The Biomaterials Silver Jubilee Compendium
The Best Papers Published in
BIOMATERIALS
1980-2004
Elsevier Internet Homepage- http://www.elsevier.com Consult the Elsevier homepage for full catalogue information on all books, major reference works, journals, electronic products and services.
All Elsevier journals are available online via ScienceDirect: www.sciencedirect.com
To contact the Publisher Elsevier welcomes enquiries concerning publishing proposals" books, journal special issues, conference proceedings, etc. All formats and media can be considered. Should you have a publishing proposal you wish to discuss, please contact, without obligation, the publisher responsible for Elsevier's materials and engineering programme:
Jonathan Agbenyega Publisher Elsevier Ltd The Boulevard, Langford Lane Kidlington, Oxford OX5 1GB, UK
Phone: Fax: E-mail:
+44 1865 843000 +44 1865 843987
[email protected] General enquiries, including placing orders, should be directed to Elsevier's Regional Sales Offices - please access the Elsevier homepage or full contact details (homepage details at the top of this page).
The Biomaterials Silver Jubilee Compendium
The Best Papers Published in
BIOMATERIALS
1980-2004
Edited by D.F. Williams
2006
iiii!!i!!!iiiilDiii!!i!iiiiiiii!!!!!!!!7~, :~:i i~l ,~':!iii!!!!!!!iii!ii
Amsterdam Paris
. Boston
. San Diego
. Heidelberg . San Francisco
. London . Singapore
' New York ' Sydney
. Oxford . Tokyo
Elsevier Ltd is an imprint of Elsevier with offices at: Linacre House, Jordan Hill, Oxford OX2 8DP, UK The Boulevard, Langford Lane, Kidlington, Oxford OX5 1GB, UK 84 Theobald's Road, London WC1X 8RR, UK Radarweg 29, PO Box 211, 1000 AE Amsterdam, The Netherlands 30 Corporate Drive, Suite 400, Burlington, MA 01803, USA 525 B Street, Suite 1900, San Diego, CA 92101-4495, USA First edition 2006 Copyright 9 Elsevier Ltd. All rights reserved No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means electronic, mechanical, photocopying, recording or otherwise without the prior written permission of the publisher Permissions may be sought directly from Elsevier's Science & Technology Rights Department in Oxford, UK: phone (+44) (0) 1865 843830; fax (+44) (0) 1865 853333; email:
[email protected]. Alternatively you can submit your request online by visiting the Elsevier web site at http://elsevier.com/locate/permissions, and selecting Obtaining permission to use Elsevier material Notice No responsibility is assumed by the publisher for any injury and/or damage to persons or property as a matter of products liability, negligence or otherwise, or from any use or operation of any methods, products, instructions or ideas contained in the material herein. Because of rapid advances in the medical sciences, in particular, independent verification of diagnoses and drug dosages should be made
British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library Library of Congress Cataloging-in-Publication Data A catalog record for this book is available from the Library of Congress For information on all biomaterials related publications visit our web site at boo ks. e lsevi er. corn Printed and bound in Great Britain 06 07 08 09 10 l0 9 8 7 6 5 4 3 2 1 ISBN-13:978-0-08-045154-1 ISBN-10:0-08-045154-3
Working together to grow libraries in developing countries www-elsevier.c~ I www.bookaid.org t w~sw.sabre.org
The Biomaterials Silver Jubilee Compendium
Table of C o n t e n t s Title, Author(s) and Reference
Page No.
Preface D. F. Williams Controlled release of biologically active compounds from bioerodible polymers J. Heller Biomaterials 1980 Jan; volume 1 issue 1:pp51-57 The response to the intramuscular implantation of pure metals A. McNamara, D.F. Williams Biomaterials 1981 Jan; volume 2 issue 1:pp33-40 Osseointegrated titanium fixtures in the treatment of edentulousness 17 P.I. Branemark, R. Adell, T. Albrektsson, U. Lekholm, S. Lundkvist, B. Rockler Biomaterials 1983 Jan;volume 4 issue 1."pp25-28 B iomaterial biocompatibility and the macrophage J.M. Anderson, K.M. Miller Biomaterials 1984 Jan," volume 5 issue 1."pp5-10
21
Systemic effects of biomaterials J. Black Biomaterials 1984 Jan; volume 5 issue 1: ppl 1-18
27
The in vitro response of osteoblasts to bioactive glass T. Matsuda, J.E. Davies Biomaterials 1987 Jul; volume 8 issue 4:pp275-284
35
Activation of the complement system at the interface between blood and artificial surfaces M.D. Kazatchkine, M.P. Carreno Biomaterials 1988 Jan; volume 9 issue 1:pp30-35
45
Dynamic and equilibrium swelling behaviour of pH-sensitive hydrogels containing 2-hydroxyethyl methacrylate L. Brannon-Peppas, N.A. Peppas Biomaterials 1990 Nov; volume 11 issue 9: pp635-644.
51
Macroencapsulation of dopamine-secreting cells by coextrusion with an organic polymer solution P. Aebischer, L. Wahlberg, P.A. Tresco, S.R. Winn. Biomaterials 1991 Jan; volume 12 issue 1" pp50-56
61
Interaction between phospholipids and biocompatible polymers containing a phosphorylcholine moiety M. Kojima, K. Ishihara, A. Watanabe, N. Nakabayashi Biomaterials 1991 Mar; volume 12 issue 2:pp121-124
69
The Biomaterials Silver Jubilee Compendium
vi
Quantitative assessment of the tissue response to implanted biomaterials D.G.Vince, J.A. Hunt, D.F. Williams Biomaterials 1991 Oct; volume 12 issue 8" pp731-736
73
Immune response in biocompatibility A. Remes, D.F. Williams Biomaterials 1992;volume 13 issue 11" pp731-743
79
Laminated three-dimensional biodegradable foams for use in tissue engineering A.G. Mikos, G. Sarakinos, S.M. Leite, J.P. Vacanti, R. Langer Biomaterials 1993 Apr; volume 14 issue 5" pp323-330
93
Late degradation tissue response to poly(L-lactide) bone plates and screws J.E. Bergsma, W.C. de Bruijn, F.R. Rozema, R.R. Bos, G. Boering Biomaterials 1995 Jan; volume 16 issue 1" pp25-31
101
Mechanism of cell detachment from temperature-modulated, hydrophilichydrophobic polymer surfaces T. Okano, N. Yamada, M. Okuhara, H. Sakai, Y. Sakurai Biomaterials 1995 Mar; volume 16 issue 4" pp297-303
109
Mechanisms of polymer degradation and erosion A. Gopferich Biomaterials 1996 Jan; volume 17 issue 2" pp 103-114
117
Stabilized polyglycolic acid fibre-based tubes for tissue engineering 129 D.J. Mooney, C.L. Mazzoni, C. Breuer, K. McNamara, D. Hem, J.P. Vacanti, et al Biomaterials 1996 Jan; volume 17 issue 2" pp 115-124 Poly(alpha-hydroxy acids): carriers for bone morphogenetic proteins J.O. Hollinger, K. Leong Biomaterials 1996 Jan; volume 17 issue 2:pp187-194
139
Response ofMG63 osteoblast-like cells to titanium and titanium alloy is 147 dependent on surface roughness and composition J. Lincks, B.D. Boyan, C.R. Blanchard, C.H. Lohmann, Y. Liu, D.L. Cochran, et al. Biomaterials 1998 Dec; volume 19 issue 23" pp2219-2232 Patterning proteins and cells using soft lithography R.S. Kane, S. Takayama, E. Ostuni, D.E. Ingber, G.M. Whitesides. Biomaterials 1999 Dec;volume 20 issue 23-24" pp2363-2376
161
Scaffolds in tissue engineering bone and cartilage D.W. Hutmacher Biomaterials 2000 Dec; volume 21 issue 24:pp2529-2543
175
The Biomaterials Silver Jubilee Compendium
vii
Topographical control of human neutrophil motility on micropatterned materials 191 with various surface chemistry. J. Tan, W.M. Saltzman. Biomaterials 2002 Aug; volume 23 issue 15" pp3215-3225. Photopolymerized hyaluronic acid-based hydrogels and interpenetrating networks Y.D. Park, N. Tirelli, J.A. Hubbell Biomaterials 2003 Mar; volume 24 issue 6" pp893-900
203
Cell sheet engineering for myocardial tissue reconstruction T. Shimizu, M. Yamato, A. Kikuchi, T. Okano Biomaterials 2003 Jun; volume 24 issue 13" pp2309-2316
211
Biomaterial-associated thrombosis: roles of coagulation factors, complement, platelets and leukocytes M.B. Gorbet, M.V. Sefton Biomaterials 2004 Nov; volume 25 issue 26" pp5681-5703
219
Author Index
243
The Biomaterials Silver Jubilee Compendium
viii
This Page Intentionally Left Blank
/x
The Biomaterials Silver Jubilee Compendium
Preface The journal Biomaterials was launched in 1980. The subject of biomaterials science was then in its infancy, being largely confined to the study of the characteristics of materials used for medical devices. In reality, few of those materials had ever been developed for this specific use and instead, were taken from other industrial applications, for example in aerospace, nuclear engineering or chemical processing, and experimented with in surgical or medical procedures. The science was largely observational, as the performance of these materials in their new surroundings was evaluated by combinations of physical, chemical, engineering, biological, pathological and clinical techniques. Over the ensuing decade, the subject evolved, as more became known about their performance and especially about the mechanisms of the interactions between the materials and the tissues that underpin the performance. This branch of biomaterials science has become associated with the term biocompatibility, a field that has been the driving force for the subject. With greater knowledge about these interactions, old serendipitous biomaterials were discarded, and new, intentionally designed, or at least modified, materials, introduced. Moreover, these materials started to find applications in related areas, and medical devices were no longer the sole home for biomaterials, as applications in pharmaceutical technology through drug and gene delivery, regenerative medicine and tissue engineering, and biotechnology have emerged and developed. Twenty-five years on, we can truly say that biomaterials science has matured at an incredible rate and now represents a formidable sector that bridges the materials sciences, advanced medical therapies, and molecular and cell sciences. This development could not have been achieved without high quality scientific journals, including those that represent the main parent disciplines and the interdisciplinary field of biomaterials science itself. Although by no means alone, the journal Biomaterials has taken centre stage here and, at the time of its silver jubilee in 2004 is widely considered to be the premier journal in this field. In order to celebrate 25 years of publishing biomaterials science, Elsevier decided to confer two awards, the Elsevier Biomaterials Gold Medal and the Elsevier Biomaterials Silver Medal. A panel of judges was established in 2005 in order to select the recipients. The Elsevier Biomaterials Gold Medal was awarded to the person judged to have made the most significant contribution to the subject of biomaterials science during
The Biomaterials Silver Jubilee Compendium
the 25 years from 1980 to 2004, irrespective of where the work was published. Elsevier were delighted to announce at the European Society for Biomaterials Annual Meeting in Sorrento, Italy, in September 2005, that the winner of the Gold Medal was Professor James Anderson, of Case Western Reserve University, Cleveland, Ohio, USA. The medal was presented at the Annual Meeting of the Tissue Engineering Society International in Shanghai, in October 2005. The Elsevier Biomaterials Silver Medal was awarded for the most significant paper published in the journal Biomaterials during the first 25 years. Over 60 papers were nominated and the judges had a very difficult time in making the selection since most of the world's leading biomaterials scientists were represented in the nomination list. The subject matter covered much of the seminal research that has set the foundation for the high quality science that is undertaken today and which will embrace the future. This Silver Jubilee Compendium consists of reprinted versions of the top 25 of these papers, arranged chronologically. The current Editor-in-Chief is both appreciative of and humbled by the decision of the panel of judges to select one of his earliest papers, with graduate student Anne McNamara, as the leading paper and recipient of the Silver Award. This Compendium is published as a landmark in biomaterials science and it is to be hoped that it will serve as a stimulus to young biomaterials scientists of the early twenty-first century for their pioneering work of the future. I have been proud to serve as Editor-in-Chief of the journal during this exciting period of its development. I place on record my thanks to previous editors of the journal, listed on a separate page, to editorial staff within Elsevier and their predecessor publishers during this 25 years and to colleagues who have served as Editorial Board members, referees and authors. I would particularly wish to thank Amanda Weaver, Publisher of the journal in Elsevier who skilfully steered this process of the medals through the company, to the panel of judges who had to work very hard on this process, and to Peggy O'Donnell, Managing Editor of the journal, who carried the full logistics burden of the medals procedure.
Professor David Williams Editor-in-Chief Liverpool
History of Editorial Appointments Stephen Bruck
Founding Editor
(1980-1983)
Garth Hastings
Founding Editor
(1980-1995)
Nicholas Peppas
Editor-in-Chief
(1983-2001)
Robert Langer
Editor-in-Chief
(1983- 2003)
David Williams
Editor-in-Chief
(1996-)
The Biomaterials Silver Jubilee Compendium
Controlled release of biologically active compounds from bioerodible polymers J. Heller
Polymer Sciences Department, SRI International, Menlo Park, CA94025, USA Received 8 June 1979; revised 1 October 1979
This article reviewsthe controlled releaseof biologically active agents by the erosion or chemical degradation of a polymer matrix into which the agent is incorporated. Chemically bound active agentsand work on steroid releasefrom cholesterol implants are not covered. The mechanismsof polymer erosion discussedare: cross4inkedscission;hydrolysis, ionization or protonation of pendant groupts; backbone cleavage. Drug releasestudiesare dealt with under each of these headings.
It is now generally recognized that the controlled release of biologically active agents to a local environment can be achieved by means of one of three general methodologies: (1) diffusion through a rate-controlling membrane, (2) use of osmotically regulated flow, and (3) release controlled by the erosion or chemical degradation of a matrix into which the active agent is incorporated 1. Each of these methodologies offers certain unique characteristics which determine the design of specific therapeutic systems. Thus, methodology (1) allows construction of drug delivery devices that release therapeutic agents by zero order kinetics and where rate of delivery can be readily adjusted by changing the rate-limiting membrane and/or memb[ane thickness and area. Methodology (2) allows construction of devices that not only release their contents by zero order kinetics but are also able to sustain high delivery rates not normally available with membranemoderated devices. Methodology (3) allows construction of drug delivery devices that have a predetermined life span and need not be removed from the site of action once their drug delivery role has been completed. Drug release from bioerodible polymers finds use in both topical applications and systemic applications. Both uses demand that the polymer degrade to nontoxic products, and polymers used in systemic applications must also degrade to low-molecular-weight fragments that can be readily eliminated or metabolized by the body. In topical applications, retainment of high molecular weight of the degradation products is desirable, since in this way no unnecessary systemic absorption of the polymer will occur, and toxicological hazards are thus reduced. The purpose of this article is to present a comprehensive review of methodology (3) where active agents are released to a surrounding aqueous environment by solubitization of the polymer matrix induced by the aqueous environment. The review is limited to devices in which the active agent is dissolved or dispersed in a polymer, and does not
cover the important work in which the active agent is chemically bound to the polymer and is released to the surrounding medium by hydrolysis of a bond between the active agent and the polymer chain2,3; nor does it cover the extensive work of Kincl and coworkers on the release of steroids from cholesterol implants 4. For this review, it is convenient to systematize polymer erosion according to the three mechanisms shown in Figure I, where ~) denotes a hydrolytically unstable bond 5. In general terms, Mechanism ] encompasses watersoluble polymers that have been insolubilized by hydrolytically unstable crosslinks; Mechanism ]! includes polymers that are initially water-insoluble and are solubilized by Mechonism T
l~,~h~ism
---'-
c
Mechanism ]]T
Figure 1 Schematic representation of degradation mechanisms; ~) denotes a hydrolyticaHy unstable bond A represents a hydrophobic substituen t and B - . C represents hydrolysis, ionization or protonation
0142-9612/80/010051-07 $02.00 9 1980 IPC BusinessPress Biomaterials 1980, Vol 1 January
51
The Biomaterials Silver Jubilee Compendium Bioerodible polymers: J. Heller
hydrolysis, ionization, or protonation of a pendant group; and Mechanism [[[ includes hydrophobic polymers that are converted to small water-soluble molecules by backbone cleavage. Clearly, these three mechanisms represent extreme cases, and erosion by a combination of mechanisms is possible.
recognized the utility of erodible hydrogels for providing sustained delivery of entrapped macromolecules. Both studies took advantage of the hydrolytic instability of crosslinks formed in vinyl polymers by using N, N~-methyl enebisacrylamide as a comonomer. Hydrolysis of the crosslinks proceeds as follows:
, 0
MECHANISM I
i-
, O~ H_I~_H i 2
= 2-R-_-NH +
Solubilization by crosslink cleavage In these systems, water-soluble polymers are insolubilized by means of hydrolytically unstable crosslinks. Consequently, the resulting matrix is highly hydrophilic and completely permeated by water. Since the active agent is in an aqueous environment, its water solubility becomes an important consideration, and compounds with appreciable water solubility will be rapidly leached out, independent of the matrix erosion rate. There are two general applications in which erodible hydrogets are useful in the controlled delivery of active agents. In the first the active agent has extremely low water solubility, and in the second the active agent is a macromolecule that is entangled in the hydrogel matrix and cannot escape until a sufficient number of crosslinks have cleaved and matrix crosslink density has been reduced. The first application is illustrated in Figure 2, which shows the release of a highly water-insoluble drug, hydrocortisone acetate, from a gelatin matrix crosslinked with formaldehyde 5. As indicated by the first-order dependence, release is by diffusion with little contribution by matrix erosion. Because the drug is very water insoluble, useful release over many days is achieved. Such a device could be used when zero-order kinetics are not important and removal of the expended device is not convenient or desirable 6. Illustrative of the second application are many examples in which water-soluble macromolecules have been immobilized in hydrogets by physical entanglement 7. However, the intent of most of these studies was to achieve long-term immobilization of enzymes or antigens; the slow diffusional escape and/or slow hydrolysis of the matrix with consequent liberation of the entrapped macromolecules was generally regarded as undesirable. Two studies, however,
40-] 3O E 2O-
I0
1
I
72 96 Time(days)
.........
J
120
t44
Figure 2 Releaseo f hydrocortisone acetate from a cross~inked gelatin matrix
52
Biornateriais 1980, Vol 1 January
where - R - represents the vinyl polymer chain. In the first study 8, bovine pancreatic insulin was immobilized in a hydrogel prepared from acrylamide and 2% N, N'-methylenebisacrylamide. Slow release of insulin from the hydrogel was inferred because insulin-containing hydrogel implants sustained diabetic animals for at least a few weeks. It is not clear from the study how much insulin was released by diffusion and how much by cleavage of crosslinks. In the second study 9, (~-chymotrypsin was immobilized in a hyclrogel prepared from N-vinyl pyrrolidone and N, N'-methylenebisacrylamide. It was found that, by varying the N, N'-methylenebisacrylamide concentration from 0.1 to 1.0w/w % with respect to N-vinylpyrrolidone, hydrogels with dissolution times of several days to practically insoluble could be prepared. However, release of =-chymotrypsin did not correlate well with hydrogel dissolution time, presumably because of diffusional escape. To prevent diffusional escape, e-chymotrypsin was acylated with acryloyl chloride and then chemically incorporated in the hydrogel by copolymerization with N-vinyl pyrrolidone and N, N'-methylenebisacrylamide. Release of e-chymotrypsin from the resulting hydrogels was then found to correlate more closely with hydrogel dissolution times.
MECHANISM II Solubilization by hydrolysis, ionization or protonation of pendant groups
Systems in this category include all polymers that are initially water-insoluble but become water-soluble as a consequence of hydrolysis, ionization, or protonation of pendant groups. Because no backbone cleavage takes place, the solubilization does not result in any significant changes in molecular weight. The major emphasis in the development of these materials has been on enteric coatings. These are coatings designed to be insoluble in a certain pH environment, usually the stomach, and then to dissolve abruptly in an environment of a different pH, such as the intestines. Usually these polymers are applied to pills as protective coatings and do not produce steady, sustained release of therapeutic agents. However, by using mixtures of enteric coatings, each with a different disintegration time, it is possible to prolong the action of therapeutic agents10. Literature on enteric coatings is much too voluminous for detailed review, but the coatings can be grouped into three categories according to their dissolution mechanism; (1) dissolution by side group hydrolysis, (2) dissolution by ionization of a carboxylic acid function and (3) dissolution by protonation of amine functions.
