstes Hämoglobin Sauerstoff schneller abgibt als Erythrozyten (wirksamer plasmatischer Transport) und dass es auch die Freisetzung des Sauerstoffs aus Erythrozyten erleichtert (Mediator-Wirkung). Das Prinzip des Blut-Additivs findet sich in der Natur bei niederen Tieren (z.B. Regenwurm) in Form riesiger, molekular proteinärer Sauerstoffträger. Nach diesem natürlichen Prinzip wurde das neue Sauerstofftransportierende Blut-Additiv (so genannte Hämoglobin-Hyperpolymere: HP3Hb) mit hoher Sauerstoff-Affinität (Halbsättigungsdruck p50 = 16 mm Hg) und hoher Kooperativität (n50 = 2,1) konzipiert und entwickelt. Das bisher im kleinen technischen Maßstab (aseptisch unter GMP-ähnlichen Bedingungen) hergestellte Produkt besteht aus polymerisiertem und pegyliertem hochreinem Schweine-Hämoglobin ohne Monomere und Oligomere, sein mittleres Molekulargewicht beträgt etwa 800 kDa. Es ist hinreichend Endotoxin-arm (< 0,029 EU/ml), mit Blut-Plasma verträglich und erhöht bei einer Wirk-Konzentration von 3 g/dl im Blut-Plasma sowohl dessen onkotischen Druck, als auch die Viskosität nur wenig. Das Produkt ist bis zu 2 Jahren lagerstabil und wird als Carbonyl-Derivat verabreicht. Die Plasma-Halbwertszeit in wachen Ratten beträgt 30 h. Es bewahrt Ratten im Ölsäure-induzierten Lungenschock vor dem Tod. Einzelne Selbstversuche Freiwilliger mit mehrfacher Anwendung ergaben keinen Einfluss auf Blutdruck und Herzfrequenz, keine Erhöhung der Transaminase-Konzentration und keine Immunreaktionen; letzteres war auch an ausgesucht sensiblen Mäusen nicht nachweisbar. Darüber hinaus ist das Blut-Additiv als künstlicher Sauerstoff-Träger universell, denn es kann, und zwar gemischt mit einem (herkömmlichen) Plasma-Expander, auch zur Therapie eines Sauerstoff-Mangels mit Blutverlust Verwendung finden.




Prof. Dr. Dr. W.K.R. Barnikol Bereich Klinische Physiologie, Private Universität Witten/Herdecke D-58448 Witten (Ruhr) Tel. +49 23 02 915-203, Fax -191 E-mail [email protected]

Introduction The need for artificial oxygen carriers for medical use has been recognised for many years, and numerous attempts have been made to investigate the necessary basic scientific foundations, to develop concepts and to finally produce clinically applicable artificial oxygen carriers. The most important developments in the area of artificial, medically applicable haemoglobin derivatives are (per-)fluorocarbons, haemosomes (liposomes filled with haemoglobin), and molecularly dispersed soluble haemoglobin derivatives [1–3]. Pharmaceutical developments have, until now, been focused on artificial oxygen carriers which can be used to substitute for lost blood or on heterologous blood transfusion necessary for blood loss therapy. This focus is evident in the earlier common use of the term ‘blood substitute’ for such preparations, which emphasises these substances as means of volume replacement that are also capable of carrying oxygen. Low tissue oxygen levels occurring together with blood loss need to be treated not only in terms of missing blood volume (using a plasma expander) but also in terms of the lacking oxygen transport capacity (using an artificial oxygen carrier). The combination of these demands thus requires an oxygen-carrying plasma replacement agent. At present two such preparations (in the form of haemoglobin derivatives dissolved in plasma) have been approved in some countries (Russia and South Africa), with several further preparations currently in late phases of clinical development. However, low tissue oxygen levels can also occur without concomitant blood loss due to a number of different pathophysiological mechanisms. These include low perfusion due to vascular diseases, thrombosis or emboli, circulatory shock, shock organs and (chronic) anaemia. Low tissue oxygen levels which arise without associated blood loss require solely an augmentation of the oxygen-carrying capacity of the already available blood. This could be in the form of an artificial oxygen carrier which can be mixed into the blood as a ‘blood additive’. In figure 1, these two clinical situations are schematically shown in terms of the changes in blood volume which occur. This contribution examines the principal conceptual differences between the various artificial oxygen carriers and also describes the present state of development of such products. To our knowledge, the concept of a blood additive has not been described previously. Both types of artificial oxygen carriers have their own spectrum of possible indications. Heavy blood loss due to accidents, natural catastrophes, acts of violence, wars, etc. requires therapy with an oxygen-carrying plasma replacement product. ‘Other’ states of low tissue oxygen occurring without blood loss – for example shock lung, myocardial infarction, stroke, arterial occlusion, shock kidney, tinnitus, sudden deafness, visual disorder, vertigo, placental insufficiency or anaemia – require an oxygen-carrying blood additive to improve oxygen supply. Its use to enhance tumour oxygenation pre-therapeuti-