The Biomaterials Silver Jubilee Compendium Bioerodible polymers: J. Heller Dissolution b y side group hydrolysis. Materials in this
category are represented by copolymers of vinyl monomers and maleic anhydride'
XI HO -C H^C CHH -CHI~ I ~ 2~/~', 0
0
Xl -CH 2-CH - I H ~ CIHcoo~ cor
0
He
He
where X is OR or H. In the anhydride form, the polymers are waterinsoluble, but on exposure to water they become soluble because of the hydrolysis of the anhydride group. A number of variables affect the rate of polymer dissolution and lag time before initiation of the dissolution process11. In general, time before dissolution increases as the size of the alkyt substituent R in the vinyl ether portion of the copolymer increases, and decreases as the pH of the aqueous environment increases. The rate of polymer dissolution also increases as the pH of the aqueous environment increases. Dissolution by i o n i z a t i o n o f c a r b o x y l groups. Currently used enteric coatings represent this type, and can be represented generally as polyacids. While in unionized form they are water-insoluble, but on ionization of the carboxylic acid functions they become water-soluble. The most widely used enteric coatings are based on cellulose acetate phthalate 12. They are insoluble in aqueous acidic media but, because of the free acid groups on the phthalate radical, dissolve in aqueous bases. Enteric coatings based on cellulose acetate succinate have also been described 13. Partially esterified copolymers of methyl vinyl ether and maleic anhydride or partially esterified copolymers of ethylene and maleic anhydride have also been investigated 14-16. It was found that these materials characteristically exhibit a pH range above which they are soluble and below which they are insoluble. This pH range is quite sharp, about 0.25 pH units, and changes with the number of carbon atoms in the ester side group of the copolymers. This behaviour can be understood readily by considering the number of ionized carboxyls necessary to drag the polymer chain into solution. With relatively small ester groups, a low degree of ionization is sufficient to solubilize the polymer; hence the dissolution pH is low. As the size of the alkyl group increases, so does the hydrophobicity, and IO0
. . . . . . . . . . . . . . . .
progressively more ionization is necessary to solubilize the polymer, resulting in increasingly high dissolution pH. The same argument holds for polymers having the same ester grouping but different degrees of esterification. The higher the degree of esterification, the more hydrophobic the polymer and consequently the higher the dissolution pH. Recently it has been shown 17 that partially esterified copolymers of methyl vinyl ether and maleic anhydride can, in a constant pH environment, release hydrocortisone incorporated therein by excellent zero-order kinetics. Figure 3 shows polymer dissolution rate and the rate of hydrocortisone release for n-butyl half-ester polymer films containing the dispersed drug. Each pair of points represents a separate device in which the amount of drug released by the device into the wash solution was determined by u.v. measurements and the amount of polymer dissolved was calculated from the total weight loss of the device. The excellent linearity of both polymer erosion and drug release over the lifetime of the device provides strong evidence for a surface-erosion mechanism and for negligible diffusional release of the drug. The latter result was independently verified by placing a drug-containing film in water at a pH low enough that no dissolution of the matrix took place and periodically analysing the aqueous solution for hydrocortisone. None was found over several days. Figure 4 shows the effect of size of alkyl group on polymer erosion rate for a series of partial esters measured at pH 7.4. Because of the linear correlation between polymer erosion and drug release, distance eroded can be directly correlated with amount of drug released and was, in fact, derived from measuring drug release. All drug release rates again show excellent linearity and strong dependence on the size of the alkyl group. Since in all experiments drug depletion also coincided with total polymer dissolution, again it can be assumed that drug release and polymer erosion occur concomitantly. The effect of pH on rate of polymer erosion and hence release of hydrocortisone dispersed in the n-butyl partial ester is shown in Figure 5. The date show a clear dependence of erosion rate and drug release on the pH of the eroding medium and, as expected, a progressive decrease in rate as the critical dissolution pH is approached. The partial ester copolymer also has been used recently as a model for a bioerodible drug delivery system that releases a therapeutic agent in response to the presence of a specific external molecule 18. In this model, hydrocortisone was incorporated into an n-hexyt half ester of a methyl vinyl ether-maleic anhydride copolymer and the 3ooI
100
.
.
.
.
.
.
.
.
.
.
.
.
.
.
"~.&250I 60
-6o
cJ_ ~4,,~ 2 o 20 0 O
o
~0 E a. 20
~ tO
20
30 40 Time (h)
50
g
60
Figure 3 Rate of polymer dissolution and rate of release of hydrocortisone for the n-butyl half-ester or methyl vinyl ether-maleic anhydride copolymer containing 10 wt-% drug dispersion. (0), drug release; (A), polymer dissolution. Reproduced from J. Appl. Polym. Sci. 1978, 22, 1991 by permission of John Wiley and Sons Inc., New York
0
I0
!
20
J
:50
1
I
40 50 Time (ram)
'
60
70
80
90
Figure 4 Effect of size ot ester group in half-esters of methyl vinyl ether-maleic anhydride copotymers on rate of erosion at pH 7.4. Reproduced from J. AppL Polym. ScL 1978, 22, 1991 by permission of John Wiley and Sons Inc., New York
Biomateriais 1980, Vol 1 January
53
The Biomaterials Silver Jubilee Compendium Bioerodible polymers:
d. Heller
20c
300
c
17
25(-
15 A
2(30=L
10 G)
~5
150-
IOO-
25 i
%
~6
20 .....30
~ ..... 50
Time (rain)
so
~o-~
9o
~oo
Figure 5 Effect of erosion medium pH on erosion rate of half-esters of methyl vinyl ether~rnaleic anhydride copolymers. A = pH 5.5, B = pH 5. 75, C = pH 6.0, D = pH 6.5, E = pH 7. Reproduced from J. AppI. Polym. Sci. 1978, 22, 1991 with permission of the copyright owner
polymer
10
100
1000
Pore diameter (~m] Figure 4 Cumulative pore size distribution of PLLA devices of one, two and three layers, measured by mercury porosimetry. 13, One layer; O, two layers; ,4, three layers.
multilayered devices, it was evident from the unhindered transport of fluids and cell suspensions across the interface of adjacent layers that the communication from layer to layer was not obstructed by the lamination process and that the interconnected pore structure was preserved. The effect of the lamination process on the creep behaviour of the polymer devices was evaluated by thermomechanical analysis. Figure 7a shows the strain measured for each device after 60 min of loading with 9.5 kPa of compressive stress. The strain measured for each of the PLLA, half-thickness PLLA (PLLA/2) and PLGA 85/15 devices was about 0.1, while the strain measured for the PLGA 50/50 devices was in the range 0.4-0.5, The number of membranes which made up the device had no effect on the measured strain. After 60 min, the stress was removed and the devices were allowed to recover unloaded for 30 min. Figure 7b shows the strain after the recovery period. For each material, nearly 50% of the deformation was recovered. In addition, the rates of strain change during a compressive creep cycle were identical for devices with different numbers of layers (Figure 8). These results are encouraging, in that once
again there was no observed effect of the lamination process on the properties of the membranes. The lamination process did not cause a weakening of the polymer foam in response to compressive forces. Laminated devices with anatomical shapes were constructed for potential use in reconstructive or orthopaedic surgery. The implant with a nose-like shape (Figure 1) was created by lamination of six PLLA membranes of an average thickness of 1727 pm and total porosity of 88%. A photograph of the implant is shown in Figure 9. We also processed laminated foams from similar PLLA membranes with shapes of metacarpalphalangeal pieces for joint repair (Figure 10). The head of each foam was prepared by lamination of four layers (discs) with orientation perpendicular to the axis of symmetry of the hemisphere and the stem was created separately by lamination of two strips. The two pieces were joined together to form a foam with the desired pinlike shape. Transplantation devices were also prepared for a variety of cells. Examples include devices for hepatocyte transplantation 12shown in Figure 11. Here, the lamination procedure was necessary to produce thick devices to accommodate a large number of hepatocytes for functional replacement. The devices were made of three layers with a catheter inserted in the centre of each device as a route for injection of hepatocytes into the bulk of the polymer. No delamination or failure of any devices due to the development of shear stresses from surrounding tissues was detected 12 from histological sections of a large number of harvested devices implanted in the mesentery of rats for a period of 35 d. The distribution of cells seeded into these devices via injection was recently modelled by our group 4 to maximize the device volume effectively employed in cell transplantation and determine the optimal surgical injection conditions. In conclusion, a new method was developed to laminate highly porous membranes and produce threedimensional foams with continuous pore structure and morphology. This method can be used to process biodegradable polymers into custom-made shaped devices with potential use in cell transplantation. With the aid of computer-assisted modelling to contour-map tissues and organs 13, we can easily construct templates with the desired implant shape. With further study of: (1] the procedures to uniformly seed large devices with cells, (2) the cell and tissue culture techniques to eliminate any mass transfer limitations of nutrients to the whole cell Biomaterials 1993, Vol. 14 No. 5
98
The Biomaterials Silver Jubilee Compendium 328
Laminated three-dimensional biodegradable foams: A.G. Mikos et al.
Figure 5 SEM photomicrographs of cross-sections of: a, three-layer laminated PLLA foam; and b, one of its constituent layers before lamination.
Figure 6 SEM photomicrographs of cross-sections of: a, three-layer laminated PLGA 85/15 foam' and b, one of its constituent layers before lamination. Biomaterials 1993, Vol. 14 No, 5
The Biomaterials Silver Jubilee Compendium
99
Laminated three-dimensional biodegradable foams: A.G. Mikos et al.
329
Figure8 Creep behaviour of PLLA (open symbols) and PLGA 50/50 (filled symbols) devices of N, !:3, one layer; O, O, two layers; and &, A, three layers as measured by thermomechanical analysis at 37~ At time zero, a constant compressive stress of 9.5 kPa was applied for 60 min. Afterwards, the stress was removed and the sample was allowed to recover for an additional time of 30 min. The strain is defined as the thickness change divided by the initial thickness.
Rgure9 Photograph of a laminated PLLA foam with noselike shape.
Figure 7 Compressive strain of devices of one, two and three layers for the four types of foams shown: a, after 60 rain under a stress of 9.5 kPa and b, after 30 min of recovery from the previous stress, measured by thermomechanical analysis at 37~ B, One layer; t~, two layers; I~, three layers.
mass and ensure the viability, growth and function of the attached cells, and (3} the surgical approaches to implant polymer-cell devices to the sites of the functioning tissues, this method could become important in the development of novel cell-based artificial organs for clinical use.
Figure lO Photograph of metacarpal-phalangeal pieces made of laminated PLLA foams (A, B) similar to a nondegradable medical implant (C).
Biomaterials 1993, Vol. 14 No. 5
100
The Biomaterials Silver Jubilee Compendium Laminated three-dimensional biodegradable foams: A.G. Mikos et al.
330
4
Wald, H,L., Sarakinos, G., Lyman, M.D., Mikos, A,G., Vacanti, ].P. and Langer, R., Cell seeding in porous transplantation devices, Bioraaterials [in press} Cima, L,G., Ingber, D.E., Vacanti, ].P. and Langer, R., Hepatocyte culture on biodegradable polymeric substrates, BiotechnoL Bioeng. 1991, 38, 145-158 Vacanti, C.A., Langer, R., Schloo, B. and Vacanti, ].P., Synthetic polymers seeded with chondrocytes provide a template for new cartilage formation, Plast. Reconstr. Surg, 1991, 88, 753-759 Cima,L.G., Vacanti, ].P., Vacanti, C., Ingber, D., Mooney, D. and Langer, R., Tissue engineering by cell transplantation using degradable polymer substrates, ]. Biomech. Eng. 1991, 113, 143-151 Freed, L.E., Marquis, I.C., Nohria, A., Emmanual, 1., Mikos, A.G. and Langer, R., Neocartilage formation in vitro and in vivo using ceils cu(tured on synthetic biodegradable polymers, ]. Biomed. Mater. Res. 1993, 27, 11-23 Mikos, A,G., Thorsen, A.]., Czerwonka, L.A., Bao, Y., Winslow, D.N., Vacanti, ].P. and Langer, R., Preparation and characterization of poly[L-lactic acid) foams for cell transplantation, Polymer {submitted} Winslow, D.N., Advances in experimental techniques for mercury intrusion porosimetry, in Surface and Colloid Science {Eds E. Matt}eric and R.]. Good}, Plenum Press, NY, USA, 1984, pp 259-282 Tsakiroglou, C.D. and Payatakes, A.C., A new simulator of mercury porosimetry for the characterization of porous materials, ]. Colloid Interface Sci. 1990, 137, 315-339 Mikos, A.G., Sarakinos, G., Lyman, M.D., Ingber, D.E., Vacanti, ].P. and Langer, R., Prevascularization of porous biodegradable polymers, Biotechnol. Bioeng. [in press] ]imenez, ]., Santisteban, A., Carazo, ].M. and Carrascosa, I.L., Computer graphic display method for visualizing three-dimensional biological structures, Science 1986, 232, 1113-1115
5 6
7
Figure 11 Photograph of hepatocyte transplantation devices made by lamination of three layers of PLLA (A), PLLAI2 (B), PLGA 85/15 (C) and PLGA 50/50 (D), with a catheter positioned in the middle of each device.
8
9
ACKNOWLEDGEMENTS Many thanks to Ms Michelle D. Lyman for excellent technical assistance. This work was supported by a grant from Advanced Tissue Sciences.
10
REFERENCES
i1
1
2 3
Vacanti, ].P., Morse, M.A., Saltzman, W.M., Domb, A.|,, Perez-Atayde, A. and Langer, R,, Selective cell transplantation using bioabsorbable artificial polymers as matrices, ]. Pediatr. Surg. 1988, 23, 3-9 Vacanti, ].P., Beyond transplantation, Arch. Surg. 1988, 123, 545-549 Mikos, A.G., Bao, Y., Cima, L.G,, Ingber, D.E., Vacanti, I.P. and Langer, R., Preparation of poly{glycolic acid] bonded fiber structures for cell attachment and transplantation, ]. Biomed. Mater. Res. 1993, 27, 183-189
.....
~
12
13
9
UJII
-
__
.
.
.
.
.
. . . .
I
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
- - ~ L
-
-
Biomedical Materials & Technologies
Biocompatibility of Medical Devices
Centro de Citologia Experimental, Porto, Portugal 28-30 June 1993 This course aims to give a comprehensive survey of current knowledge in the fields of biomaterials and biocompatibility, and will appeal to a multidisciplinary audience. The structure and properties of biornaterials will be reviewed, and particular attention paid to their uses in orthopaedics, dentistry and cardiovascular surgery. The biocompatibility of different materials will be fully examined. The interactions of host cells with biomaterials will be described and discussed, with emphasis being placed on the importance of biocompatibility to implant survival.
For further information and registration details please contact: COMETT Course Secretary, Department of Clinical Engineering, University of Liverpool, PO Box 147, Liverpool L69 3BX, UK. Fax" + 44 051 706 5803 _ .
_
-
9
, ......
,
,,.
J
L,J
- -
Biomaterials 1993, Vol. 14 No. 5
,,,
.
.
.
.
.
.
,, ,,~,,,__..,
101
The Biomaterials Silver Jubilee Compendium
.B U . T . T. E.R .W. O. R T H E I N E M A N
~
Biomaterials16 (1995)25-31 9 1995 ElsevierScience Limited Printed in Great Britain. All rights reserved 0142-9612/95/$10.00
N
Late degradation tissue response to poly(L-lactide) bone plates and screws J.E. Bergsma, W.C. de Bruijn*, F.R. Rozema, R.R.M. Bos and G. Boering Department of Oral and Maxillofacial Surgery, University Hospital Groningen, PO Box 30.001, 9700 RB Groningen, The Netherlands; *AEM Unit, Clinical Pathological Institute I, Erasmus University of Rotterdam, The Netherlands
Patients with fractures of the zygomatic bone were treated with high molecular weight poly(L-lactic) acid (PLLA) bone plates and screws. Three years after implantation four patients returned to our department with a swelling at the site of implantation. At the recall of the remaining patients we found an identical type of swelling after the same implantation period. To investigate the nature of the tissue reaction, eight patients were reoperated for the removal of the swelling. The implantation period of the PLLA material varied from 3.3 to 5.7 years. Microscopic evaluation and molecular weight measurements were performed. The excised material showed remnants of degraded PLLA material surrounded by a dense fibrous capsule. Ultrastructural investigation showed crystal-like PLLA material internalized by various cells. The results of this investigation suggest that the PLLA material slowly degrades into particles with a high crystallinity. The intra- and extracellular degradation rate of these particles is very low. After 5.7 years of implantation, these particles were still not fully resorbed. Biomaterials (1995), 16 (1), 25--31 Keywords: Poly (L-lactide), biodegradation, tissue response, enzyme activity
Received 28 November 1993; accepted 25 April 1994
In a study on rats the biocompatibility and degradation characteristics of PLLA were investigated TM. The histological reaction to the implanted PLLA material was very mild: only a slight foreign body reaction was observed after a follow-up of 2.8 years. The implanted PLLA material showed a rapid decrease of molecular weight but only a small mass loss. Total resorption of the PLLA material was not observed in this study on rats, but was estimated to be about 3.5 years. Based on the positive results in animal studies, a limited investigation in humans was set up. PLLA bone plates and screws were used for the fixation of unstable zygomatic fractures 17. Three years after implantation, four patients returned to our department spontaneously because of a swelling at the site of implantation TM. Another five patients were recalled and all showed identical swellings. The aim of this study is to characterize the remainder of the PLLA material after an implantation period of 3.3 and 5.7 years, in order to gain an insight to the nature and course of the swelling at a light microscopical, ultrastructural and cytochemical level.
Metallic bone plates and screws are commonly used in oral and maxillofacial surgery for internal fracture fixation. Although good fracture healing is obtained, the disadvantages of metallic plates and screws are the possibility of bone atrophy due to stress-shielding and the obligation to remove these devices in a second operation 1-3. Bone plates and screws made of a biodegradable material are considered to be a good alternative for metallic ones. The main advantage of biodegradable plates and screws is that they lose their mechanical properties due to degradation so that loads are gradually retransferred to the bone, preventing stress-shielding of the healed bone. Moreover, if the material fully degrades, a reoperation for the removal of the plate and screws can be avoided 4-6. Because of these advantages, biodegradable polyesters such as poly(L-lactide) or polyglycolide have been studied extensively during the last two decades. These biodegradable polyesters have been used as internal fixation devices in the shape of rods, screws and bone plates 7-9. At our department high molecular weight aspolymerized poly(L-lactide) (PLLA) has been a material of special interest because of its gaod mechanical properties ~~ To gain insight into the mechanical behaviour during degradation, PLLA was used for fracture fixation of the mandible of dogs 13 and sheep 14, and for orbital floor reconstructions in goats 15. In all cases the PLLA plates gave sufficient stability to enable undisturbed fracture healing 13-15.
MATERIALS AND METHODS
Patients From 1986 to 1988, ten patients (mean age 39.6 yr; range 20.2-61.8 yr) with solitary displaced unstable fractures of the zygoma were treated with high molecu-
Correspondence to Dr J.E. Bergsma. 25
Biomaterials 1995, Vol. 16 No. 1
102
The Biomaterials Silver Jubilee Compendium Late degradation tissue response to poly(L-lactide) bone plates and screws' J.E. B e r g s m a
26
lar weight as-polymerized poly0.-lactide) bone plates and screws (My = 1 x 106, calculated using the formula [~1] = 5.45 • 104 l~//~73; and Mn = 7.6 x 105, calculated using the formula [77]= 3.25 x 104 ~'~n0"77).Unfortunately one patient died of a cerebrovascular accident one year after implantation. Three years after implantation, four out of nine patients returned to our department at their own initiative because of a painless swelling at the site of implantation TM. At a recall, the remaining five patients showed a similar swelling. In a period of two years, seven patients agreed to have a reoperation for the exploration of the area of swelling. The postoperative implantation period varied from 3 years 4 months to 5 years 8 months. The reoperation was performed under general anaesthesia. The tissue in the area of the swelling was excised via an incision laterally in the eyebrow. Samples of the screw-holes were taken by trephination of the bone.
Characterization of the degraded PLLA From a part of the excised tissue the remainder of the PLLA material was mechanically removed and was subsequently treated with trypsin 2.5% in Hank's balanced salt solution and collagenase ]a for the removal of organic components. The PLLA particles were then dried under vacuum for 17 h at 1 0 -3 bar to constant weight. Nuclear magnetic resonance (NMR) measurements for the determination of the molecular weight were carried out on a Varian 300 NMR spectrometer. The 1H NMR spectra were obtained from polymer solutions in deuterated chloroform in 5 m m tubes. For scanning electron microscopy (SEM) analysis, dried PLLA particles were gold-palladium sputter-coated and photographed in a DS 130 (ISI) scanning electron microscope.
Histological procedures Slices 2 m m thick were cut perpendicular to the long axis of the excised tissue mass and fixed in 2% v/v glutaraldehyde in 0.1 M phosphate buffer of pH 7.4 for at least 48 h at 4~ For light microscopy, sections were dehydrated in graded series of ethanol. The sections were embedded in glycol-methacrylate, polymerized for 24 h at -20~ Sections of 2 pm were made (]ung microtome 1140/autocut), which were stained with toluidine blue and basic fuchsin. For electron microscopy, postfixation was performed with 1 wt% OsO4 to which K4Fe(CN)~. 3H20 was added to a final concentration of 0.05 M. Subsequently, the material Table 1
et al.
was dehydrated in series of 70, 80, 90 and 100% acetone. The material was embedded in LX 112 epoxy resin and polymerized for 24 h at 60~ Based on light microscopic observations, ultrathin 70 nm sections were acquired at selected sites, around the screw-head and at the periphery of the bone plate. These ultrathin sections were stained with uranyI acetate/lead citrate. For transmission electron microscopy a Zeiss EM 902 was used, operating at 80 kV.
Histochemical procedures To investigate possible enzymatic activity towards the membrane-bound PLLA conglomerates, cytochemical reactions were performed on the material implanted for 5.7 years. For the demonstration of acid phosphatase and alkaline phosphatase, aldehyde-fixed tissue was used. For the demonstration of lactate dehydrogenase (LDH), fresh 5 m m tissue slices were quickly frozen at -80~ Perpendicular to the bone plate axis, sections were cut of 50 gm thickness in a cryostate microtome at -28~ The histochemical methods for the enzymes investigated are summarized in Table 1.
RESULTS Material characterization The mean number molecular weight (~/,) of the PLLA particles as determined by NMR was respectively 5600 and 5400 for the 3.3 and 5.7 years implanted PLLA. The plates and screws were machined out of a block of as-polymerized with an N/n of 7.6 X 105. Ultrastructural SEM examination of the 3.3 years material revealed particles varying in size from 1 to 1500 pm (Figure 1). The larger PLLA fragments showed numerous whitish cracks. Many smaller particles seemed to be adhered to the surface. Higher magnification of a particle showed an irregular surface structure. The SEM appearance of the material implanted for 5.7 years showed at some parts a microporous structure, again with many smaller particles attached to it. The mean particle size seemed to be smaller compared with those of the 3.3 years implanted material.
Histological analysis The material excised after an implantation period of 3.3 years showed a firm consistency and the contours of some of the screw-heads could still be seen. Light
Histochemical methods for the enzymes investigated. .
.
.
.
.
.
.
.
Enzyme
Substrate, cofactors, coupling agent
Buffer and pH
Incubation (time, temp)
Reference
Acid phosphatase
1.5 mg ml-1 Sodium-~-glycerophosphate 1 mM Cerium chloride
0.08 M Tris-mateate pH 5.0
30 min, 37~'C
Hulstaert et al. 2~
Alkaline
1.5 mg m1-1 Sodium-,8-glycerophosphate 1 mM Cerium chloride 4 mM Magnesium chloride
0.1 M Tris-maleate pH 9.2
30 min, 37:C
Hulstaert et al. 2~
Lactate dehydrogenase
2.5 mg m1-1 Lactic acid lithium salt 0.05 M potassium ferricyanide 0.5 mg m -~ NAD +
0.1 M Phosphate buffer pH 7.2
60 min, 37~C
Hanker et al. 21
phosphatase
(LDH) .
..
9
B i o m a t e r i a l s 1995, Vol. 16 No. I
,
.
,,
103
The Biomaterials Silver Jubilee Compendium Late degradation tissue response to poly(L-lactide) bone plates and screws: J.E. Bergsma et a/.
27
Figure 1 Scanning electron micrograph of the remnants of the 3.3 year implanted poly(L-lactic) acid material (scale bar = 500/zm, 40 kV).
microscopic analysis of this material revealed a fibrous tissue capsule that enveloped large areas of foreign body material. Under crossed Nicol prisms the areas with the foreign body material were brilliantly birefringent representing the PLLA material. At low magnification of a section through a screw-head and bone plate, areas with densely packed birefringent PLLA-like material were seen. These areas were separated by fibrous tissue that varied in thickness throughout the section. In parts where the capsule measured about 150 #m, a sharp boundary between the closely packed PLLA material and the fibrous tissue was seen (Figure 2a). Here, no birefringent material was situated in the capsule. These parts of the capsule, with a sharp interface PLLA/fibrous tissue, were investigated ultrastructurally. Transmission electron microscopy (TEM) revealed densely packed foreign body material with a lamellar or needle-like structure in close contact with orientated bundles of collagen fibres (Figure 2b). These non-electron dense needle-like particles represented the PLLA material. Between these fibres long slender cells were present that could be characterized on morphological grounds to be fibrocytes. Virtually no other ceils like macrophages, foreign body giant cells or lymphocytes were seen in this area. No PLLA was situated in the cytoplasm of cells or in the extracellular space between the collagen fibres. Light microscopically it was observed that in other parts of the section the fibrous tissue spread out, and PLLA particles were seen between bundles of collagen and cells. In these parts the cells most frequently present were long slender fibrocytes. Only a small number of macrophages or lymphocytes were seen. Electron microscopy revealed that fibrocytes possessed well developed organelles like rough endoplasmic reticulum and golgi apparatus. In a number of cells with internalized PLLA, a clear deposition of glycogen around phagosomes was seen. In the plasma membranes of the fibrocytes a high number of endocytotic vesicles was observed. Intracellularly, fibrocytes showed a profuse amount of PLLA material which was mostly packed in membrane-b~ vacuoles which could be described as phagosomes (Figure 3). Fusion
Figure 2 a, Micrograph of the 3.3 year implanted material, taken under crossed Nicols prisms, of the fibrous capsule with centrally removed poly(L-lactic) acid (PLLA) particles (RP). The arrows indicate a part of the capsule with a sharp interface PLLA/fibrous tissue. In other parts birefringent PLLA particles (P) were situated between bundles of collagen (C) (original magnification x40). b, Transmission electron microscopic (TEM) photograph of the 3.3 year implanted material. The arrows indicate a sharp interface of densely packed needle-like PLLA material (P) and orientated bundles of collagen fibres (C). These orientated bundles of collagen are intermingled wffh fibrocytic (F) cells (scale bar = 2.5 #m).
of a lyosome with a packed vacuole forming a phagolysosome was observed only infrequently. In a small number of fibrocytes the incorporated PLLA particles were also situated apparently freely in the cytoplasm. The cells with incorporated PLLA material had swollen parts of endoplasmic reticulum, which could indicate an increased protein synthesis, and swollen mitochondria that lacked the cristae suggesting physiological damage. Although PLLA particles were seen between collagen fibres and cells, the bulk of the PLLA material was still situated extrae~llularly and was not interlaced with collagen fibres and cells. Macroscopical investigation of the 5.7 years material showed a tissue mass which lacked the firm consistency of the 3.3 years material. The contours of some of the screw-heads were still visible. Microscopic examinatiort at low magnification showed a sharp outline of a thin fibrous capsule. At the peripheral Biomaterials 1995, Vol. 16 No. 1
104
The Biomaterials Silver Jubilee Compendium 28
Late degradation tissue response to poly(L-lactide) bone plates and screws: J.E. Bergsma et al.
Figure 3 Transmission electron micrograph of a fibrocyte with glycogen accumulation (arrows) around phagosomes and swollen mitochondria (M). The cell is situated in fields of poly(L-lactic) acid particles (P) and sheets of collagen (C) (scale bar = 1.7 #m).
Figure 4
Birefringent poly(L-lactic) acid particles intraceliularly in foamy macrophages (arrows), 5.7 years after implantation (original magnification • taken under crossed Nicol prisms).
parts of the screw-head region, blood capillaries, nerve fibres and fat deposition were observed. More centrally in the screw-head region, randomly orientated bundles of collagen were seen amidst various kinds of cells. In this section, no large areas of densely packed extracellular PLLA particles were found. A section through the bone plate showed large fields of cells with intracellularly positioned birefringent PLLA material (Figure 4). This section was composed of mainly foamy macrophages surrounded by fibrous tissue. Electron microscopic observations revealed that most of the PLLA material was situated intracellularly in various cells. These phagocytizing cells formed clusters that were encapsulated by mature and randomly orientated bundles of collagen. The number of elongated fibrocytic cells with internalized PLLA material had diminished as compared to the situation after 3.3 years. The number of macrophages and foreign body giant cells had increased and represented the Biomaterials 1995, Vol. 16 No. 1
Figure 5 Membrane-bound conglomerates of poly (L-lactic) acid particles (arrows) described as phagosomes in the cytoplasm of a phagocytic cell. All cells are embedded in a mature fibrous capsule (C) (scale bar - 2.5/~m).
major phagocytizing cellular component. In these cells the PLLA particles were no longer found freely in the cytoplasm, but entirely as membrane-bound conglomerates (Figure 5). The morphology of phagocytizing cells showed minimal signs of cell damage. The cytoplasmic organelles, such as mitochondria and rough endoplasmic reticulum, were of normal appearance. A sample of trephined bone was obtained from a patient after an implantation period of 5.7 years. The tapped screw-holes were still visible and not fully filled in with cortical bone. On a section perpendicular to the long axis of the screw-hold, birefringent PLLA particles were still densely packed in the shape of the screw-thread. These densely PLLA particles were not interlaced with collagen fibres or cells. A fibrous capsule was situated between cortical bone and the bulk of the PLLA particles and had at some parts spread out into the lacunae of the cortical bone (Figure 6). Fields of PLLA particles were seen up to 0.5 mm from the original implant site showing birefringent PLLA material between sheets of collagen and in various cells. Ultrastructural investigation showed that the PLLA material had the same lamellar or needlelike structure as observed in the soft tissue. As an indicator of cell damage or high lactic acid concentrations, possibly released from the PLLA particles, the presence of lactate dehydrogenase (LDH) was investigated. The presence of LDH in mitochondria was demonstrated by a cytochemical reaction: Hatchett's brown depositions amplified by treatment with 3,3'-diaminobenzidine (DAB) and osmication were seen in close relation to the cristae, the intracristate spaces and the intermembrane spaces of mitochondria (Figure 7). There was no evidence of LDH-related precipitates in the membrane-bound conglomerates, or in close contact with the PLLA particles; nor were extracellular LDH-related precipitates, possibly released by damaged or lytic cells demonstrated. Control specimens, without nicotinamide adenine dinucleotide (NAD) and/or DAB/osmium showed no Hatchett's brown depositions in mitochondria. Acid phosphatase could be demonstrated in lysosomes in a
105
The Biomaterials Silver Jubilee Compendium Late degradation tissue response to poly(L-lactide) bone plates and screws: J.E. Bergsma et al.
Figure 6 Densely packed poly(L-lactic) acid (PLLA) particles (P) in the shape of the screw-thread surrounded by cortical bone (B). Fields of PLLA particles were seen at some distance of the original implant in the cortical bone (P') (original magnification x40).
Figure 7 Hatchett's brown depositions amplified by treatment with 3,3'-diaminobenzidine and osmication were seen in close relation to the cristae, the intracristate spaces and the intermembrane spaces of mitochondria (scale bar = 0.25/~m).
limited number of macrophages that had internalized PLLA particles. Some of these lysosomes were seen in close contact with the PLLA-bearing phagosomes. Fusion of a primary lysosome and a phagosome forming a phagolysosome was rarely seen. DISCUSSION The total resorption time of as-polymerized PLLA was estimated in previous studies to be 3.5 years 15' 16. The results of this experiment show that the PLLA bone plate and screws, implanted for 3.3 years, had degraded into fragments and disintegrated into particles that have a needle-like structure on TEM. Ultrastructural TEM analysis of the PLLA material with an implantation period of 5.7 years shows a comparable morphology. SEM analysis would suggest that the average particle size of the materia] implanted
29
for 5.7 years is much smaller. Between 3.3 and 5.7 years the PLLA material degrades from fragments into particles that have a needle-like structure an TEM. Light microscopic observations suggested that the number of PLLA particles that were internalized by cells had increased with longer implantation periods. The molecular weight, about 5000, is identical for both implantation periods. Rozema 21 described that an M, of 5000 may be a break-even point as a start of relative high disintegration. However, the PLLA particles have a rather high crystallinity 21 which is probably one of the factors that makes them very stable and not very susceptible to hydrolysis. This may explain the very limited progression of the degradation of PLLA particles in the period from 3.3 to 5.7 years. Substantial mass loss or total resorption had not taken place up to 5.7 years. If a PLLA particle degrades, it is probably in non-detectable oligomers that are washed away with tissue fluids and are not detected in the material analysis. This mechanism may account for the same values of molecular weight and crystallinity for both implantation periods. The origin of the described swelling is not quite clear. Maybe the swelling is initiated by a gradual disintegration of the PLLA bone plate and screws into fragments. Bergsma et al. TM described how during degradation the PLLA plates and screws disintegrate into small fragments which may lead to an increased volume in comparison with the volume of the intact bone plate and screws. In a cross-section of tissue implanted for 3 years, the surface area occupied by the a-cellular PLLA particles was estimated to be 65% of the total surface area. The remaining 35% of the cross-section was occupied by the enveloping fibrous capsule. B6stman eta]. 22, in a study with intraosseously placed polyglycolide screws and pins, suggest that an increased osmotic intracavital pressure associated with the degradation of polyglycolide and the resistance of the surrounding tissue may determine the formation of a sinus. The origin of the described swelling may possibly be explained by a combination of factors such as the disintegration of the PLLA material into small particles, and an increased osmotic pressure caused by these fragments and the, compared to bone, low resistance of the subcutaneous tissue. Another mechanism that may induce or maintain the swelling is given by Fornasier et oi. 23, who described a correlation between the presence of birefringent polyethylene particles, the density of histiocytes and the thickness of a fibrohistiocytic membrane all of which showed an increase with time. A section obtained from the material with an implantation period of 5.7 years consists of a thin fibrous capsule and sheets of collagen interlaced with various cells. In contrast to the material that was implanted for 3.3 years, scarcely any PLLA material can be found in the extracellular space. The majority of the PLLA crystals has been internalized by phagocytizing ceils in membrane-bound vacuoles. These results may lead to the conclusion that with longer implantation periods there is a gradual shift of PLLA particles from extra- to intracellular in phagocytic cells that are imbedded in a fibrous matrix. The presence of macrophages and fibrocytes in response to the PLLA particles can be Biomaterials 1995, Vo]. 16 No. 1
106
The Biomaterials Silver Jubilee C o m p e n d i u m
30
Late degradation tissue response to poly(L-lactide) bone plates and screws: J.E. Bergsma et al.
expected since macrophages are known to phagocytize and remove foreign body material. As a response to internalization of the foreign body material macrophages can activate and attract fibroblast-like ceils. The extracellular degradation of the PLLA particles is probably a hydrolytic process. However, phagocytizing cells, especially macrophages, can release a number of lysosomal hydrolytic enzymes that may influence the degradation. If this is the case then an increased concentration of the lysosome guide enzyme, acid phosphatase, would be expected. Acid phosphatase is present in all lysosomes and its easy identification makes it an excellent marker. In the tissue with implantation periods of 5.7 years the presence of acid phosphatase was demonstrated, although not in abundance. Another enzyme that has been studied is lactic dehydrogenase (LDH). LDH converts lactic acid into pyruvate that can be metabolized in the citric acid cycle. If a substantial amount of intracellular PLLA particles degrades into lactic acid an increase might be expected. Again, the presence of enzyme-related precipitates were demonstrated but not in large amounts. Although a very limited number of enzymes were investigated these results may lead to the conclusion that the PLLA particles are eventually all internalized by phagocytizing cells that cannot actively degrade the PLLA particles. Hydrolysis is probably the only degradation mechanism and the highly crystalline particles seem to degrade very slowly. This implies that there is a long lasting presence of intracellular PLLA particles or that the particles are egested into the extracellular space because the cell cannot actively degrade the particles. Indigestible foreign body particles may cause a continuous attraction of macrophages that may again phagocytize the PLLA particles and thus repeat the intracellular cycle. Based on the literature on silicone implants another possibility may be that PLLA particles, or macrophages with PLLA particles, migrate to nodal tissue from the implant site 24. In this study no lymph nodes were excised, but perhaps in future studies the possibility of migration of PLLA particles to lymph nodes should be investigated. In the orthopaedic literature many studies have been published about aseptic loosening of prosthetic joints due to the presence of particulate polymer debris found within fibrous tissue, macrophages and foreign body cells. Horowitz et al. 25 described in an in vitro study that exposure to polymethylmethacrylate (PMMA) particles inhibits macrophage DNA synthesis, impairs their cytotoxic ability and eventually kills the cells. In our study cells that had internalized the lamellar or needle-like PLLA particles showed signs of mild cell damage such as enlarged rnitochondria and accumulation of glycogen. Human fibroblasts in culture accumulate glycogen in their cytoplasm as they approach senescence. In the 5.7 year specimens no signs of cell damage were observed. When an implanted material causes cellular damage, an increase in the leakage of intracellular lactate dehydrogenase may be expected. The damaging effect of the PLLA particles seems to be very low, no increased amounts Biomaterials 1995, Vol. 16 No. 1
of mitochondrial LDH could be demonstrated, so it may be assumed that the internalized PLLA crystals do not cause severe cell injury or cell death. The PLLA particles will probably induce a macrophage and fibrocyte response. The time needed for total hydrolytic degradation of the PLLA crystals will probably determine the duration of the swelling. The results of the trephined bone from the patient with an implantation period of 5.7 years, show a number of differences compared to the results of subcutaneously implanted material. The degradation of the PLLA screw-thread resembles the degradation of the PLLA bone plate, but the screw remnants are not interlaced with collagen fibres and internalization of PLLA particles by phagocytic cells is very limited. These results may indicate that there can be a variation in the degradation mechanism between subcutaneous and intraosseous PLLA implants and the histological reaction the implant induces. These differences may be explained by the fact that perhaps cortical bone can withstand the osmotic pressure of the degrading material and thus prevent swelling of the PLLA material. The PLLA material remains densely packed which perhaps prevents cellular ingrowth and internalization of PLLA particles. In summary, the disintegration of PLLA into particles with the accompanying increase in volume of the PLLA material itself and the fibrous tissue, may explain the origin of the described swelling. The PLLA particles with a very slow hydrolytic degradation rate, although not very irritable to the cell, do induce a cellular reaction. These are processes that resemble those seen in aseptic bone loosening in orthopaedic applications. The biocompatibility of the non-degraded PLLA material has been established in a number of studies. The degraded PLLA particles do not cause major cell injury but they can induce and maintain a clinically detectable swelling which could imply that these PLLA particles can no longer be considered to be fully biocompatible. Future research has to focus on biodegradable polymers that do not disintegrate into highly crystalline particles to avoid very long degradation periods, and in some applications a clinically detectable swelling.