270


Fig. 1. Schematic representation of volume changes and therapy of tissue oxygen deprivation with an oxygen carrier dissolved in blood plasma following a heavy blood loss (*) and b oxygen shortage without blood loss (**, e.g., insufficient perfusion). S = Oxygen-transporting plasma expander (‘blood substitute’); A = oxygen-transporting blood additive.

cally could be a further application. The range of indications includes a number of widely occurring civilisation diseases. States of oxygen deficiency occurring without blood loss account for the largest number of indications both quantitatively and qualitatively.

Efficiency and Advantages of Plasma-Dissolved Haemoglobin–Based Artificial Oxygen Carriers An oxygen carrier – whether naturally occurring or artificial – needs to be effective in terms of being able to take up sufficient oxygen from the alveoli and releasing it in the tissue capillaries, irrespective of the prevailing conditions. For this, a rapid release of oxygen from the different preparations of haemoglobin added to blood is vital. Possible preparations in this context include erythrocytes, haemosomes and free haemoglobin solutions. The rate of oxygen release can be assessed using a rapid-mix chamber (stopped-flow technique). Such experiments were performed by Farmer et al. [4] and showed that the de-saturation of small haemosomes (0.2 µm) and freely dissolved haemoglobin takes place 11 and 14 times faster respectively than that found in erythrocytes. The form in which haemoglobin is present is thus of great importance for the efficiency of oxygen release; the most rapid release of oxygen is found with freely dissolved haemoglobin.

Barnikol/Poetzschke

The Authors’ Concept of a Blood Additive Over the past years a number of investigations have led to the recognition of the need for the development of a well-tolerable and efficient haemoglobin-based, artificial, oxygen-carrying blood additive. A basic demand on a blood additive is that it must be introducible into the existing blood without causing a noticeable increase in the in vivo blood volume. The blood plasma volume and the distribution of the various body liquids

Haemoglobin Hyperpolymers, a New Type of Artificial Oxygen Carrier

pO2, mm Hg

Page et al. [5] simulated the geometric conditions found in the microcirculation and photometrically measured the diffusional release of oxygen during flow through an artificial capillary. The capillary was made of silicon, the oxygen content of which was reduced to zero by flushing with nitrogen. In this way, the capillary material, rather than the surrounding cells, acted as an oxygen sink. Oxygen release curves were constructed on the basis of experimentally obtained values, thus showing the dependence of the oxygen saturation of haemoglobin as a function of its duration in the artificial capillaries. The bovine haemoglobin content of all liquids measured was the same, i.e. 10.2 g/dl. Dissolved haemoglobin releases oxygen considerably faster than erythrocytes. If both preparations are mixed with a ratio of 1:1, the oxygen is released just as rapidly as it is from a pure haemoglobin solution. The finding that more than 50% of the oxygen can be released indicates that the dissolved bovine haemoglobin can act as a mediator for the release of oxygen from the erythrocytes. This is just as effective as were the haemoglobin in the erythrocytes also freely dissolved, showing the high effectiveness of the mediator function of dissolved haemoglobin. When freely dissolved haemoglobin and erythrocytes are mixed with a ratio of 1:9 (i.e. 10% of the oxygen carrier is freely dissolved in the plasma), then the rate of oxygen release increases greatly as compared with red blood cells. Therefore, in terms of oxygen release, haemoglobins dissolved in plasma appear to exhibit two very favourable properties: firstly, the capability of efficient plasmatic transport and secondly, the mediator effect. Besides their transport efficiency properties, artificial oxygen carriers dissolved in plasma demonstrate a number of other benefits, including a long shelf life, blood group independency and a lack of potential infection problems such as hepatitis or acquired immunodeficiency syndrome (AIDS). In addition, oxygen carriers dissolved in plasma are not dependent on the availability of blood donations, can pass through constricted blood vessels and present only a negligible burden on the reticuloendothelial system. Haemoglobin-based oxygen carriers are also suitable for use in experimental organ perfusion or for the preservation of transplantable organs. A further application for such carriers is in the provision of oxygen to cells growing in culture, with the carrier dissolved in the culture medium.

0.5

1.0

Fig. 2. A hyperbolic (dashed line) and a sigmoidal (continuous line) oxygen binding curve with the same affinity. The curves are shown as oxygen release curves. pO2 = Oxygen partial pressure; SO2 = saturation of the oxygen binding sites; (GO2 = proportional oxygen content of the solution); A = alveolus/alveolar; L = lungs; X = peripheral tissue; BE = Bohr effect.

in the fluid compartments of an organism are determined by the oncotic (colloid-osmotic) pressure of the blood plasma which, being a colligative property, is in turn proportional to the number of dissolved macromolecules. A blood additive should therefore be strongly hypo-oncotic (principle of hypooncosis) in comparison to blood plasma so that it does not cause any considerable increases in the plasma oncotic pressure upon application. Thus, in order to keep the number of administered molecules of a haemoglobin derivative-based blood additive low, the molecular mass of the derivative needs to be as great as possible. This can be achieved by an intermolecular networking (polymerisation) of native haemoglobin molecules. A further very important property is an optimal oxygen affinity, which is not even found in normal blood which exhibits a half-saturation pressure (p50) of approximately 25 mm Hg. Investigations carried out by Conover et al. [6] showed that the provision of oxygen to organs by means of plasmatic oxygen carriers occurs best when these agents exhibit a half-saturation pressure of between 15 und 20 mm Hg. The oxygen-transporting blood additive should be able to compensate for a poor tissue oxygenation whereby it is probable that the haemoglobin saturation will need to fall below 50%, as suggested by measurements of tissue oxygen partial pressure below 25 mm Hg (already in normal tissue, see e.g. [7–9]). In such instances, a highly cooperative oxygen binding is necessary to achieve the desired effect, i.e. a pronounced sigmoidality of the binding curve. This requirement can be seen in figure 2.


271

Fig. 3. Schematic representation of the manufacturing process for the hyperpolymer, divided into three main synthesis and modification steps. a Using native haemoglobin as the basis, a chemical polymerisation (1) is carried out which results in b polymerised haemoglobin from which, by means of molecular concealment (2), c a masked, polymerised product is obtained. Finally, this product is separated (3) into d monomers and oligomers as well as (hyper-)polymers. A = blood additive); S = blood substitute.

The partial oxygen pressure in the oxygen release curve is likewise functionally the intracapillary driving force for the diffusion of oxygen from the capillary to the tissue cells. As a result of the S-form of the binding curve, the blood in the lungs is almost completely saturated at an alveolar oxygen partial pressure of only 60 mm Hg. At the same time, at a desaturation down to 25%, the capillary oxygen partial pressure still remains above 20 mm Hg. Both of these cannot however be provided by the hyperbolic binding curve. Rather, a greater capillary oxygen partial pressure (as the driving force for the diffusion of oxygen from the capillaries into the tissue) is indeed necessary if hypoxic tissue is to be sufficiently supplied with oxygen. Further demands on the additive relate to its use as an oxygen-transporting blood volume replacement agent (together with a plasma expander), and its compatibility with human blood plasma. In principle, a preparation with low oncotic pressure is favourable; a low viscosity and a long vascular halflife are desirable as well. The additive should not be excreted via the kidneys or leave the blood vessels (no extravasation). Immunotolerance is also a necessity. With regard to an eventual commercial application of a blood additive, a reasonably priced manufacturing process is desirable so that the final medicinal product is not too expensive. Factors which can influence the manufacturing costs are the simplicity (or complexity) of the technical organisation of the apparatus and processes together with the material and running costs. An oxygen-transporting blood additive which is soluble in plasma, and complies with the demands outlined above must consist of molecules possessing a very large number (hundreds) of binding sites for oxygen, in other words, ‘molecular erythrocytes’ [10–12]. Naturally occurring oxygen-transporting additives do exist. Indeed, lower forms of life such as snails and worms possess giant proteins with molecular weights of 8.4 and 3.6 MDa respectively, known as haemocyanin and erythrocruorin. The latter possesses approximately 200 binding