REFERENCES 1
2
3
4
Akeson WH, Woo SL-Y, Coutts RD, Matthews JV, Gonsalves M, Amiel D. Quantitative histological evaluation of early fracture healing of cortical bones immobilized by stainless steel and composite plates. Calcif Tiss Res 1975; 19: 27-37. Paavolainen P, Karaharju E, Slatis P, Ahonen J, Holmstorm T. Effect of rigid plate fixation on structure and mineral content of cortical bone. Clin Orthop Rel Res 1987; 136: 287-293. Simon BR, Woo SL-Y, McCarthy M, Lee S, Akeson WH. Parametric study of bone remodeling beneath internal fixation plates of varying stiffness. J Bioeng 1978; 2: 543-556. Tunc DC, Rohovsky MW, Lehman WB, Strongwater A, Kummer F. Evaluation of body absorbable bone fixation devices. Proc 31st Ann Orthop Soc, Las Vegas, Jan 1985: 165.
107
The Biomaterials Silver Jubilee Compendium Late degradation tissue response to poly(L-lactide) bone plates and screws: J.E. Bergsma et ai.
5
6
7
8
9
10
11 12
13
14
15
Christel P, Vert M, Chabot F, Abols Y, Leray JL. Polylactic acid for intramedullary plugging. In: Advances in Biomaterials Vol. 5 (ed Ducheyne P}. Amsterdam: Elsevier Science; 1984: 1-6. Vert M, Christel P, Chabot F, Leray J. Bioresorbable plastic materials for bone surgery. In: Macromolecular Biomaterials, (eds Hastings GW, Ducheyne P). Boca Raton, FL: CRC Press; 1984: 120-142. B6stman O, M~ikel~i EA, T6mtil~i P, Rokkanen P. Transphyseal fracture fixation using biodegradable pins in children. J Bone Joint Surg [Br] 1989; 17-B: 7067O7. Suuronen R. Comparison of absorbable self-reinforced poly-L-lactide screws and metallic screws in the fixation of mandibular condyle osteotomies: An experimental study in sheep. J Oral Maxi]lofac Surg 1991; 49: 989-995. Leenslag JW, Pennings AJ, Bos RRM, Rozema FR, Boering G. Resorbable materials of poly(L-lactide). VI. Plates and screws for internal fracture fixation. Biomaterials 1987; 8: 70-73. Gogolewski S, Pennings AJ. Resorbable materials of poly(L-lactide) 2. Fibers spun from solutions of poly(Llactide) in good solvents. ! Appl Polym Sci 1983, 28: 1045-1061. Gogolewski S, Pennings AJ. Resorbable materials of poly{L-lactide). 3. Porous materials for medical application. Colloid Sci 1983; 261: 477-484. Leenslag JW, Gogolewski S, Pennings AJ. Resorbable materials of paly(t.-lactide). 5. Influence of secondary structure on the mechanical properties and hydrolyzability of poly(u-lactide) fibers produced by a dryspinning method. J Appl Polym Sci 1984; zg: 2829-2842. Bos RRM, Rozema FR, Boering G, Nijenhuis AJ, Pennings AJ, Verwey AB. Bioabsorbable plates and screws for internal fixation of mandibular fractures. A study in six dogs. Int J Oral Maxillofac Surg 1989; 18: 365-369. Bos RRM, Rozema FR, Boering G, Nijenhuis AJ, Pennings AJ, Jansen HWB. Bone plates and screws of bioabsorbable poly(L-lactide). An animal pilot study. Br J Oral Maxillofac Surg 1989; 27: 467-476. Rozema FR, Bos RRM, Pennings AJ, Jansen HWB. Poly(L-lactide) implants in repair of defects of the orbital floor. An animal study. ] Oral Maxillofac Surg 1990; 48: 1305-1309.
16
17
18
19
20
21
22
23
24
25
31
Bos RRM, Rozema FR, Boering G et al. Degradation of and tissue reaction to biodegradable poly(L-lactide) for use as internal fixation of fractures. A study in rats. Biomateriols 1991; lZ: 32-36. Bos RRM, Rozema FR, Boering G, Nijenhuis AJ, Penniags AJ, Verwey AB. Resorbable poly(L-lactide) plates and screws for the fixation of unstable zygomatic fractures. J Oral Maxillofac Surg 1987; 45: 751-753. Rozema FR de Bruijn WC, Bos RRM, Boering G, Nijenhuis AJ, Pennings AJ. Late tissue response to bone-plates and screws of poly(L-lactide) used for fracture fixation of the zygomatic bone. In: Doherty PJ, Williams RL, Williams DF, eds, Biomaterial - Tissue Interfaces, Advances in Biomaterials vol. 10. Amsterdam: Elsevier, 1992: 349-355. Bergsma JE, Rozema FR, Bos RRM, de Bruijn WC. Foreign body reactions to resorbable poly(L-lactide) bone plates and screws used for the fixation of unstable zygomatic fractures. J Oral Maxillofac Surg 1993; 51: 666-670. Hulstaert CE, Kalicharan D, Hardonk MJ. Cytochemical demonstration of phophatases in the rat liver by a cerium-based method in combination with osmiumtetroxide and potassium ferrocyanide post-fixation. Histochemistry 1983; 78: 71-79. Hanker JS, Kusyk CJ, Bloom FE. The demonstration of dehydrogenases and monoamine oxidase by the formation of osmium blacks at the sites of Hatchett's brown. Histochemie 1973; 33: 205-230. BSstman O, P~iiv~irinta U, Manninen M, Rokkanen P. Polymeric debris from absorbable polyglycolide screws and pins. Acta Orthop Scand 1992; 63: 555559. Fornasier V, Wright J, Seligman J. The histomorphologic and morphometric study of asymptomatic hip arthroplasty. A postmortem study. Clin Orthop 1991; 271: 272-282. Dolwick MF, Aufdemorte TB. Silicone-induced foreign body reaction and lymphadenopathy after temperomandibular joint arthroplasty. Oral Surg Oral Med Oral Pathol 1985; 59: 449-452. Horowitz SM, Gautsch TL, Frondoza CG, Riley Jr L. Macrophage exposure to polymethyl methacrylate leads to mediator release and injury. J Orthop Res 1991; 9: 406-413.
Biomaterials 1995, Vol. 16 No. 1
108
This Page Intentionally Left Blank
The Biomaterials Silver Jubilee Compendium
109
The Biomaterials Silver Jubilee Compendium
Biomaterials 16 (1995) 297-303 9 1995 Elsevier Science Limited Printed in Great Britain. All rights reserved 0142-9612/95/$10.00
r~u TT ERWO R T H ~l~E I N E M A N N _
Mechanism of cell detachment from temperature-modulated, hydrophilichydrophobic polymer surfaces Teruo Okano, Noriko Yamada, Minako Okuhara, Hideaki Sakai and Yasuhisa Sakurai
Institute of Biomedical Engineering, Tokyo Women's Medical College, 8-1 Kawada, Shinjuku, Tokyo 162, Japan
Poly(N-isopropylacrylamide) (PIPAAm), exhibiting a lower critical solution temperature (LCST) at 25 ~C in physiological phosphate buffered saline solution (pH 7.4) and at 32~ in pure water, was grafted onto the surfaces of commercial polystyrene cell culture dishes. This PIPAAm-grafted surface exhibited hydrophobic surface properties at temperatures over the LCST and hydrophilic surface properties below the LCST. Endothelial cells and hepatocytes attached and proliferated on PIPAAmgrafted surfaces at 37~ C, above the LCST. The cultured cells were readily detached from these surfaces by lowering the incubation temperature without the usual damage associated with trypsinization. In this case, the optimum temperature for cell detachment was 10~ for hepatocytes and 20~ for endothelial cells. Cell detachment was partially inhibited by sodium azide treatment, suggesting that cell metabolism directly affects cell detachment. Morphological changes of the adherent cells during cell detachment experiments indicated further involvement of active cellular metabolic processes. Cells detached from hydrophobic-hydrophilic PIPAAm surfaces not only via reduced cellsurface interactions caused by the spontaneous hydration of grafted PIPAAm chains, but also by active cell morphological changes which were a function of cell metabolism. Biomaterials (1995) 16 (4), 297-303
Keywords: Thermoresponsive polymer surface, cell culture, cell detachment, hepatocyte, endothelial cell Received 21 December 1993; accepted 25 April 1994
Poly(N-isopropylacrylamide) (PIPAAm), a thermoresponsive polymer, exhibits a lower critical solution temperature (LCST) of about 32 ~ in water a.2. PIPAAm is fully hydrated with an extended chain conformation in aqueous solutions below 32~ and is extensively dehydrated and compact above 32~ These unique thermosensitive polymers and their copolymers have therefore been utilized in temperature-modulated bioconjugates constructed by a bioactive molecule and a stimuli-responsive PIPAAm chain. Thermally modulated to induce soluble-insoluble changes in solution, these new bioconjugates have attracted considerable attention in both fundamental research and practical application, such as bioactivity control and bioseparations in protein engineering 3-~. Further, cross-linked PIPAAm and its copolymers have been developed as thermal on-off switching polymers for drug permeation and release 7-1~ We have studied thermal on-off modulation of hydrophilic-hydrophobic changes on PIPAAm-grafted surfaces 11'12. Cells cultured on hydrophobic PIPAAmgrafted surfaces at 37~ (above the LCST of 32~ were prompted to detach spontaneously by lowering
the medium temperature and changing the hydration of the PIPAAm chains. Cultured cells generally will adhere to hydrophobic surfaces but not on highly hydrated hydrophilic surfaces 13'14. Our study clearly demonstrated the feasibility of a new recovery strategy for harvesting cultured cells by external modulation of thermoresponsive surfaces. In fact, PIPAAm-grafted surfaces demonstrate very effective thermal switching to reverse hepatocyte and endothelial cell attachment and detachment without cell damage 11'15. Cell adhesion onto a material surface can be arbitrarily classified as a two-step mechanistic process: the first stage is controlled by complex combinations of physicochemical interactions including hydrophobic, coulombic, and van der Waals forces between the cell membrane and the material surface. This process might be termed 'passive adhesion' according to this adsorption mechanism. The second stage might be considered as 'active adhesion', because of the participation of cellular metabolic processes. Attached ceils are well-known for changing their shapes and expending metabolic energy in order to stabilize the interface between their membrane and the underlying materials, by both physicochemical and biological mechanisms16,17
Correspondence to Dr Teruo Okano. 297
Biomaterials 1995, Vol. 16 No. 4
110
The Biomaterials Silver Jubilee Compendium 298
When surface properties of the material are changed from hydrophobic to hydrophilic, cells often attempt to detach themselves from the surface as mentioned above. In this case, cells may not be able to detach from the surface without involving cellular metabolic processes which actively change membrane morphology. This paper attempts to clarify the influence of cell metabolic processes on cell detachment. No metabolic effects on cell detachment would be manifested by increasingly rapid cell detachment with decreasing temperature as PIPAAm hydration is enhanced when temperature is reduced. Mechanisms of cell detachment were discussed with regard both to effects of temperature-modulated surface changes and cell metabolic changes.
MATERIALS AND M E T H O D S Preparation of poly(IPAAm)-grafted surfaces The procedures for the preparation of PIPAAm-grafted cell culture dishes are described elsewhere 11'15. NIsopropylacrylamide monomer (IPAAm) (Eastman Kodak, Rochester, NY, USA) was dissolved in isopropyl alcohol. IPAAm solution (45wt%, 0.1ml) was added to each polystyrene tissue culture dish (Falcon 3001, diameter 35mm, Falcon Becton Dickinson Labware, Oxnard, CA, USA) and then irradiated with a 0.25 MGy electron beam (200kV, under 1.3 x 1 0 -4 Pa) using an Area Beam Electron Processing System (Nisshin High Voltage, Kyoto, Japan). IPAAm was polymerized and grafted onto the surfaces of the dishes using an electron beam. The PIPAAm-grafted dishes were rinsed with cold distilled water to remove non-grafted IPAAm, dried under nitrogen gas and gassterilized by ethylene oxide before use in cell culture experiments. Untreated Falcon 3001 dishes were used as controls. Homogeneous coverage of PIPAAm-grafted dishes was confirmed using field emission scanning electron microscopy. The amount of grafted PIPAAm polymer can be controlled by the concentration of IPAAm solution in each preparation. Surfaces of these PIPAAm-grafted dishes change reversibly between hydrophilic and hydrophobic by controlling temperature, as previously reported '5.
Cell culture Endothelial cells were isolated from bovine thoracic aorta by a dispase digestion method described previously' 5. Endothelial cells were cultured in Dulbecco's modified Eagle's Minimal Essential Medium (DMEM) (Gibco, Grand Island, NY, USA) supplemented with 10% fetal bovine serum (FCS) (Gibco), lOOUm1-1 of penicillin (Gibco), 100~gm1-1 of streptomycin (Gibco) and 2.Spgm1-1 of fungizone (Gibco) at 37~ in a fully humidified atmosphere of 5% CO2 in air. The cells were subcultured by treatment with 0.05% trypsin and 0.02% ethylenediaminetetraacetic acid (EDTA) solution (Gibco) for 5 min after confluent cell monolayers had formed. Endothelial ceils from the third to sixth passages were used in all experiments. Rat hepatocytes were isalated from 5-week-old male Biomaterials 1995, Vol. 16 No. 4
Cell detachment from thermoresponsive surface: T. Okano et al.
Wistar rats, weighing about 150g, using an in situ collagenase perfusion method previously reported '5"18. More than 98% of the cells obtained were parenchymal cells as determined by phase-contrast microscopy, and more than 90% were viable as measured by trypan blue dye exclusion. The hapatocytes were cultured in Williams E medium (Gibco) supplemented with 5% FCS, 10 ngm1-1 human epidermal growth factor (hEGF) (Wakunaga Pharmaceutical, Osaka, Japan), 10mM nicotinamide (Wako, Pure Chemical Industries, Osaka, Japan), 5 U m l - ' aprotinin (Wako), 1 0 - 7 M insulin (Sigma Chemical, St Louis, MO, USA), lO-aM dexamethasone (Sigma) and 50mgml-1 canamaicin sulphate (Gibco) at 37 ~ under a humidified atmosphere of 5% CO2 in air.
Measurement of DNA Cell numbers were calculated by DNA content in cultured cells. The amount of DNA was assayed fluorometrically with calf thymus DNA (type 1, Sigma) as the standard 15,19. Briefly, cells were solubilized with 10mM EDTA solution, pH 12.3 (Kanto Chemical, Tokyo, Japan) for 30min at 37~ and neutralized by the addition of 1M potassium dihydrogen phosphate (KH2PO4) (Kanto) solution and then mixed with 2'-(4hydroxyphenyl)-5-(4-methyl-l-piperazinyl)-2, 5'-bi-lHbenzimidazole (Hoechst 33258) solution (Sigma). DNA content was determined by fluorimetry (JASCO FP-770 spectrofluorometer; Japan Spectroscopic Co, Tokyo, Japan) at 360 nm excitation and 450 nm emission.
Influence of temperature on cell detachment Rat hepatocytes and endothelial cells were seeded onto PIPAAm-grafted and control dishes at a density of 4 x 1 0 4 cells cm -2 and cultured in their respective culture medium at 37 ~ under a humidified atmosphere of 5% CO2 in air. After 2 d, the temperature of the cell culture system was decreased from 37~ to T~ (4, 10, 15, 20 and 27 ~C) by changing with medium of T~ and cooling the culture dishes. Culture dishes containing cells were incubated at T~ for 30 min and cells detached from cell culture dish surfaces were estimated after 5min additional incubation at 25~ Detached cells from cell culture dishes were collected and cell number was estimated from DNA measurement.
Effect of sodium azide on hepatocyte detachment Sodium azide (Nacalai Tesque, Kyoto, Japan) was dissolved in culture medium for hepatocytes and adjusted to concentrations of 0, 0.2, 1.0 and 2.0ruM. Rat hepatocytes were seeded on PIPAAm-grafted and control dishes at an initial density of 4 x 1 0 4 cells c m -2 and cultured for 2 d in culture medium for hepatocytes at 37~ under a humidified atmosphere of 5% CO2 in air. Culture dishes were then changed with a culture medium containing sodium azide at each concentration, and incubated for 60rain at 37 C,C. After treatment with sodium azide, the culture dish incubation temperature was reduced and incubated at 10~ for 30min followed by additional incubation at 25~ for 5min. Cells detached from cell culture dishes were then collected and cell number was determined by
111
The Biomaterials Silver Jubilee Compendium Cell detachment from thermoresponsive surface: T. Okano et al.
DNA measurement as described above. Moreover, to clarify the influence of cell metabolism in cell detachment, the effect of sodium azide on the temperature dependence of cell detachment was also investigated. Hepatocytes were cultured on PIPAAm-grafted dishes at 37~ for 2 d as described above. After pretreatment with or without 2 mM sodium azide solution for 60min at 37 ~ C, the culture dishes were incubated at T~ (4, 10, 15, 20 and 27~ for 30min and an additional 5 m i n at 25 ~ The percentage of detached cells was calculated and cell numbers were determined.
Cell morphology by optical and scanning electron microscopy Hepatocytes were cultured on PIPAAm-grafted dishes as described above. After 2 d culture at 37~ the culture dishes were incubated at 10~ for 30min and an additional 5 m i n at 25~ The morphological changes of detaching hepatocytes on dish surfaces were directly and continuously observed by phasecontrast microscopy (Nikon Diaphot-TMD, Tokyo, Japan) using a micro-cool plate (Kitazato Supply, Sizuoka, Japan). For scanning electron microscopy (SEM), detaching hepatocytes on dish surfaces were fixed at 10~ for 6 0 m i n with 2% glutaraldehyde (EM Sciences, Fort Washington, USA) in 0.1M cacodylatebuffered solution (EM Science), pH 7.4. The fixed cells were washed with cacodylate-buffered solution and then lyophilized. After sputter-coating with gold, the samples were observed using a scanning electron microscope (JEOL JSM5300LV, Tokyo, Japan).
RESULTS AND DISCUSSION
Influence of temperature on hepatocyte detachment For primary rat hepatocytes, cell growth curves were observed to be similar on PIPAAm-grafled and control dishes as reported previously 15. After hepatocytes were cultured for 2 d at 37 ~C, the temperature of the cell culture systems was decreased from 37 to T~ by both cooling the culture dishes and exchanging the medium with fresh medium at T~ After 30min incubation at T ~C, an additional 5 min incubation was performed at 25~ (Figures 1, curve A) and compared with results at constant temperature of T~ (curve B). Figure I shows the correlation between the percentage of detached cells and incubation temperature, T~ Cells remained over 85% attached at both 30 and 35 rain incubation times after the cell culture systems were decreased from 37 to T ~C. At lower temperatures, 4 and 10 ~C, the number of detached cells was smaller than at higher temperature, 15 and 20 ~C, at both 30 min (C) and 35 min (B) incubation times. These results show that cell detachment is not directly correlated with reduced temperature. Grafted PIPAAm chains are assumed to be hydrated and maintain expanded conformations at lower temperatures resulting in reduced interactions between the cells and grafted surfaces of the cell culture dishes. Hydration of PIPAAm at the cell-material interface, therefore, does not completely govern cell detachment from culture dish surfaces. As cell metabo-
299
100 80 A tj "{3 t3 to
60
40
Q
20
_ ~
0
..
9
.... !
L
10
_
I .
20
L
II
,,
30
_
40
Temperature (~ Figure 1 Hepatocyte detachment from PIPAAm-grafted surfaces by reducing temperature, and subsequent additional 5-min incubation at 25~ (A) and T~ (B) after
30-min incubation at T~
(C) (T = 4, 10, 15, 20, 27 ~C).
lism is suppressed by decreasing temperature, influences of the cell metabolic processes as well as hydration of the culture surface on cell detachments are implicated. In fact, 30 min incubation at T~ followed by a temperature change to 25~ in order to increase cell metabolism drastically enhances cell detachment, as shown in Figure 1 (curve A). In this case, numbers of detached cells show a maximum at 10 ~C. These results demonstrate that cell detachment is controlled not only by the hydration of grafted PIPAAm on the culture dishes but also by active cellular metabolism. Morphologies of detaching cells were observed by both optical and electron microscopies over time after 30 rain incubation at 10 ~C followed by a temperature increase to 25 ~C as shown in Figures 2 and 3, respectively. Cells start to change their shape from a spread to a rounded form, and finally cells are observed to detach completely from the surface. After 10min incubation at 25~ 100% of cells were detached. These results clearly demonstrate that the cell detachment process involves cell shape changes accompanying a consumption of cellular metabolic energy. Hydration changes of grafted PIPAAm at the cell-material interface is an important initial stimulus to induce active cell detachment mediated by cellular processes.