272


sites for oxygen. Experimental work has shown that, under physiological conditions, erythrocruorin from the earthworm shows p50 of 9 mm Hg and a cooperativity (n50) of 12 [13]. The functioning, naturally occurring examples provide conceptual security for the development of an artificial oxygentransporting blood additive: Suitable synthesised highly molecular oxygen-binding substances will most certainly also be able to deliver oxygen to the cells in sufficient quantities, particularly considering the fact that the earthworm possesses a capillary system as well. The goal is therefore the development of a hypo-oncotic oxygen carrier with an enhanced oxygen affinity and a high cooperativity, which is able to fulfil the already described demands. Since earthworm erythrocruorin (just as human haemoglobin) consists of haem-carrying subunits with a molecular weight of 15 kDa, the goal is evident: Mammalian haemoglobin should be networked into giant molecules under defined conditions and then masked to reduce molecular interactions. An optimal starting material is of course human haemoglobin. Its availability is however limited although gene technological production at a moderate price may be possible in the future. Our choice was therefore porcine haemoglobin (PHb) since it is freely available and very similar to human haemoglobin, both functionally and structurally. It has the same effector (2,3-diphosphoglycerate) and shows the same characteristics for oxygen binding. Additionally, the problem of BSE (bovine spongiform encephalopathy) is not an issue. Because of the pronounced similarities of human haemoglobine and PHb, a single manufacturing process can be used for both of these forms of haemoglobin. The process itself should be uncomplicated and reasonably priced. Figure 3 shows a schematic representation of the manufacturing process. The process basically involves three steps. In the first of these, a covalent network is formed from the haemoglobin molecules. Here, the central cavity of the haemoglobin molecule is reversibly blocked with inositol hexaphosphate. The products of this process exhibit a very broad distribution of molecular

Barnikol/Poetzschke

Table 1. Characteristic values for the oxygen binding of human blood and aqueous solutions of various haemoglobins at 37 °C, pCO2 40 mm Hg and pH 7.4a

Fig. 4. Transmission electron bright field micrograph of a very large single hyperpolymeric haemoglobin.

weights, ranging from monomer haemoglobin to giant molecules. The second step is the concealment of the networked products by covalent binding of polyethylene glycol. This is performed to prevent the occurrence of unwanted interactions with blood plasma proteins, especially with larger molecules such as fibrinogen and certain antibodies. Both of these processes can take place in a single reactor, without the need for intermediate steps so that the manufacturing process is considerably simplified and reasonably priced. The production output is approximately 70%. The third step is an ultrafiltration (using filters with a nominal molecular weight separation limit of 1,000 kDa) which separates the concealed, networked product into two approximately equal parts: a low and high molecular fraction. The low molecular fraction is a high-quality ‘waste product’ since it can be used – for correspondingly appropriate indications – as an oxygen-transporting plasma expander (volume replacement) to treat low oxygen levels associated with blood loss. The second fraction (also approximately 35%) is hyperpolymer haemoglobin, which can be used as blood additive to alleviate low oxygen levels in the many instances where these occur without accompanying blood loss.

Characteristics of the Blood Additive – Actual Drug Developments The production of haemoglobin hyperpolymers has been optimally developed and standardised for aseptic production on a small technical scale. The median molecular weight of the product as produced under these optimised conditions is of 800 kDa, with 10 and 90% percentiles of approximately 180 and 3,600 kDa. The final product thus contains practically no haemoglobin in monomer or low oligomer forms. The endotoxin activity of the preparation was found to be less than 0.029 EU/ml, (the lower detection limit of the endotoxin test). Figure 4 shows an electron micrograph of a very large haemoglobin hyperpolymer molecule. The picture gives the impression that the hyperpolymer is very bulky and that the molecule is not as compact as a globular protein. In accordance


Haemoglobin / derivative

p50, mm Hg

n50

Human blood Human haemoglobin PHb HP3Hb

25 15 15 16

2.5 2.5 2.5 2.1

aMeasurements

with the native haemoglobins and the haemoglobin hyperpolymers (HP3Hb) were performed in the absence of organic phosphate, especially 2,3-bisphosphoglycerate and adenosine triphosphate.

with this, the polymer chemical Mark-Houwink coefficient for the hyperpolymer in aqueous solution is 0.38 [14]. Characteristic values for the oxygen binding of the artificial carrier together with those for other haemoglobin preparations are given in table 1. The values provided are the p50 as a measure of the mean oxygen affinity, and the n50 (Hill coefficient) which provides an indication of the sigmoidality of the oxygen binding curve; an n50 of 1 signifies a lack of sigmoidality (and cooperativity), and instead is suggestive of a binding curve with a hyperbolic form. When viewing table 1, it is evident that human haemoglobin and PHb are functionally very similar. They show the same characteristic values for oxygen bindung (p50 and n50). Additionally, the primary structures of these substances are very similar [15]. The synthesised blood additive exhibits a p50 value of 16 mm Hg and a cooperativity of 2.1, which, according to Conover et al. [6], results in optimised conditions for oxygen release. An effective concentration of the carrier of approximately 3 g/dl in blood plasma is reasonable. In comparison, the content of native haemoglobin in blood is 15 g/dl so that, if the blood cell volume is also taken into account, the content of the artificial additive would be only 1.5 g/dl, i.e. only one tenth of the whole blood volume. Viscosity measurements in an aqueous electrolyte (AE) solution and in fresh blood plasma are shown in figure 5. Not only is the cinematic viscosity for the artificial additive, as measured in fresh blood plasma, greater than that measured in the electrolyte but also the increase seen with greater concentrations is intensified. The dependency is in both cases not linear. Even at the intended therapeutic concentration of 3 g/dl, the viscosity remains below that of blood. Data from colloid osmotic (oncotic) pressure measurements obtained under similar conditions with respect to solutes as those used in viscosity measurements are presented in figure 6. In contrast to the viscosity-related findings, with colloid osmotic pressure almost linear dependencies were seen. However, when the additive is dissolved in fresh human blood plasma, the increase in colloid osmotic pressure caused by the ad-


273

cmHb, g/l

Q/cSt

cmHb, g/l

cmHb, g/l Fig. 6. The colloid-osmotic pressure (πonc) of solutions of the blood additive (HP3Hb) in an AE solution and in fresh human plasma as a function of haemoglobin content (cmHb) at room temperature. πoncBL = Oncotic pressure of blood (particularly of plasma).

ditive is greater than that obtained in an electrolyte solution. At the designated therapeutic concentration, the increase seen in the blood plasma oncotic pressure is approximately 12 mbar. It is imperative that the artificial oxygen carrier is compatible with fresh human plasma over the relevant physiological pH range from 6.9 to 7.7. This was verified by precipitation experiments. The results obtained with two preparations at different concentrations are shown in figure 7. The concentrations in the supernatant of the centrifuged mixture showed no pHdependent changes over a pH range from 6.5 to 8.2. Considering the above-mentioned properties, the carrier appears to be suitable for an application to an organism, for example in small portions via venous infusion.

274

Fig. 7. Assessment of the compatibility of the blood additive when mixed with fresh human plasma. The concentration of the haemoglobin hyperpolymer (cmHb) in the supernatant after standardised incubation is shown as a function of proton activity (as indicated by pH).

cPL, g/l

Sonc, mbar

Fig. 5. Cinematic viscosity (ν) of solutions of the blood additive (HP3Hb) in AE solution and in fresh blood plasma as a function of haemoglobin content at 37 °C, cmHb = Haemoglobin concentration; νBL = cinematic viscosity of blood.


t, h Fig. 8. Course of the decrease in the concentration of the blood additive in blood plasma (cPL) in vivo as measured in conscious rats receiving different initial additive concentrations. t = Time after additive administration; 1/2= half of the initial concentration.