Effect of sodium azide on hepatocyte detachment Sodium azide is a known inhibitor of cytochrome C oxidase in mitochondoria and decreases ATP generat i o n 2~ resulting in the disruption of cellular activities which require ATP. The effect of sodium azide on cell detachment was investigated to clarify the role of cell metabolism in cell detachment. Cultured cells were damaged and detached from Biomaterials 1995, Vol. 16 No. 4
112
The Biomaterials Silver Jubilee Compendium 300
Cell detachment from thermoresponsive surface: T. Okano et al.
Figure 2 Phase contrast micrographs showing the process of hepatocyte detachment from PIPAAm-grafted surfaces at 25 ~ C after 30-min incubation at 10 ~ C.
culture surfaces when sodium azide over 10raM was added. At concentrations below 2mM sodium azide, cells were not observably detached from culture dish surfaces and cell metabolism was only partially inhibited. After 2 d cultured cells were treated with sodium azide for 60min at 37~ then incubated at 10~ for 30min followed by additional incubation at 25~ for 5 min. Cell detachment was scarcely observed with and without additions of up to 2mM sodium azide, as shown in Figure 4. By contrast, the temperature-responsive cell recovery system shows significant inhibition of cell detachment at reduced temperatures. This inhibition on PIPAAm-grafted surfaces was increased with increasing sodium azide. Since sodium azide treatment under these conditions may not completely inhibit cellular ATP generation, the inhibition of cell detachment observed is partial but not complete, as shown in Figure 4. Figure 5 shows detached cell percentages from PIPAAm-grafted dishes with or without sodium azide treatment followed by incubation at T~ for 30min and additional 5 min at 25 ~C. The inhibition effects of 2ram sodium azide treatment on cell detachment was observed over all temperatures. These results strongly suggest that active cellular processes enabling cell morphological changes are essential procedure to complete cell detachment from hydrated surfaces.
Mechanism of endothelial cell detachment Endothelial cells are readily cultured on the PIPAAmgrafted surfaces and proliferate the same as on commercial dishes, as shown in a previous paper 15. Detachment of endothelial cells was investigated using the same experiments for temperature-modulated cell detachment on PIPAAm-grafted surfaces. More significant cell Biomaterials 1995, Vol. 16 No. 4
detachment maxima for endothelial cells are observed, as shown in Figure 6. The maximum at 20 ~C - - a significantly higher temperature than in the case of hepatoc y t e s - - i s evident. As discussed above with regard to hepatocytes, the detachment of endothelial cells is also controlled by two steps: the initial temperature-responsive PIPAAm surface hydration, and active processes of cell detachment accompanying cell shape changes. Cell detachment is not significant at lower temperatures, even if PIPAAm swelling is more significant at lower temperatures. The maximum peak for endothelial cell detachment at 20~ suggests that endothelial cell metabolism is inhibited more significantly at lower temperatures compared with hepatocytes. Since different cell lines exhibit different temperature sensitivities, the maximum peak for detachment of endothelial cells at a higher temperature than that for hepatocytes is another important result to support a metabolically related cell detachment mechanism. After the culture dish was changed from 37 to 25 c' C, the temperature-induced increase in PIPAAm swelling was no enough to initiate cell detachment. However, dishes reduced from 37 to 20~ increased PIPAAm hydration sufficiently to initiate cell detachment. A 30-min incubation at 20~ allows PIPAAm chains to hydrate and expand their conformations. However, this alone is not sufficient to induce cell detachment because the temperature is not high enough to allow cellular metabolism and accompanying morphological changes. Therefore, cell detachment is significantly enhanced by increasing the temperature at 25~ sufficient to induce observable cellular shape changes. Even if PIPAAm chains dehydrate slightly by increasing this temperature from 20 to 25~C, metabolic changes of cells at this higher temperature seem to be much more significant than hydration changes of PIPAAm-grafted chains.
113
The Biomaterials Silver Jubilee Compendium 301
Cell detachment from thermoresponsive surface: T. Okano et al.
Figure 4 Effect of sodium azide concentration on hepatocyte detachment. Hepatocytes were pretreated with sodium azide for 60min at 37~ and then detached by 30-rain incubation at 10~ and an additional 5min at 25 ~
Figure 3 Scanning electron micrographs showing the process of hepatocyte detachment from PIPAAm-grafted surfaces at 25~ after 30-rain incubation at 10~ C_ a, 0 rain. b, 3 min. c, 10 rain.
This result demonstrates that cells are very sensitive to hydratian changes on surfaces of the culture dishes. Effective cell detachment requires an increase in temperature to recover cell metabolism after increasing the hydration of PIPAAm by decreasing temperature. As different cells have different metabolic requirements, the optimum temperature is different for different cell lines.
Mechanism of the cell detachment from thermoresponsive polymer surfaces Figure 7 represents cell adhesion and detachment data on material surfaces. Cells are small particles but are
Figure 5 Effect of sodium azide on hepatocyte detachment by reducing temperature, 30-rain incubation at T~C and an additional 5rain at 25 ~ Hepatocytes on PIPAAm-grafted dishes were pretreated with ( 9 and without (0) 2 m i sodium azide. Hepatocytes on control dishes were pretreated with 2 m u sodium azide (I-l).
distinctly different from small artificial particles in adhesion because cells have metabolism. After cells contact surfaces (passive adhesion), cells are always dynamically altering their cell membrane and its morphology to optimize interactions and to stabilize the cell-material surface interface (active adhesion), both physicochemically and biologically. Therefore, cell adhesion should be divided into two stages: passive adhesion and active adhesion, as shown in
Figure 7. Biomaterials 1995, Vol. 16 No. 4
114
The Biomaterials Silver Jubilee Compendium Cell detachment from thermoresponsive surface: T. Okano et al.
302
Figure 7
Mechanism of the cell attachment to and detachment from material surfaces.
100
different temperature-sensitive cell metabolic efficiencies.
80
A
m m
0
0 t~
60
ACKNOWLEDGEMENTS The authors gratefully acknowledge Dr David Grainger for valuable discussions. This research was supported by the Ministry of Education (grant no. 04453108), Japan.
A
40 20
00
REFERENCES
1
B, C 10 20 30 Temperature (~
2
40
3
Figure 6 Endothelial cell detachment from PIPAAm-grafted
surfaces by reducing temperature and subsequent additional 5-min incubation at 25~ (A) and T~C (B) after 30-min incubation at T~ (T = 4, 10, 15, 20, 27).
When cultured cells in active adhesion seek to detach themselves from a surface, cell shape changes which consume energy are necessary as shown in this paper. W h e n the temperature of culture dishes originally incubated at 37~ is decreased, PIPAAm chains start to hydrate below 32 ~ This remarkable hydration change initiates cell detachment. Further cell d e t a c h m e n t from these temperature-responsive surfaces requires adherent cells to change their m e m b r a n e shape, consuming internal metabolic energy. Lower temperatures provide more hydrated PIPAAm chains but reduce cell metabolism. Therefore, o p t i m u m temperatures are observed to recover cells fully self-detached from temperature-responsive surfaces. As different cells have different temperature sensitivities for cellular metabolism, hepatocytes and endothelial cells require different o p t i m u m temperatures for their detachment. Also, subtle thermal control of cell d e t a c h m e n t is an important basis for advanced technologies, not only for cell culture but for purification, sorting and separation of ceils on the basis of Biomaterials 1995, Vol. 16 No. 4
4
5
6
7 8
9
10 11
Bae YH, Okano T, Kim SW. Temperature dependence of swelling of crosslinked poly(N,N'-alkyl substituted acrylamide) in water. J Polym Sci: Polym Phys 1990; 28: 923-936. Heskins M, Guillent JE, James E. Solution properties of poly(N-isopropylacrylamide). J Macromol Sci Chem 1968; A2: 1441-1455. Takei YG, Aoki T, Sanui K, Ogata N, Okano T, Sakurai Y. Temperature-responsive bioconjugates. 1. Synthesis of temperature-responsive oligomers with reactive end groups and their coupling to biomolecules. Bioconjugate Chem 1993; 4: 42-46. Takei YG, Aoki T, Sanui K, Ogata N, Okano T, Sakurai Y. Temperature-responsive bioconjugates. 2. Molecular design for temperature-modulated bioseparations. Bioconjugate Chem 1993; 4: 341-348. Chen JP, Yang HJ, Hoffman AS. Polymer-protein conjugates I. Effect of protein conjugation on the cloud point of poly(N-isopropylacrylamide). Biomaterials 1990; 11: 625-630. Chen JP, Hoffman AS. Polymer-protein conjugates II. Affinity precipitation separation of human immunogammaglobulin by a poly(N-isopropylacrylamide)protein A conjugate. Biomaterials 1990; 11: 631-634. Okano T, Bae YH, lacobs H, Kim SW. Thermally on-off switching polymers for drug permeation and release. J Control Rel 1990; 11: 255-265. Yoshida R, Sakai K, Okano T, Sakurai Y, Bae YH, Kim SW. Surface-modulated skin layers of thermal responsive hydrogels as on-off switches: I. Drug release. J Biomater Sci Polym Edn 1991; 3:155-162. Yoshida R, Sakai K, Okano T, Sakurai Y. Surfacemodulated skin layers of thermal responsive hydrogels as on-off switches: II. Drug permeation. / Biomater Sci Polym Edn 1991; 3: 243-252. Yoshida R, Sakai K, Okano T, Sakurai Y. Pulsatile drug delivery systems using hydrogels. Adv Drug Delivery Rev 1993; 11: 85-108. Yamada N, Okano T, Sakai H, Karikusa F, Sawasaki Y,
115
The Biomaterials Silver Jubilee Compendium Cell detachment from thermoresponsive surface: T. Okano et al.
12
13
14
15
16
Sakurai Y. Thermo-responsive polymeric surfaces: Control of attachment and detachment of cultured cells. Makromol Chem, Rapid Commun 1990; 11: 571576. Takei YG, Aoki T, Sanui K, Ogata N, Okano T, Sakurai Y. Dynamic contact angle measurement of temperature-responsive surface properties for poly(Nisopropylacrylamide) grafted surface. Macromolecules (in press). Ratner BD, Horbett T, Hoffmen AS. Cell adhesion to polymeric materials; implications with respect to biocompattibility. J Biomed Mater Res 1975; 9: 407422. McAuslan BR, Johnson G. Cell response to biomaterials I: Adhesion and growth of vascular endohelial cells on poly(hydroxyethyl methacrylate) following surface modification by hydrolytic etching. J Biomed Mater Res 1987; 21: 921-935. Okano T, Yamada N, Sakai H, Sakurai Y. A novel recovery system for cultured cells using plasma-treated polystyrene dishes grafted witth poly(N-isopropylacrylamide). J Biomed Mater Res 1993; 27" 1243-1251. Kataoka K, Okano T, Sakurai Y, Maruyama A, Tsuruta
303
17 18 19
20
21 22
T. Controlled interactions of cells with multiphasestructured surfaces of block and graft copolymers. In: Tsuruta T, Nakajima A, eds. Multiphase Biomedical Materials. Tokyo: VSP; 1989: 1-19. Baier RE, DePalma VA, Goupil DW, Cohen E. Human platelet spreading on substrata of known surface chemistry. J Biomed Mater Res 1985; 19: 1157-1167. Seglen PO. Preparation of isolated liver cells. Method Cell Biol 1976; 13: 29-83. West DC, Sattar A, Kumar S. A simplified in situ solubilization procedure for determination of DNA and cell number in tissue culture. Anal Biochem 1985; 147: 289-295. Yonetani T, Ray GS. Studies on cytochrome oxidase VI. Kinetics of the aerobic oxidation of ferrocytochrome c by cytochrome oxidase. J Biol Chem 1965; 240: 33923398. Wilson DF, Chance B. Reversal of azide inhibition by uncoupler. Biochim Biophys Res Commun 1966; 23: 751-756. Palmieni F, Klingenberg M. Inhibition of respiration under tthe control of azide uptake by mitochondria. Eur J Biochem 1967; 1: 439-446.
Biomaterials 1995, Vol. 16 No. 4
116
This Page Intentionally Left Blank
The Biomaterials Silver Jubilee Compendium
117
The Biomaterials Silver Jubilee Compendium
Biomateria]s 17 (1996) 103-114 9 1996 Elsevier Science Limited Printed in Great Britain. All rights reserved 0142-9612/96/$15.00
ELSEVIER
Mechanisms of polymer degradation and erosion Achim G6pferich
Department of Pharmaceutical Technology, University of Erlangen-NOrnberg, CauerstraBe 4, 91058Erlangen, Germany The most important features of the degradation and erosion of degradable polymers in vitro are discussed. Parameters of chemical degradation, which is the scission of the polymer backbone, are described such as the type of polymer bond, pH and copolymer composition. Examples are given how these parameters can be used to control degradation rates. Degradation leads finally to polymer erosion, the loss of material from the polymer bulk. The resulting changes in morphology, pH, oligomer and monomer properties as well as crystallinity are illustrated with selected examples. Finally, a brief survey on approaches to polymer degradation and erosion is given.
Keywords: Polymers, erosion, degradation, modelling, mechanisms Received 29 October 1994; accepted 30 December 1994
At present, tremendous progress is being made in the medical sciences toward the advancement of medical therapies through the application of degradable polymers. Degradable materials are used for the local treatment of cancer 1, the development of vaccines 2'3, the manufacture of nanoparticles with increased plasma half.life 4, 5, self-regulated drug delivery systems 6'7, orthopaedic fixing devices 8 and the fight against organ failure 9. Concomitantly, investigations of these sophisticated applications, however, have raised serious questions about the suitability of degradable polymers in some cases. Examples are the stability of sensitive compounds such as protein and peptide drugs, or the survival of living cells, in the constantly changing chemical environment of an eroding polymer. Other concerns are related to the loss of mechanical stability of polymers during erosion TM, which can be undesirable when occurring too fast, or the toxicity of high concentrations of degradation products. A physical chemical understanding of polymer degradation and erosion processes is the key for a better understanding of these problems and maybe also for their solution. Polymer degradation and erosion play a role for all polymers. The distinction between degradable and non-degradable polymers is, therefore, not clean-cut and is in fact arbitrary, as all polymers degrade. It is the relation between the time-scale of degradation and the time-scale of the application that seems to make the difference between degradable and non-degradable polymers. We usually assign the attribute 'degradable' to materials which degrade during their application, or immediately after it. Non-degradable polymers are those that require a substantially longer time to
degrade than the duration of their application. Degradation and erosion are investigated in many fields of science, such as waste management I1'12 and space science 13. Therefore, many different definitions for degradation and erosion exist in the current literature and sometimes vary markedly from one another 14. The following definitions are adapted for this review. The process of 'degradation' describes the chain scission process during which polymer chains are cleaved to form oligomers and finally to form monomers. 'Erosion' designates the loss of material owing to monomers and oligomers leaving the polymer 15 There are different types of polymer degradation such as photo-, thermal-, mechanical and chemical degradation 16'17. All polymers share the property that they erode markedly under the influence of UV light or 7-radiation TM. For polymer biomaterials, such effects are of minor importance, unless they are submitted to 7-sterilization, after which a significant loss of molecular weight can be observed 19. Thermal degradation plays a greater role for non-degradable polymers 2~ Mechanical degradation affects those biodegradable polymers that are subjected to mechanical stress, such as non-degradable polymers 21 or biodegradable polymers used as fixture or suture material 22. All biodegradable polymers contain hydrolysable bonds. Their most important degradation mechanism is, therefore, chemical degradation via hydrolysis or enzyme-catalysed hydrolysis. The latter effect is often referred to as biodegradation, meaning that the degradation is mediated at least partially by a biological system TM. The processes involved in the erosion of a degradable polymer are complicated. Water enters the polymer bulk, which might be accompanied by swelling. The
Correspondence to Dr A. G6pferich. 103
Biomaterials 1996, Vo]. 17 No. 2
The Biomaterials Silver Jubilee Compendium
104
intrusion of water triggers the chemical polymer degradation, leading to the creation of oligomers and monomers. Progressive degradation changes the microstructure of the bulk through the formation of pores, via which oligomers and monomers are released. Concomitantly, the pH inside pores begins to be controlled by degradation products, which typically have some acid-base functionality. Finally, oligomers and monomers are released, leading to the weight loss of polymer devices. The development of biodegradable polymers during the last two decades has increased exponentially, going hand in hand with new applications for such materials. In early applications, degradable polymers were used as resorbable suture materials 2a. Poly(lactic acidJ and poly(glycolic acid) served as raw materials for such applications. Since the 1970s, these polymers have been used for drug delivery 24 because of their excellent biocompatibility. Soon it was realized, however, that these polymers would not fit the needs of a growing number of applications. Therefore, new polymers were synthesized. Poly(ortho esters) 2'~ poly(anhydrides) 26 and many other polymers emergeci as new materials in the early 1980s. Since then, numerous polymers have been manufactured to keep pace with a steadily increasing demand 27'2~. It is not possible to elucidate the details of erosion for all these polymers in this article. The intention of this review is rather to summarize the most important features of chemical polymer degradation and erosion in vitro and to show how these effects may be described by theoretical simulations.
POLYMER DEGRADATION Polymer degradation is the key process of erosion. There are two principal ways by which polymer bonds can be cleaved: passively by hydrolysis or actively by enzymatic reaction 2~. The latter option is only effectively available for naturally occurring biopolymers like polysaccharides, proteins (gelatin and collagen :~~ and poly(fl-hydroxy acids) 31, where appropriate enzymes are available. A detailed list of enzymatically degradable polymers can be found in Ref. 32. For most biodegradable materials, especially artificial polymers, passive hydrolysis is the most important mode of degradation. There are several factors that influence the velocity of this reaction: the type of chemical bond, pH, copolymer composition and water uptake are the most important. Chemical and physical changes go along with the degradation of biodegradable polymers, like the crystallization of {~ oligomers:: and monomers :~4 or pH changes 5 Some of these factors can have a substantial feedback effect on the degradation velocity. The most important parameter for monitoring degradation is molecular weight. Besides loss of molecular weight, other parameters have been proposed as a measure for degradation, like loss of mechanical strength, complete degradation into monomers or monomer release. All of these are related but need not necessarily obey the same kinetics. For example, complete degradation of poly(L-lactic acid) is known to take B i o m a t e r i a l s 1996, Vol. 17 No. 2
118 Mechanisms of polymer degradation and erosion A. Gopferich
substantially more time than the loss of tensile strength as. Aqueous solutions of lactic acid form spontaneously poly(lactic acid) oligomers that might affect molecular weight measurements :~6, and monomers from copolymers need not be released with identical kinetics during erosion~~. The specific relation between the erosion parameters varies according to the type of polymer. There are, however, basic principles, according to which degradation proceeds, and how degradation can be influenced.
The importance of the type of chemical bond for polymer degradation It is mainly the type of bond within the polymer backbone that determines the rate of hydrolysis :~7. Several classifications for ranking the reactivity exist which are either based on hydrolysis kinetics data for polymers :~':~~ or are extrapolated from low-molecularweight compounds containing the same flmctional group 3z. A brief list is given in Table 1. Anhydrideand ortho-ester bonds are the most reactive ones, followed by esters and amides. Such rankings must be viewed, however, with circumspection. Reactivities can change tremendously upon catalysis 4~J'4j or by altering the chemical neighbourhood of the functional group 4z through steric and electronic effects. The substitution of hydrogen by chlorine in the acid :~position of ethyl acetate, for example, increases the reaction rate constant for hydrolysis in neutral media from 2.5 x 1 0 ~ (s ~) to 1.1x I0 ~ (s ~) through a negative inductive effect 42. The influence of steric effects on degradation can be seen with poly(~.hydroxy esters). The slower degradation of poly(lactic acid) is partially due to the steric effects 4:~. because the voluminous alkyl group hinders the attack of water.
Table 1 Classes of hydrolysable bonds with half-lives according to References 32 and 39 Polymer class o
Half-life poly(anhydrides)
0.1 h
poly(ortho esters)
4h
poly(esters)
3.3 yrs
poly(amides)
83 000 y rs
II
R~C--O~
O~(~C CH:~
i i HH i N--C--
I
R
The Biomaterials Silver Jubilee Compendium
119
Mechanisms of polymer degradation and erosion: A. Gdpferich
The effect of pH on polymer degradation The pH affects reaction rates through catalysis. After shifts in pH, reaction rates of esters, for example, may change some orders of magnitude due to catalysis 42. Ester hydrolysis can, thereby, be either acid or base catalysed 4~ The effect of pH on degradation has been investigated carefully for most biodegradable polymers. For poly(glycolic acid) and poly(lactic-coglycolic acid) sutures, the breaking strength was found to depend markedly on the pH of the degradation medium and was found to be highest at neutral pH 44, reflecting the fastest degradation at low and high pH. This faster chain scission at low pH explains the heterogeneous erosion of poly(lactic acid) due to autocatalysis. The generated monomers, which are carboxylic acids, accelerate polymer degradation by lowering pH 45. For poly(bis-(p-carboxyphenoxy)propane anhydride) cylinders, for example, the degradation rate increases by a factor of 10 when increasing the pH of the degradation medium from 7.4 to 10 (Ref. 46). Poly(ortho esters), in contrast, are resistant against basic pH and degrade substantially faster at acidic compared to neutral pH 47. By using acidic or basic excipients, the degradation rate of polymer hydrolysis can be varied in a controlled way. The internal pH can, thus, effectively be used to influence the degradation rate of polymers 48.
The effect of copolymer composition on polymer degradation By introducing a second monomer into the polymer chain, many properties of the original polymer can be influenced, such as crystallinity or glass transition temperature 49. Such changes have been observed for poly(anhydrides) 5~ where degradation also depends on the copolymer composition. It was shown for poly(1,3-bis-p-carboxyphenoxypropane-co-sebacic acid) (p(CPP-SA)) that degradation depends markedly on the CPP content. Increasing the content of the aromatic monomer from 50 to 100% was reported to increase the time of erosion substantially 51. Other examples are poly(lactic-co-glycolic acid) copolymers, where the decrease of molecular weight during degradation was found to be accelerated with increasing glycolic acid content 52'53. Other factors that depend on the copolymer composition, such as the glass transition temperature and the crystallinity of a polymer, can have additional indirect effects on degradation rates. In general, it can be concluded that the degradation rates of degradable polymers depend on the prevailing type of bond.
The effect of water uptake an degradation Hydrolysis is a bimolecular reaction in which water and the functional group possessing the labile bond are involved. The reaction velocity is determined by the 'concentration' of both reaction partners 54. Lipophilic polymers cannot take up large quantities of water and decrease, thereby, their degradation velocity 46'55. Hydrophilic polymers, in contrast, take up large quantities of water and increase, thereby, degradation rates. The uptake of water is especially important in the area of drug delivery. Hydrogels, for example, may undergo
105
substantial swelling, which for some polymers is the decisive parameter for controlling the release of drugs, and may be more important than polymer degradation.
Influencing polymer degradation In cases where polymers tend to degrade too slowly for a specific application, one might choose to regulate the velocity of chemical degradation. In most cases this is achieved by adding excipients that regulate pH. In drug delivery applications, these can be the drugs themselves that are incorporated into the polymers, such as alkaloids and other bases 56'57 or acids 58. For poly(ortho esters), magnesium hydroxide47 and carboxylic acid anhydrides 59 have been used to modify the degradation of the polymer. The anhydrides accelerate hydrolysis through acid catalysis, whereas magnesium hydroxide decreases degradation rates due to the increased stability of orthoesters in basic media. In the case of poly(e-caprolactone) low-molecular-weight compounds, like ethanol, pentanol, oleic acid, decylamine and tributylamine, were also reported to enhance degradation 6~ Besides adding pH regulating substances, changing the polymer matrix structure has been shown to be a useful tool in controlling degradation rates. There are two principal ways by which this can be achieved: copolymerization and polymer blending. Copolymerization of lactic acid with glycolic acid, for example, increases degradation rates 61. Heller and co-workers have shown that the introduction of acidic and hydrophilic monomers increases water uptake and enhances the autocatalytic degradation 62. Pitt and coworkers observed an increase in degradation rates for soluble blends of poly(vinyl alcohol) and poly(lacticco-glycolic acid) 63. A completely different approach to influencing degradation rates has been proposed by Kost and Langer using ultrasound 64. For poly(anhydrides) a strong dependence of degradation rates on the application of ultrasound w a s f o u n d 65'66, which might be useful for the external regulation of polymer degradation in vivo.
POLYMER EROSION All degradable polymers share the property of eroding upon degradation. Degradation and erosion are the decisive performance parameters of a device made of such materials. To classify degradable polymers a distinction is made between surface (or heterogeneous) and bulk (or homogeneous) eroding materials 67, which is illustrated in Figure 1. During an application, surface eroding polymers lose material from the surface only. They get smaller but keep their original geometric shape. For bulk eroding polymers, degradation and erosion are not confined to the surface of the device. Therefore, the size of a device will remain constant for a considerable portion of time during its application 68. The advantage of surface eroding polymers is the predictability of the erosion process 69. This is desirable when using such polymers for drug delivery, where the release of drugs can be related directly to the rate of polymer erosion 7~ Surface and Biomaterials 1996, Vol. 17 No. 2
The Biomaterials Silver Jubilee Compendium
120 Mechanisms of polymer degradation and erosion" A. G6pferich
106
surface erosion
y
v
/
bulk erosion
,..._ ,...--
Figure I erosion.
pores: macropores with a diameter of approximately 100 ~m which stem from the formation of cracks and micropores with a diameter of approximately 0.1~m that stem from the erosion of polymer bulk. Figure 3 shows that for p(CPP-SA) 20:80 the number of smaller pores increases during erosion, while the number of macropores remains the same. Well defined erosion zones are visible under the light microscope for surface eroding polymers like poly(anhydrides) 77 and poly(ortho esters) 47. Figure 4(a) illustrates an erosion front in a p(CPP-SA) 20:80 disc. Inversely moving erosion fronts have been observed for the autocatalytic degradation of poly(D,L-lactic acid) and poly(D,l.-lacticco-glycolic acid) rods and discs TM that move from the inside of the polymer outward. The preferential erosion of amorphous compared to crystalline polymer parts was observed for enzymatic 79 as well as non-
Schematic illustration of surface erosion and bulk
bulk erosion are ideal cases to which most polymers cannot be unequivocally assigned. Polymer erosion is far more complex than degradation, because it depends on many other processes, such as degradation, swelling, the dissolution and diffusion of oligomers and monomers, and morphological changes. Even more parameters apply to some special types of polymers like electrically erodible materials 71, or during in viva applications 72. Although degradation is the most important process of erosion, depending on the type of polymer, other parameters may also become critical in controlling erosion behaviour. The knowledge of the erosion mechanism is, therefore, most important for the successful application of a degradable polymer. In tissue engineering, surface properties or porosity determine the performance of implantable scaffolds 73. In drug delivery, swelling and porosity are critical to the release behaviour of drugs 68. As with degradation, many different indicators of erosion have been proposed, such as molecular weight loss, sample weight loss and changing geometry. These parameters need not change at the same velocity as the example of poly(anhydrides) illustrates. The molecular weight loss of poly(anhydrides) can be substantial during the first 12 h TM, while there is almost no loss of weight and no change in geometry 34. Erosion is, like degradation, again an individual process for each polymer.
Figure 2 Picture of a p(CPP-SA) 20:80 polymer matrix disc surface after 18.5h in phosphate buffer, pH 7.4, at 37~ taken by scanning confocal microscopy (scale bar m 100/~m). (Reproduced with permission from Ref. 34, ',~(~: 1993, Wiley & Sons.)
Morphological changes during erosion The first morphological changes during erosion are confined to the polymer surface. For poly(anhydrides) the formation of cracks can be observed immediately after contact with buffer 34. Figure 2 shows the surface of a poly(anhydride) after 18.5 h in phosphate-buffered saline, pH 7.4 taken by scanning confocal microscopy, which is covered with cracks. The surface of poly(ortho ester) investigated by atomic force microscopy shows an increasing surface roughness 75. With proceeding erosion, polymers change to more porous structures. Such changes can be detected by mercury intrusion porosimetry TM. The investigation of poly(anhydrides) revealed that there are two types of Biomaterials 1996, Vol. 17 No. 2
Figure 3 Pore size distribution of eroding p(CPP-SA) 20:80 polymer matrix discs determined by mercury intrusion porosimetry (model Poresizer 9 2 2 0 , Micromeritics, Norcross, GA, USA). Penetrometer volume 6ml, size of sample discs: 7mm diameter, 2mm height. Pressure 0.530000psi (3.45-207000 kPa). Pore sizes calculated from Washburn equation for a guessed contact angle of 160'; between mercury and polymer. 9 Day 1; A, day 2; [], day 4.
121
The Biomaterials Silver Jubilee Compendium 107
Mechanisms of polymer degradation and erosion: A. GOpferich ....... . . . . . . .
Thus, anhydrides are cleaved into carboxylic acids, esters and orthoesters into alcohols and carboxylic acids. The degradation products, therefore, influence pH in the degradation medium as well as inside pores. Anhydrides, for example, were shown to affect the pH of the erosion medium substantially. Using pH sensitive dyes in combination with fluorescence scanning confocal microscopy, pH gradients in nonstirred buffered degradation media were detected when approaching the surface of eroding p(CPP-SA) 20:80 discs 34. Figure 5 shows such a profile. It was found that the pH inside the pores of eroding anhydrides is between 4 and 5, which is identical with the pKa of the monomers and far less than the pH of the degradation medium, which was 7.4 (Ref. 34). The findings agree with the results from earlier studies where the pH inside eroding anhydrides was measured using a glass electrode 84. Even more severe deviations of the pH inside the eroding polymer from the pH in the degradation medium were observed for poly(lactic acid) and its copolymers 78, which is due to the higher solubility and the low pKa compared to the poly(anhydride) monomers, pH values as low as 1.8 were measured inside eroding polymer rods 78.
The behaviour of oligomers and monomers during erosion
Figure 4 SEM picture of eroding p(CPP-SA) 20:80 polymer matrix discs, a, Erosion front (middle of the picture) separating eroded (left part) from non-eroded (right part) polymer. b, Eroded spherulite. (Reproduced with permission from Ref. 34, 9 1993, Wiley & Sons.)
enzymatic degradation a~ For poly(3-hydroxybutyrate) this is visible from the appearance of the crystalline spherulitic skeleton of the material a2. The same was observed for poly(anhydrides), where the amorphous parts of p(CPP-SA) 20:80 were less resistant ta erosion than the amorphous ones 34. Figure 4(b) shows the crystalline skeleton of an eroded spheru1Re. The amorphous regions of the spherulite have been eroded, while the crystalline skeleton is still largely in place. These findings were also confirmed for other poly(anhydrides) based on 1,6-bis(p-carboxyphenoxy)hexane, (carboxyphenoxy)methane and 5-(pcarboxyphenoxy]-valeric acid 83.
Changes in pH As already mentioned, the degradation rate depends strongly on pH. Through the chain scission, polymers are transformed into oligomers and monomers, which have different functional groups than the polymer.
During the degradation of polymer chains, oligomers and monomers are created which need not necessarily be released immediately. Lactic acid oligomers, for example, have been reported to form salts that have properties different from those of the protonated compounds aS. Li and Vert observed that poly(D,L-lactic acid) was able to crystallize during the degradation of the polymeric chains 86'a7. They identified the crystals as an oligomeric stereocomplex consisting of poly(Dlactic acid) and poly(L-lactic acid) chains 33, which has been identified and characterized earlier as-92. Monomers created by degradation have also been reported to crystallize during erosion. Differential scanning calorimetry and X-ray diffraction data suggest that the monomers of poly(sebacic acid) and 7.00
J ,
6.75 6.50 6.25 6.00 5.75 5.50 w 9I ' I'--" I ' I -200 -175 -150 -125 -It3() -75
"
]
w"
-50
I
-25
'
l
()
distance from surface [yml Figure 5 pH profile above an eroding p(CPP-SA) 20:80 polymer matrix disc determined by scanning confocal microscopy using fluorescein-5- (and 6)-sulphonic acid as a pH-sensitive fluorescent probe. (Reproduced with permission from Ref. 34, 9 1993, Wiley & Sons.)
Biomaterials 1996, Vol. 17 No. 2
The Biomaterials Silver Jubilee Compendium
122 Mechanisms of polymer degradation and erosion: A. G6pferich
108
p(CPP-SA) polymers crystallize inside eroding polymer matrix d i s c s 34' 93. The release of oligomers and monomers from the polymer bulk has been studied for many polymers. Oligomers have been reported to be released from poly(D,u-lactic acid) microspheres 94 and to increase drug release rates 95. Monomer release profiles for poly(sebacic acid) have a short induction period. The release rate is highest at early times and declines in a concave manner 96. More complex release profiles were obtained for L-lactic acid release from poly(Du-lactic acid). Induction periods are increased to 8 weeks, giving the release profile of L-lactic acid a sigmoidal shape 97, a clear sign of the lower reactivity of the ester bond compared to the anhydride bond. Similar profiles were obtained for the release of lactic and glycolic acids from poly(L-lactic-co-glycolic acid) 75:25 (Ref. 53) and poly(L-lactic-co-glycolic acid) 50:50 (Ref. 15), whereby glycolic acid left the devices approximately twice as fast as lactic acid. Remarkable are the monomer release profiles obtained from poly(anhydride) copolymers 34. The release profiles of sebacic acid are again concave whereas the release profiles of CPP monomer are sigmoidal, as shown in Figure 6. For such differences in the release of individual monomers from copolymers, two mechanisms have been proposed.
degradation was shown to determine the pH inside pores, thereby limiting the solubility of CPP. Whenever sebacic acid has left the device, which according to Figure 6 is after approximately 8 days, the solubility of CPP increases tremendously, visible from the increase in release rate. This example illustrates how intricate the erosion mechanism of biodegradable polymers can be.
1. Erosion controlled release: assuming that the different types of possible bonds in the copolymer backbone are cleaved at different rates, the monomers are set free and, therefore, released at divergent rates. 2. Diffusion controlled release: differences in solubility and diffusivity account for different release rates.
MODELLING OF POLYMER DEGRADATION AND EROSION
For poly(anhydride) copolymers it was proposed that different hydrolysis rates of SA-SA bonds and CPPCPP bonds might lead to different rates at which the monomers are created 15. More important, however, seems to be the solubility of monomers. In the case of poly(anhydrides) this is at any pH approximately 1:10 in favour of sebacic acid 34. Sebacic acid created by t20 10() "~,
80
_~
6o
There are two general sources of crystallinity changes during polymer erosion. One is the generation of crystallized oligomers and monomers. The other stems from the behaviour of partially crystalline polymers during erosion. Due to the faster erosion of amorphous compared to crystalline polymer regions, the overall crystallinity of samples increases, and has been measured for poly(L-lactic acid) 81 and poly(fl-hydroxy butyrate) derived materials 98. Crystallinity also increases during the erosion of intrinsically amorphous polymers like quenched samples of poly(Llactic acid). When introducing these samples to erosion media, their glass transition temperature is lowered due to the uptake of water, which leads to the recrystallization of the polymer 8~
There are many reasons for trying to model polymer degradation and erosion. It would, for example, be very useful if one could predict pH changes on the surface of polymeric scaffolds used in tissue engineering to ensure it is tolerable to attached cells. In drug delivery, proteins and peptides incorporated into polymers might become unstable at extreme pH values, which could be avoided if pH were predictable. The formation of crystallites due to the preferential erosion of amorphous polymer parts might decrease the biocompatibility of implants. All of these problems can only be partially addressed at present because none of the existing models takes into account all of these parameters. In addition, degradation and erosion are often simplified as separate events in modelling schemes, which is not generally the case.
Modelling of polymer degradation J.
4() {}
Changes in crystallinity
2O ()
9
()
,
14
7
9
,
,' ....
r
9
2
time [daysl Figure 6 Release of CPP and sebacic acid (SA) monomer from p(CPP-SA) 20:80. 9 1,3-Bis-p-Carboxyphenoxypropane (CPP); Q, (SA). (Reproduced with permission from Ref. 34, :i" 1993, Wiley & Sons.) B i o m a t e r i a l s 1996, Vol. 17 No. 2
Degradation modelling is not trivial. Major problems arise from investigating the process experimentally, which is necessary to obtain data on which a prospective model can be based. Most degradable polymers are not water soluble and their degradation is influenced by additional factors like swelling, or the kinetics of water uptake. Nevertheless, degradation data were obtained by investigating water soluble oligomers~},.l(,o by using polymer solutions in organic solvent-water mixturesl~ or by investigating degradation in bulk at elevated temperatures ~~ ao4. All these approaches have disadvantages. For example, the degradation of water soluble oligomers can be in equilibrium with the formation of oligomers from monomers in aqueous solutions 36, or the degradation mechanisms
123
The Biomaterials Silver Jubilee Compendium Mechanisms of polymer degradation and erosion: A. G6pferich
109
can be changed by the addition of organic solvents. In most modelling approaches degradation is regarded as a random scission process 1~ assuming first- or secondorder kinetics 1~176To describe the formation of oligomers, random theory has recently been applied in more depth to the description of degradation 1~ and allows the description of the formation of oligomers upon degradation. Little attention, however, has been given to the degradation of copolymers so far.
Modelling of polymer erosion Erosion modelling is even more complex than degradation modelling because of the multitude of involved processes. There are only few appraaches to erosion modelling but none of them covers all processes that are involved in erosion. In early approaches, only heterogeneous erosion was modelled. It was assumed to advance at constant velocity 1~ Similar assumptions were made to investigate spheres and cylinders with concentric bores l~ Next, diffusion theory was introduced to describe the diffusion of low-molecularweight compounds from eroding polymers 1~ Later, moving erosion fronts as well as dissolution fronts for crystalline matter were introduced 11~ A substantial improvement was made when combining the diffusion equation with a reaction term accounting for the degradation of the polymer 112. The degradation of polymer was included into the models under the premise of first-order kinetics for the chain scission 1~3. Recently, the formation and release of oligomers and molecular weight changes were taken into account, also using a diffusion/reaction equation 114'115. All these approaches relied on differential equations for describing erosion. A completely new way of modelling erosion takes advantage of random theory. The erosion of small polymer pieces is regarded as a random event, that cannot be predicted when it will occur, but the likelihood of which is known for any time 116. Similar approaches have been used before for modelling the erosion of controlled release devices 117 and have been developed recently for the optimization of drug release from bioerodible materials 118. The advantage is the inclusion of parameters such as shape, crystallinity, porosity and tortuosity. First, polymer matrices are partially covered using a twodimensional computational grid, as shown in Figure 7(a). The grid divides cross-sections into individual pixels representing crystalline and amorphous polymer areas. Figure 7(b) shows such a grid, on which dark pixels represent crystalline polymer areas and white pixels represent amorphous areas. As poly(anhydrides) are surface eroding, it was assumed that only pixels in contact with the buffer medium can erode. Erosion was assumed to be a Poisson process. The lifetime of a pixel, i.e. the time between the first contact with the erosion medium and the erosion, for such a process is distributed according to a first-order Erlang distribution. Crystalline and amorphous pixels differ by their erosion rate constants, which provide higher likelihood for amorphous pixels to erode. By removing eroded pixels continuously from the grid, time series like the one shown in Figure 8 are obtained 1~6. From such simulations, many experimen-
Figure 7 a, Schematical representation of a polymer matrix cut-out by a computational grid. b, Theoretical representation of a polymer matrix prior to erosion (black pixels, crystalline areas; white pixels, amorphous areas). (Reproduced with permission from Ref. 116, '~ 1993, ACS.)
tally measurable parameters can be calculated, like porosity or weight loss. Figure 9 shows the fit to experimental data for the erosion of p(CPP-SA) 20:80. The fit allows the determination of the erosion rate constants and illustrates that the model is quite well able to adjust to the experimental data. The Monte Carlo model, unfortunately, does not account for the release of incorporated drugs, oligomers or monomers from Biomaterials 1996, Vol. 17 No. 2
The Biomaterials Silver Jubilee Compendium
124 Mechanisms of polymer degradation and erosion: A. G6pferich
110
dependence of solubility on pH (Ref. 120). Such comprehensive models can then be used to describe the complex behaviour of monomer release. Figure 10 shows the fit of the model to the data of Figure 6. As a better quality criterion than the apparently good fit, the model's ability to predict experimental data was tested by predicting functions other than monomer release. Figure 11 shows the prediction of the suspended mass of monomers from the same erosion experiment as in Figure 10. It is apparent that such modelling approaches can be used for predicting data for eroding systems. Despite some progress in the area of modelling, much more data and more sophisticated models are needed to apply these approaches to other degradable polymers. A S P E C T S OF F U T U R E R E S E A R C H The future research in this area will have to focus on experimental aspects of eroding systems. More information on the processes are needed for a better Figure 8 Simulation of polymer erosion using a Monte Carlo model (black pixels, non-eroded areas; white pixels, eroded areas). (Reproduced with permission from Ref. 116, ,~) 1993, ACS.)