The length of time over which the carrier stays in the organism’s blood plasma is also of great importance, a property which can be characterised using the plasma half-life. This was assessed in conscious rats. The results of these experiments are shown in figure 8. It is evident that the plasmatic half-life increases with rising initial concentrations. At a concentration of 2.2 g/dl, the half-life is 18 h, at 3.0 g/dl it is almost 30 h (and at 1 g/dl approximately 9 h). Similar measurements performed in mice indicated a half-life of 15 h with an initial blood plasma concentration of the carrier of 3 g/dl. Mice are metabolically more active than rats which are in turn metabolically more active than humans. The efficacy of the blood additive was assessed following the induction of acute lung injury (or shock lung) using oleic acid

Barnikol/Poetzschke

BP, mm Hg HR, min–1

Survival time, h

a

t, min

Animal number

in barbiturate-anaesthetised rats [16]. As shown in figure 9, the amount of oleic acid applied intravenously was chosen in a way that animals not receiving the carrier died after approximately 3 h (control group). All animals of the verum group, which were treated with the additive (initial concentration in the plasma approximately 2.5 g/dl) survived the acute lung injury and were euthanised after approximately 7 h. Thus, the blood additive was able to protect these animals from death due to acute lung injury. In voluntary exploratory self-experiments the new artificial carrier was administered in small doses to humans. Of special interest was if the additive would cause an increase in blood pressure. Figure 10 shows the course of blood pressure changes during this experiment in 2 persons. In the male person, the blood pressure rose from 140 to 165 mm Hg over the period (approximately 40 min) prior to administration of the additive. Subsequent to the administration of the carrier (1.16 g), the blood pressure remained constant. The diastolic blood pressure and pulse rate were also found to be constant over the whole measurement period. The female person undergoing the same self-experiment showed constant values for all parameters. Thus, in the cases shown, administration of the oxygen-transporting blood additive had no effect on blood pressure and pulse rate. A further interesting and important question is whether the additive, when dissolved in plasma, causes damage to the liver. The liver has an open vascular system, so that the carrier can come into direct contact with hepatocytes. One index of liver damage is an increase in blood levels of the transaminases GOT and GPT. However, when administering the artificial oxygen carrier three times with 14-day intervals between the applications, no significant changes in the blood transaminase levels could be found (fig. 11). In order to test the immunogenicity of the oxygen carrier, self-experiments were done using doses of 0.6–2.4 g applied


BP, mm Hg HR, min–1

Fig. 9. Demonstration of effectiveness: The survival time of anaesthetised rats following oleic acid-induced acute lung injury a without treatment (control) and b with administration of the blood additive (treatment).

b

t, min

Fig. 10. Systolic (syst.) and diastolic (diast.) blood pressure (BP) and heart rate (HR) as a function of the observation time (t) following intravenous administration (in each instance 1.16 g) of the blood additive (HP3Hb) in humans. a Volunteer P1, 70 kg body weight, b volunteer P2, 65 kg body weight.

3 times with 14-day intervals. After the 3rd application, neither of the persons showed any clinical signs of immunological reactions. The immunogenicity of the blood additive, together with that of monomeric PHb was also assessed in mice of the C3H/HeN strain which are naturally highly sensitive to haemoglobin. The immunisation scheme comprised of 3 applications each 1 week apart. After the 3rd application of native haemoglobin, a very low titre was seen (about 1:200), with no symptoms of shock being observed in the mice. With the blood additive, the titres for the three categories of immunoglobulins assessed were found to be slightly increased (to about 1:400), and the animals sat still for approximately 15–20 min after the 3rd application. A further important criterion for the clinical usefulness of the blood additive is that it should be storable in a simple fashion for a sufficiently long period of time. The structural stability is not a limiting factor as the hyperpolymers were found to be stable for a number of months, even at a storage temperature of 45 °C, with no signs of molecular alteration or turbidity


275

cPL, U/l Fig. 11. Transaminase levels (cPL) before (a) and after (p) intravenous administration of different amounts of the carrier in two volunteers (P1 and P2). P1 70 kg body weight, 1 = 0.58 g, 2 = 2.32 g. P2 65 kg body weight, 1, 2, 3 = 1.16 g each. Shaded columns = GPT, empty columns = GOT.

being seen. At room temperature (22 °C), the stability could, until now, be verified for a period of 9 months. The irreversible conversion of (reduced) functional haemoglobin into oxidised methaemoglobin (haemiglobin), which is incapable of oxygen transport in blood, can be effectively reduced (approximately 100 times) by ligandation of the oxygen binding sites with carbon monoxide. This results in a daily increase in the portion of methaemoglobin of less than 0.001% at 5 °C and of approximately 0.01% at 22 °C (

Autologous Transfusion - From Euphoria to Reason: Clinical Practice Based on scientific knowledge - Nottwil, January 16-17, 2004: Proceedings

Scientific American January 2004

Sound on Sound (January 2004)

Clinical Transfusion Medicine (Vademecum)

Knowledge-Based Bioinformatics: From analysis to interpretation

Autologous Fat Transfer: Art, Science, and Clinical Practice

Autologous Fat Transfer: Art, Science, and Clinical Practice

Autologous Fat Transfer: Art, Science, and Clinical Practice

Transfusion Medicine: A Clinical Guide

Knowledge-Based Software Engineering: Proceedings of the Fifth Joint Conference on Knowledge-Based Software Engineering

PHP Magazin (January 2004)

Scientific American (January 2007)

Scientific American - January 2011

Psychoneuroendocrinology: The Scientific Basis of Clinical Practice

Evidence-Based Practice: Toward Optimizing Clinical Outcomes

Practice-based Evidence for Healthcare: Clinical mindlines

Evidence-Based Clinical Supervision: Principles and Practice

Handbook of Community-Based Clinical Practice

Case Based Echocardiography: Fundamentals and Clinical Practice

Evidence-Based Practice: Toward Optimizing Clinical Outcomes

Scientific american (June 2004)

Scientific american (October 2004)

Introduction to Vascular Biology: From Basic Science to Clinical Practice

Introduction to Vascular Biology: From Basic Science to Clinical Practice

Scientific american (May 2004)

Scientific american (December 2004)

Scientific american (November 2004)

Scientific american (September 2004)

Knowledge-based clustering: from data to information granules

Knowledge-Based Clustering: From Data to Information Granules

Scientific american (March 2004)

Autologous Transfusion - From Euphoria to Reason: Clinical Practice Based on scientific knowledge - Nottwil, January 16-17, 2004: Proceedings

Scientific American January 2004

Sound on Sound (January 2004)

Clinical Transfusion Medicine (Vademecum)

Knowledge-Based Bioinformatics: From analysis to interpretation

Autologous Fat Transfer: Art, Science, and Clinical Practice

Autologous Fat Transfer: Art, Science, and Clinical Practice

Autologous Fat Transfer: Art, Science, and Clinical Practice

Transfusion Medicine: A Clinical Guide

Knowledge-Based Software Engineering: Proceedings of the Fifth Joint Conference on Knowledge-Based Software Engineering

PHP Magazin (January 2004)

Scientific American (January 2007)

Scientific American - January 2011

Psychoneuroendocrinology: The Scientific Basis of Clinical Practice

Evidence-Based Practice: Toward Optimizing Clinical Outcomes

Practice-based Evidence for Healthcare: Clinical mindlines

Evidence-Based Clinical Supervision: Principles and Practice

Handbook of Community-Based Clinical Practice

Case Based Echocardiography: Fundamentals and Clinical Practice

Evidence-Based Practice: Toward Optimizing Clinical Outcomes

Scientific american (June 2004)

Scientific american (October 2004)

Introduction to Vascular Biology: From Basic Science to Clinical Practice

Introduction to Vascular Biology: From Basic Science to Clinical Practice

Scientific american (May 2004)

Scientific american (December 2004)

Scientific american (November 2004)

Scientific american (September 2004)

Knowledge-based clustering: from data to information granules

Knowledge-Based Clustering: From Data to Information Granules

Scientific american (March 2004)

Recommend Documents