1.2 1.0 9.
0.8 0.6 0.4
0.2 0.0 0
7
14
21
time[days] Figure 10 Fit of erosion model to experimental data. 9 1,3-bis-p-Carboxyphenoxypropane (CPP); I , sebacic acid (SA). (Reproduced with permission from Ref. 120, :.~:ii 1994, Elsevier.)
time
60
[daysl
Figure 9 Fit of Monte Carlo model to experimental data. 9 Erosion front position; Q, relative polymer matrix disc mass. (Reproduced with permission from Ref. 116, :~ 1993, ACS.)
50,,.-..,
40 30
eroding polymer matrix discs. For the description of such transport phenomena, diffusion theory has to be applied. Equation 1 describes the one-dimensional diffusion equation in porous media119:
0 0 OC(x,t) O--tC(x, t)e(x, t) - - ~ DeffC(x,t)~:(x, t) Ox
o 6,
20 1() ()
(1)
where C, e, D~ff, t and x are concentration of the diffusant inside pores, porosity, effective diffusivity, time- and space-variables, respectively. The function can be expanded to describe additional phenomena such as the dissolution of suspended drug, or the Biomaterials 1996, Vo|. 17 No. 2
9
()
1
2
3
4
5
6
7
time Idays] Figure 11 Predicted and experimentally measured mass of suspended monomers contained in an eroding p(CPP-SA) 20:80 poly(anhydride) disc. 9 1,3-bis-p-Carboxyphenoxypropane (CPP); Q, sebacic acid (SA). (Reproduced with permission from Ref. 120, (~:~1994, Elsevier.)
The Biomaterials Silver Jubilee C o m p e n d i u m
125
Mechanisms of polymer degradation and erosion: A. G6pferich
understanding of the b e h a v i o u r of d e g r a d a b l e polymers. I n f o r m a t i o n on pH, osmotic pressure, the fate of m o n o m e r s , c h a n g e s in crystallinity as well as m o n o m e r a n d oligomer release and solubility are vital for a better design of devices m a d e of d e g r a d a b l e polymers. Once such information is available, i m p r o v e d m o d e l s can be d e v e l o p e d that m i g h t h e l p to p r e d i c t d e g r a d a t i o n and erosion of p o l y m e r s m o r e accurately. C o n c e p t s a d v a n c e d by such m o d e l s m a y be n e c e s s a r y to fully explore the e n o r m o u s potential of b i o d e g r a d a b l e polymers. REFERENCES
1
2
3 4 5
6 7 8
9 10 11
12 13 14 15 16 17
111
18 19
20 21
22
Brem H, Walter K, Langer R. PoLymers as controlled drug delivery devices for the treatment of malignant brain tumors. Eur ] Pharmacol Biopharmacol 1993;
23
39: 2-7.
24
Alonso MJ, Cohen S, Park, TG, Gupta R, Siber G, Langer R. Determinants of release rate of tetanus vaccine from polyester microspheres. Pharmaceut Res 1993; 10', 945-953. Esparza I, Kissel T. Parameters affecting the immunogenicity of microencapsulated tetanus toxoid. Vaccine 1992; 10' 714-720. Gref R, Minamitake Y, Peracchia MT, Trubetskoy V, Torchilin V, Langer R. Biodegradable long-circulating polymeric nanospheres. Science 1994; 383: 1600-1603. G6pferich A, Gref R, Minamitake J e t al. Drug delivery from bioerodible polymers: systemic and intravenous administration. In: Cleland J, Longer R, eds. Protein Formulations and Delivery (ACS Symposium Series 567). Washington, DC: ACS, 1994; 242-277. Heller J. Chemically self-regulated drug delivery systems. J Controlled Rel 1988; 8: 111-125. Kost J, Langer R, Responsive polymer systems for controlled delivery of therapeutics. Trends Biotechnol 1992; 10: 127-131. Leenslang JW, Pennings AJ, Ruud RM, Rozema FR, Boering G. Resorbable materials of poly(L-lactide). VI. Plates and screws for internal fracture fixation. Biomateria]s 1987; 8: 70-73. Langer R, Vacanti JP. Tissue engineering. Science 1993; 260: 920-926. Engelberg I, Kohn J. Physico-mechanical properties of degradable polymers used in medical applications. Biomaterials 1991; 12: 292-304. Garton A, Henry JL, McLean PD, Stevenson WT. Laboratory simulation of space and salt water environment effects on polymers. Polym Engng Sci 1989; 29: 23-28. Doi Y, Kanesawa Y, Tanahashi N. Biodegradation of microbial polyesters in the marine environment. Polym Degrad Stab 1992; 36:173-177. Golub MA. Reactions of atomic oxygen [O(3P)] with polymer films. Makromol Chem Makromo] Syrup 1992; 53: 379-391. Vert M, Feijen J, Albertson A, Scott G, Chiellini E. Degradable Polymers and Plastics. Melksham: Redwood Press Ltd, 1992: 73-92. Tamada J, Langer R. Erosion mechanism of hydrolytically degradable polymers. Proc Nat Acad Sci USA 1993; 90: 552-556. Banford CH, Tipper CFH. Comprehensive Chemical Kinetics (Vol 14: Degradation of Polymers). New York: Elsevier, 1972. Grassie N, Scott G. Polymer Degradation and Stabilization. New York: Cambridge University Press, 1985.
25
26 27 28 29 30 31 32 33 34
35
36 37 38 39
Sepp/il/i J, Linko Y-Y, Su T, Photo- and biodegradation of high volume thermoplastics. Acta Polytech Scand 1991; 198: 10-12. Spenlehauer G, Vert M, Benoit JP, Boddaert A. In vitro and in rive degradation of poly(D,U lactide/glycolide) type microspheres made by solvent evaporation method. Biomaterials 1989; 10: 557-563. Chatfield DA, Einhorn IN. Stepwise thermal degradation of a polybenzimidazole foam. J Polym Sci Chem 1981; 19: 601-618. Brandwood A, Noble KR, Schindhelm K et al. Influences of stress on biodegradation of novel biomedical polyurethanes. Adv Biomater 1992; 10: 413-420. Miller ND, Williams DF. The in rive and in vitro degradation of PGA suture material as a function of applied strain. Biomaterials 1984; 5: 365-368. Herrmann JB, Kelly RJ, Higgins GA. Polyglycolic acid sutures. Arch Surg 1970; 100: 486-490. Wise DL, Fellmann TD, Sanderson JE, Wentworth RL. Lactic/glycolic acid polymers. In: Gregoriadis G, ed. Drug Carriers in Biology and Medicine. New York: Academic Press, 1979' 237-270. Heller J, Himmelstein KJ. Biodegradable poly(ortho esters) as drug delivery forms. In: Borchardt RT, Arnold J, Stella vJ, eds. Directed Drug Delivery. Clifton, NJ: Humana, 1985: 171-188. Tamada J, Langer R. The development of polyanhydrides for drug delivery applications. J Biomater Sci Polym Edn 1992; 3.4: 315-353. Shalaby SW. Biomedical Polymers --Designed-toDegrade Systems. Munich: Hanser Publishers, 1994. Shalaby SW, Ikada Y, Langer R, Williams J. Polymers of biological and biomedical significance. ACS Symposium Series 1994; 540. Lenz R. Biodegradable polymers. In: Peppas NA, Langer RS, eds. Advances in Polymer Science (No 107: Biopolymers I). Berlin: Springer Verlag, 1993: 1-40. Tomihata K, Burczak K, Shiraki K, Ikada Y. Crosslinking and biodegradation of native and denaturated collagen. Polym Prepr 1992; 33: 534-535. Cox MK. The effect of material parameters on the properties and biodegradation of 'Biopol'. Biodegrad Polym Plastics 1992; 109: 95-105. Park K, Shalaby WSW, Park H. Biodegradable Hydrogels for Drug Delivery. Lancaster: Technomic Publ., 1993. Li S, Vert M. Crystalline oligomeric stereocomplex as an intermediate compound in racemic poly(DL-lactic acid) degradation. Polym Int 1994; 33: 37-41. G6pferich A, Langer R. The influence of microstructure and monomer properties on the erosion mechanism of a class of polyanhydrides. J Polym Sci 1993; 31: 24452458. Laitinen O, T6rm/il~ P, Taurio R, Skutnabb K, Saarelainen K, Iivonen T, Vainionp~/i S. Mechanical properties of biodegradable ligament augmentation device of poly(L-lactide) in vitro and in vivo. Biomaterials 1992; 13: 1012-1016. Holten CH. Lactic Acid, Properties and Chemistry of Lactic Acid and Derivatives. Weinheim: Verlag Chemie, 1971. Fan LT, Singh SK. Controlled Release--A Quantitative Treatment. Berlin: Springer-Verlag, 1989: 105-106. Baker R. Controlled Release of Biologically Active Agents. New York: John Wiley & Sons, 1987: 84-131. St. Pierre T, Chiellini E. Biodegradability of synthetic polymers used for medical and pharmaceutical applications: Part 1--principles of hydrolysis mechanisms. J Bioact Compatible Polym 1986; 1: 467---497. Biomaterials 1996, eel. 17 No. 2
The Biomaterials Silver Jubilee Compendium Mechanisms of polymer degradation and erosion: A. G6pferich
112 40 41 42
43
44 45 46
47 48 49 50
51
52
53
54
55 56 57
58 59
126
Sykes P. A Guidebook to Mechanism in Organic Chemistry (4th edn). London: Longman Group Ltd, 1975: 232-239. Shih C, Higuchi T, Himmelstein KJ. Drug delivery from catalysed erodible polymer matrices of poly(ortho ester)s. Biomaterials 1984; 5: 237-240. Kirby AJ. Hydrolysis and formation of esters of organic acids. In: Bamford CH, Tipper CFH, eds. Comprehensive Chemical Kinetics (Vol 10: Ester Formation and Hydrolysis and Related Reactions). Amsterdam: Elsevier, 1972: 57-202. Pitt CG. Non-microbial degradation of polyesters. In: Vert M, Feijen J, Albertson A, Scott G, Chiellini E, eds. Biodegradable Polymers and Plastics. Melksham: Redwood Press Ltd, 1992: 7-20. Chu CC. A comparison of the effect of pH on the biodegradation of two synthetic absorbable sutures. Arch Surg 1982; 1913: 55-59. Vert M, Li S, Garreau H. More about the degradation of LA/GA-derived matrices in aqueous media. J Controlled Rel 1991; 16: 15-26. Leong KW, Brott BC, Langer R. Bioerodible polyanhydrides as drug-carrier matrices. I: Characterization, degradation and release characteristics. J Biomed Mater Res 1985; 19: 941-955. Heller J. Controlled drug release from poly(orthoesters}. A surface eroding polymer. J Controlled Ret 1985; 2: 167-177. Heller J. Poly(ortho esters). In: Peppas N, Langer R, eds. Advances in Polymer Science (Vol 107). Berlin: Springer Vertag, 1993: 41-93. Hiemenz PC. Polymer Chemistry. New York: Marcel Dekker, 1984: 423-504. Mathiowitz E, Ron E, Mathiowitz G, Amato C, Langer R. Morphological characterization of bioerodible polymers. 1. Crystallinity of polyanhydride copolymers. Macromolecules 1990; 23: 3212-3218. Chasin M, Domb A, Ron E et al. Polyanhydrides as drug delivery systems. In: Chasin M, Langer R, eds. Biodegradable Polymers as Drug Delivery Systems. New York: Marcel Dekker, 1990: 43-71. Miller R, Brady J, Cutright D. Degradation rates of oral resorbable implants (polylactates and polyglycolares): rate modification with changes in PLA/-GA copolymer ratios. J Biomed Mater Res 1977; 11: 711719. Li SM, Garreau H, Vert M. Structure-property relationships in the case of the degradation of massive poly(a-hydroxy acids) in aqueous media. Part 2: degradation of lactide-glycolide copolymers: PLA37.5 GA25 and PLA75GA25. J Mater Sci Mater Med 1990; 1: 131-139. Pitt CG, Chasalow FI, Hibionada YM, Klimas DM, Schindler A. Aliphatic polyesters. I. The degradation of poly(e-caprolactone) in vivo. J Appl Polym Sci 1981; 26" 3779-3787. Ron E, Turek T, Mathiowitz E et al. Controlled release of polypetides from polyanhydrides. Proc Nat Acad Sci USA 1993; 90: 4176-4180. Cha Y, Pitt CG. The acceleration of degradation o controlled drug delivery from polyester microspheres. J Controlled Rel 1989; 8: 259-265. Yoshioka S, Kishida A, Izumikawa S et al. Baseinduced polymer hydrolysis in poly(fl-hydroxybutyrate/fl-hydroxyvalerate) matrices. J Controlled Rel 1991; 16: 341-348. McGinity JW. Influence of drugs and enzymes on the stability of bioerodible polymers. Polym Prepr (Am Chem Soc, Div Polym Chem) 1990; 31: 187-188. Shih C, Higuchi T, Himmelstein KJ. Drug delivery from
Biomaterials 1996, Vol. 17 No. 2
60
61 62
63
64
65
66
67
68 69 70 71 72 73
74 75
76 77 78 79
catalysed erodible polymeric matrices of poly(ortho ester)s. Biomaterials 1984; 5: 237-240. Pitt CG, Gu Z-W. Modification of the rates of chain cleavage of poly(~:-caprolactone) and related polyesters in the solid state. J Controlled ReI 1987" 4: 283-292. Dahlmann I, Rafter G, Fechner K, Mehlis B. Synthesis and properties of biodegradable aliphatic polyesters. Br Polym J 1990; 23: 235-240. Heller J, Penhale DW, Fritzinger BK, Ng SY. The effect of copolymerized 9,10-dihydroxystearic acid on erosion rates of poly(ortho esters) and its use in the delivery of levonorgestrel. J Controlled Rel 1987' 5' 173-177. Pitt CG, Cha Y, Shah SS, Zhu KJ. Blends of PVA and PGLA: control of the permeability and degradability of hydrogels by blending. J Controlled Rel 1992' 19: 189200. Kost J, Leong K, Langer R. Ultrasound-enhanced polymer degradation and release of incorporated substances. Proc Nat Acad Sci USA 1989" 86: 76637666. Liu L-S, D'Emanuele A, Kost J, Langer R. A study on the mechanism of polymer degradation enhanced by low intensity ultrasound. Proc Int Syrup Control Rel Bioact Mater 1991" 18: 225-226. Liu L-S, D'Emanuele A, Kost J, Langer R, Experimental approach to elucidate the mechanism of ultrasoundenhanced polymer erosion and release of incorporated substances. Macromolecules 1992; 25: 123-128. Langer R, Peppas N. Chemical and physical structure of polymers as carriers for controlled release of bioactive agents' a review. J Macromol Sci-Rev Macromol Chem Phys 1983-C23: 61-126. Longer R. New methods of drug delivery. Science 1990' 2 4 9 : 1527-1532. GSpferich A, Langer R. Predicting drug release from cylindric polyanhydride matrix discs. Eur J Pharmacol Biopharmaco] 1995; 41" 81-87. GSpferich A, Langer RS. Modeling of polymer erosion in three dimensions - - Rotationally symmetric devices. AIChE J 1995' 41: 2292-2299. Kwon IC, Bae YH, Kim SW. Electrically erodible polymer gel for controlled release of drugs. Nature 1991' 394" 291-293. Williams DF. Mechanisms of biodegradation of implantable polymers. Clin Mater 1992" 10: .9-12. Vacanti CA, Vacanti JP, Langer R. Tissue engineering using synthetic biodegradable polymers. In' Shalaby SW, Ikada Y, Longer R, Williams J, eds. Polymers of Biological and Biomedica] Significance (ACS Symposium Series 540). Washington, DC: ACS" 1994. D'Emanuele A, Hill j, Tamada ]A, Domb A. Longer R. Molecular weight changes in polymer erosion. Pharmaco] Res 1992; 9: 1279-1283. Shakesheff KM, Davies MC, Domb A et al. Visualizing the degradation of polymer surfaces with an atomic force microscope. Proc Int Symp Control Rel Bioact Mater 1994; 21' 618-619. Mikos AG, Thorsen AJ, Czerwonka LA, Bao Y, Langer R. Preparation and characterization of poly(I.-lactic acid) foams. Polymer 1994; 35: 1067-1078. Mathiowitz E, Ron E, Mathiowitz G et al. Surface morphology of bioerodible polyanhydrides. Polym Prepr 1989" 30:460-461. Herrlinger M. In Vitro Polymerabbau und Wirksttofffeigabe yon Poly-oL-Laktid-Formlingen. PhD thesis, Heidelberg, 1994. Nishida H, Tokiwa Y. Effects of higher-order structure
127
The Biomaterials Silver Jubilee Compendium Mechanisms of polymer degradation and erosion- A. G6pferich
80
81 82
83
84
85 86
87
88 89
90
91
92
93
94 95
of poly(3-hydroxybutyrate) on its biodegradation. II. Effects of crystal structure on microbial degradation. J Environ Polym Degrad 1993; 1" 65-80. Li SM, Garreau H, Vert M. Structure-property relationships in the case of the degradation of massive poly(~hydroxy acids) in aqueous media. Part 3: influence of the morphology of poly(L-lactic acid). ] Mater Sci Mater Med 1990; 1: 198-206. Pistner H, Bendix DR, Mfihlig J, Reuther jF. Poly(Llactide): a long-term degradation study & viva. Biomateriols 1993; 4: 291-298. Doi Y, Kumagai Y, Tanahashi N, Mukai K. Structural effects on biodegradation of microbial and synthetic poly(hydroxyalkanoates). Biodegrad Polym Plastics 1992; 109: 139-148. Mathiowitz E, Jacob J, Pekarek K, Chickering DIII. Morphological characterization of bioerodible polymers. 3. Characterization of the erosion and intact zones in polyanhydrides using scanning electron microscopy. Macromolecules 1993; 28: 6756-6765. Laurencin C. Novel bioerodible polymers for controlled release analyses of in vitro~in viva performance and characterizations of mechanism. PhD thesis. MIT, 1987. Wada R, Hyon S-H, Ikada Y. Salt formation of lactic acid oligomers as matrix for sustained release of drugs, l Pharm PharmacoI 1991; 43: 605-608. Li S M, Vert M. Morphological changes resulting from the hydrolytic degradation of stereocopolymers derived from L- and ~3L-|actides. Macromolecules 1994; 27: 1307-1310. Li SM, Garreau H, Vert M, Therin M, Christel P e t al. In viva degradation mechanism of massive aliphatic polyesters derived from lactic and glycolic acids. In: Vert, M, Feijen J, Albertson A, Scott G, Chiellini E, eds. Biodegradable Polymers and Plastics. Redwood Press Ltd, 1992: 7-20. Ikada Y, Jamshidi K, Tsuji H, Hyon S-H. Slereocomplex formation between enantiomeric poly(lactides). Macromo]ecules 1987; 20: 904-906. Tsuii H, Horii F, Hyon S-H, Ikada Y. Stereocomplex formation between enantiomeric poly(L-lactic acid)s. 2. Stereacomplex formation in concentrated solutions. Macromolecules 1991; 24: 2719-2724. Tsuji H, Hyon S-H, Ikada Y. Stereocomplex formation between enantiomeric poly(L-lactic acid)s. 3. Calorimetric studies on blend films cast from dilute solution. Macromolecules 1991; 24: 5651-5656. Tsuji H, Hyon S-H, Ikada Y. Stereocomplex formation between enantiomeric poly(L-lactic acid)s. Differential scanning calorimetric studies on precipitates from mixed solutions of poly(o-lactic acid) and poly(L-lactic acid). Macromolecules 1991; 24: 56575662. Tsuji H, Hyon S-H, Ikada Y. Stereocomplex formation between enantiomeric poly(IAactic acid)s. 5. Calorimetric and morphological studies on the stereocomplex formed in acetonitrile solution. Macromolecules 1992; 25: 2940-2946. G6prerich A, Langer R. The effect of erosion on the microstructure of biodegradable polymers. Proc Int Syrup Control Rel Bioact Mater 1993; 20: 133134. Park TG. Degradation of poly(O-L-lactic acid) microspheres: effect of molecular weight. ] Controlled Rel 1994; 30: 161-173. Bodmeier R, Oh KH, Chen H. The effect of addition of low molecular weight poly(D,L-lactide} on drug release from biodegradable poly(D,h-lactide) drug delivery systems. Int 1 Pharmacol 1985; 51: 1-8.
113
96 97
98
99
100
101 102 103
104
105
106
107
108
109 110
111
112
113
Tamada J, Longer R. Mechanism of the erosion of polyanhydride drug delivery systems. Proc lnt Syrup Rel Bioact Mater 1990; 17: 156-157. Li SM, Garreau H, Vert M. Structure-property relationships in the case of the degradation of massive aliphatic poly(~-hydroxy acids) in aqueous media, Part 1: poly(BL-lactic acid). ] Mater Sci Mater Med 1990; 1: 123-130. Yasin M, Tighe BJ. Strategies for the design of biodegradable polymer systems: manipulation of polyhydr0• materials. Plastic Rubber Camp Proc Appl 1993; 19: 15-27. Maniar ML, Kalonia DS, Simonelli AP. Use of liquid chromatography and mass spectroscopy to select an oligomer representative of polyester hydrolysis pathways. J Pbarmacol Sci 1989; 78" 858-862. Maniar ML, Kalonia DS, Simonelli AP. Determination of specific rate contents of specific oligomers during polyester hydrolysis. ] Pharmacol Sci 1991; 80" 778782. Domb A, Langer R. Solid-state and solution stability of poly(anhydrides} and poly(esters). Macromolecules 1989; 22: 2117-2122. Zhang X, Wyss UP, Pichora D, Goosen MFA. An investigation of poly(lactic acid) degradation. J Bioact Compat Polym 1994; 9: 80-100. Buchholz B. Accelerated degradation test on resorbable polymers. In: Planck H, Dauner M, Renardy M, eds. Degradation Phenomena on Polymeric Biomaterials. Heidelberg: Springer Verlag, 1992: 67-75. Wallis KH, Miiller RH. Comparative measurement of nanoparticle degradation velocity using an accelerated hydrolysis test. Pharmacol Ind 1993; 55: 168170. Pitt CG, Gratzel MM, Kimmel JL et a]. Aliphatic polyesters. II. The degradation of poly(D-L-lactide), poly(e-caprolactone), and their copolymers in viva. Biomaterials 1981; 2: 215-220. Granger G. Stochastic modeling for oligomers produced by degradation of linear polymers. In: Vert M e t al. eds. Degradable Polymers and Plastics. Melksham: Redwood Press Ltd, 1992:191-199. Hopfenberg HB. Controlled release from bioerodible slabs, cylinders and spheres. In: Paul DR, Harris FW, eds. Control]ed Release Polymeric Formulations (ACS Symp Set No 33). Washington, DC: ACS, 1976: 26-32. Cooney DO. Effect of geometry on the dissolution of pharmaceutical tablets and other solids: surface detachment kinetics controlling. AIChE J 1972; 18: 446-449. Baker RW, Lonsdale HK. Erodible controlled release system. Am Chem Sac Div Org Coat Plast Chem Prepr 1976; 3: 229. Lee PI. Diffusional release of a solute from a polymeric matrix--approximate analytical solutions. ] Membr Sci 1980; 7: 255-275. Thombre AG, Himmelstein KJ. Modeling of drug release kinetics from a laminated device having an erodible drug reservoir. Biomaterials 1984; 5: 250254. Thombre AG. Theoretical aspects of polymer biodegradation: mathematical modeling of drug release and acid-catalyzed poly(ortho-ester) biodegradation. Biodegrod Polym Plastics 1992; 109: 214-225. Heller J, Baker RW. Theory and practice of controlled drug delivery from bioerodible polymers. In: Baker RW, ed. Controlled Release of Bioactive Materials. New York: Academic Press, 1980: 1. Biomaterials 1996, Vol. 17 No. 2
The Biomaterials Silver Jubilee Compendium Mechanisms of polymer degradation and erosion: A. G6pferich
114
114 115 116 117
128
Himmelstein KJ, George J. Erosion of polymer with autocatalysis. Proc Int Syrup Control Rel Bioact Mater 1993; 20: 53-54. Himmelstein KJ, Heterogeneous mechanisms of polymer erosion. Polym Prepr (Am Chem Soc, Div Polym Chem 1992; 33: 48-49. G6pferich A, Langer R, Modeling polymer erosion. Macromolecules 1993; 16: 4105-4112. Zygourakis K. Development and temporal evolution of erosion fronts in bioerodible controlled release
Biomaterials 1996, Vol. 17 No. 2
118 119 120
devices. Chem Engng Sci 1990; 45: 2359-2366. Zygourakis K. Design and optimization of bioerodible devices with optimal release characteristics. Mater Res Soc Proc 1993; 331: 79-84. Saltzmann WM, Langer R. Transport rates of proteins in porous materials with known microgeometry. Biophys J 1989; 55: 163-171. G6pferich A, Langer R. Modeling monomer release from bioerodible polymers. ] Controlled Rel 1995: 33: 55-69.
129
The Biomaterials Silver Jubilee Compendium
Biomaterials 17 (1996) 115-124 (~) 1996 Elsevier Science Limited Printed in Great Britain. All rights reserved 0142-9612/96/$15.00
ELSEVIER
Stabilized polyglycolic acid fibrebased tubes for tissue engineering D.J. Mooney *t* C L. Mazzoni* C. Breuer* K. McNamara* D. Hern*, J.P. Vacanti* and R. Langer* 9
"
~
9
9
*Department of Chemical Engineering, Massachusetts Institute of Technology, Cambridge, MA 02139, USA; tDepartment of Surgery, Harvard Medical School and Children's Hospital, Boston, MA 02115, USA
Polyglycolic acid (PGA) fibre meshes are attractive candidates to transplant cells, but they are incapable of resisting significant compressional forces. To stabilize PGA meshes, atomized solutions of poly(L-lactic acid) (PLLA) and a 50/50 copolymer of poly(D,L-lactic-co-glycolic acid) (PLGA) dissolved in chloroform were sprayed over meshes formed into hollow tubes. The PLLA and PLGA coated the PGA fibres and physically bonded adjacent fibres. The pattern and extent of bonding was controlled by the concentration of polymer in the atomized solution and the total mass of polymer sprayed on the device. The compression resistance of devices increased with the extent of bonding, and PLLA bonded tubes resisted larger compressive forces than PLGA bonded tubes. Tubes bonded with PLLA degraded more slowly than devices bonded with PLGA. Implantation of PLLA bonded tubes into rats revealed that the devices maintained their structure during fibrovascular tissue ingrowth, resulting in the formation of a tubular structure with a central lumen. The potential of these devices to engineer specific tissues was exhibited by the finding that smooth muscle cells and endothelial cells seeded onto devices in vitro formed a tubular tissue with appropriate cell distribution.
Keywords: Tissue engineering, polyglycolic acid, polylactic acid, smooth muscle cells, endothelial cells Received 26 October 1994; accepted 5 January 1995
to engineer a variety of tissues, including liver, cartilage and intestine 3. This class of polymers degrades by a simple hydrolysis mechanism, and by varying the ratio of lactic and glycolic acids in the polymer one can control the crystallinity of the polymer, and thus its degradation rate and mechanical properties 4. Furthermore, these polymers can be processed to yield a variety of different structures, including fibres, hollow tubes and porous sponges 5-7. Non-woven meshes of polyglycolic acid (PGA) fibres have been particularly attractive materials for use as cell delivery devices as they are highly porous, permitting diffusion of nutrients throughout the device following implantation while allowing subsequent neovascularization of the developing tissue, and they can be easily fabricated into devices with varying geometry. However, this material lacks the structural stability to withstand compressive forces in vivo, and external supports are necessary if one desires to form a stable three-dimensional structure (e.g. a tube) from this material 8' 9. In this study, we investigated whether threedimensional structures capable of resisting large compressive forces and guiding the formation of a desired tissue structure could be formed from PGA fibre meshes by physically bonding adjacent fibres using a spray casting method. Poly(L-lactic acid)
While organ transplantation and tissue reconstruction are highly successful therapies for a variety of maladies, a shortage of donor tissue limits their application to a percentage of those who could potentially benefit from these therapies. For example, over 83 000 people either died or were maintained on less-thanoptimal therapies due to a lack of donated organs in the USA in 19901. To aid these people, a variety of investigators have proposed to engineer new tissues by transplanting isolated cell populations on biomaterial scaffolds to create functional new tissues in vivo 2. To engineer complex tissues such as blood vessels or intestine, cells must be localized to a specific site in vivo, and the formation of an appropriate tissue structure from the implanted cells and the host tissue must be promoted. Biodegradable materials are particularly attractive for fabricating the devices utilized ta transplant cells and engineer new tissues because they can be designed to erode after tissue development is complete, leaving a completely natural tissue 2'3. Templates synthesized from polymers of the lactic and glycolic acid family have previously been utilized tCurrent address: Departments of Biological and Materials Sciences and Chemical Engineering, University of Michigan, Ann Arbor, MI 48109, USA. Correspondence to Prof. R. Longer. 115
Biomaterials 1996, Vol. 17 No. 2
130
The Biomaterials Silver Jubilee Compendium 116
(PLLA) or a 50/50 copolymer of lactic and glycolic acids (PLGA) was dissolved in chloroform, atomized and sprayed over a PGA mesh formed into a tubular structure. Following solvent evaporation, a physically bonded structure resulted, and the pattern and extent of PGA fibre bonding was controlled by the processing conditions. These tubular devices were capable of withstanding large compressive forces in vitro (50200mN) and maintained their structure in vivo. The specific mechanical stability was dictated by the extent of physical bonding and the polymer utilized to bond the PGA fibres.
MATERIALS The PGA mesh (fibre diameter approximately 121~m; mesh thickness=0.3 ram, specific gravity= 80.2mgcm-3, porosity= 97%) was purchased from Albany Int. (Taunton, MA, USA), PLLA and the poly(D,L-lactic-co-glycolic acid) from Medisorb (Cincinnati, OH, USA), the lactic dehydrogenase kit, glycolic and lactic acid standards, and 4,5-dihydroxy2,7-naphthalenedisodium salt were purchased from Sigma Chemical Co. (St Louis, MO, USA), chloroform from Mallinckrodt (Paris, KY, USA), phosphatebuffered saline and DMEM medium from Gibco (Grand Island, NY, USA), Tmax film from Kodak, Lewis rats (250-300g) from Charles River (Wilmington, MA, USA), calf serum from Hyclone Lab. Inc. (Logan, UT, USA), penicillin and strepromycin from Irvine Scientific (Santa Ana, CA, USA), and methoxyflurane from Pitman-Moore Inc. (Mundelein, IL, USA).
METHODS Tube fabrication Rectangles (1.3 x 3.0cm) of the non-woven mesh of PGA fibres were wrapped around a Teflon cylinder (outside d i a m e t e r - 3.0 mm) to form a tube, and the two overlapping ends were manually interlocked to form a seam. The Teflon cylinders were then rotated at 20rpm using a stirrer (Caframo; Wiarton, Ontario, Canada). Solutions of PLLA and PLGA dissolved in chloroform (1-15%, w/v) were placed in a dental atomizer (Devilbus Corp.) and sprayed over the rotating PGA mesh from a distance of 6in (.~15cm) using a nitrogen stream (18 psi (.~.124.2kPa)) to atomize the polymer solution. The PLGA and PLLA had molecular weights (Mw) of 43 400 (M,,/M,, = 1.43) and 74100 (M,./M~ = 1.64), respectively. Molecular weights were determined by gel permeation chromatography as described previously 7. While PLLA and copolymers of lactic and glycolic acids are soluble in chloroform, PGA is very weakly soluble in this solvent. Thus, the PGA fibres are largely unchanged by the process. After spraying was completed, the tubes were lyophilized to remove residual solvent, removed from the Teflon cylinder and cut into specific lengths. The tubes were sterilized by exposure to ethylene oxide for 24h, followed by degassing for 24 h. Biomaterials 1996, Vol. 17 No. 2
Stabilized PGA tubes D.J, Mooney et al.
Device characterization The mass of PLLA and PLGA that bonded to the PGA scaffolds was determined by weighing PGA devices before and after spraying. For scanning electron microscopic examination, samples were gold coated using a Sputter Coater (Desk II, Denton Vacuum, Cherry Hill, NJ, USA). An environmental scanning electron microscope (Electro Scan , Wilmington, MA, USA) was operated at 30 kV with a water vapour environment of 5 torr (~665 Pa) to image samples. Photomicrographs were taken with Polaroid 55 film. Thermal mechanical analysis was performed with a TMA 7 (Perkin Elmer Corp, Norwalk, CT, USA) using a compression probe with a circular tip (d -- 3.0 mm). All testing was done at a constant temperature of 37"C. Tubes were placed on their sides for testing (axis of tube lumen perpendicular to the axis of force application), and the change in device diameter (parallel to the direction of force application) was followed during and after force application. The compressional forces applied to the tubes in vivo will presumably also be in a radial direction. The resulting deformations were normalized to the initial device diameter. Some samples were pre-wet by placing them in a vial containing phosphate-buffered saline and incubating at 37"C for 24 h. All tests were performed in triplicate, and representative data are given. The erosion characteristics of bonded devices were assayed by' placing individua| tubes in 5 ml of phosphate-buffered saline, pH 7.4, and incubating under static conditions at 37'C. The mass loss was analysed by weighing lyophilized devices before and after the incubation period. The release of lactic acid was assayed enzymatically with lactic dehydrogenase using a kit from Sigma. The release of glycolic acid was quantitiated with a colorimetric assay ~~ which involves decarboxylating glycolic acid in the presence of concentrated sulphuric acid to form formaldehyde, followed by reaction of formaldehyde with chromotropic acid to yield a coloured product which can be quantitated spectrophotometrically.
Implantation of tubes Polymer constructs were implanted into the omentum of syngeneic Lewis rats as described previously ~. NIH guidelines for the care and use of laboratory animals (NIH Publication No. 85-23 Rev. 1985) have been observed in all experiments involving animals. Inhalation anaesthesia with methoxyflurane was always utilized. The omental tissue was rolled around the devices to promote tissue invasion and neovascularizalion of the implants from all sides. Implants were secured in place with sutures of 7-0 Maxon (Davis and Geck). Recipients of polymer devices were killed on post-implantation days 3 and 18. The implants were removed, fixed in 10% buffered formalin and thin sections were cut from paraffin-embedded tissue. Histological sections were stained with haematoxylin and eosin. Photomicrographs were taken with Kodak Tmax fihn.
Cell seeding on devices To introduce bovine aortic smooth muscle cells (passage 6-9) into the polymeric: delivery devices,
The Biomaterials Silver Jubilee Compendium
131
Stabilized PGA tubes: D.J. Mooney et al.
l ml of a cell suspension containing 5 - 2 0 x 1 0 5 cellsm1-1 was injected into the interior of each tube using a l m l syringe and a 22-gauge needle. The cell suspension was retained in the tubes by placing a small plug of the PGA fibres at both ends of the tubes during the cell adhesion period. Devices were incubated at 37~ in an atmosphere of 10% CO2 to allow for cell adhesion and proliferation. The tubes were manually rotated periodically using sterile forceps during the period of cell adhesion to promote even cell seeding. Cell-polymer devices were kept in DMEM medium, containing 5% calf serum, 100Um1-1 penicillin and 100mgm1-1 streptomycin, during this time. The seeding protocol was repeated 7 days later to ensure even seeding of cells within the devices. Ten days later, a cell suspension of bovine aortic endothelial cells (passage 6-9) was similarly seeded onto the tubes. After 4 more days the devices were fixed in formalin, embedded in paraffin, sectioned and stained (haematoxylin and eosin) using standard techniques. Sections were stained for the presence of desmin (a smooth muscle specific protein) and Factor 8 (specific for endothelial cells) using standard immunohistochemical protocols. Antibodies for this analysis were purchased from Shandon (Pittsburgh, PA, USA). The endothelial cells and smooth muscle cells were isolated from bovine aortas using a collagenase digestion, and were a gift from Dr Judah Folkman.
RESULTS Bonding tubes with PLLA To determine whether PGA scaffolds could be stabilized by physically bonding adjacent fibres, chloroform containing dissolved PLLA (1-15% w/v) was sprayed over the exterior surface after the PGA mesh was wrapped around a Teflon cylinder to form a tube. The PLLA formed a coating over the exterior PGA fibres after the solvent evaporated, and physically bonded adjacent fibres. The tubes formed in this manner could be easily removed from the Teflon cylinder for characterization and use, The pattern of bonding was controlled by the concentration of the PLLA in the atomized solution (Figure I), even though the time of spraying was adjusted to maintain an approximately constant mass of PLLA on the devices under the various conditions (Table 1). Spraying with a solution containing 1 or 5% PLLA resulted in extensive bonding of PGA fibres without significantly blocking the pores of the PGA mesh. Spraying with a 10% solution of PLLA also bonded fibres, but resulted in the formation of a PLLA film on the exterior surface of the PGA mesh that contained only small pores. Spraying with a solution containing 15% PLLA had a similar effect, although the polymer film that formed was less organized. In all cases, the PLLA coated and bonded fibres only on the exterior surface of the PGA mesh, as no coating or bonding of fibres was observed on the interior surface of the PGA mesh (Figure 2). The compression resistance of bonded tubes was assessed in vitro to determine which patterns of
117
bonding resulted in the most stable devices. Unbonded tubes were completely crushed by a force of 5 mN, but banded tubes were capable of resisting forces in excess of 200raN. However, the ability of bonded tubes to resist a given compressional force was dependent on the pattern of bonding (Figure 3). For example, tubes bonded with 1 or 15% PLLA were significantly compressed by a force of 200mN, while tubes bonded with a solution of 5 or 10% PLLA were only slightly compressed by this force. The compression was viscoelastic in all cases, as the devices only partially decompressed after the force was removed. Uniform properties were observed with respect to the position along and around a tube. To determine if the extent, as well as the pattern, of bonding could vary the compression resistance of tubes, an atomized dispersion of 5% PLLA was then sprayed over the devices for different times. Lengthening the spraying time from 10 to 60s increased the mass of PLLA on the devices (Table 2). Infrequent bonds between adjacent fibres resulted from spraying for 10 s. Spraying for more extended periods increased the PLLA coating over the PGA fibres, and the extent of bonding (Figure 4). The ability of these tubes to resist compressional forces and maintain their shape was quantitated again using thermal mechanical analysis. The compression resistance strongly depended on the extent of bonding, as tubes that were more extensively bonded had a greater resistance to deformation (Figure 5A). The compression that did occur under these conditions was again a combination of a reversible, elastic strain and an irreversible deformation. Some tubes were also exposed to an aqueous environment before testing to determine whether this environment for 24h would destabilize the tubes. The aqueous environment had a slight detrimental effect on the stability of bonded tubes, but they were still capable of resisting large compressive forces (Figure 5B).
Bonding tubes with PLGA To determine whether this technique of stabilizing PGA devices could be utilized with a variety of polymers, the previous study was repeated using a 50/50 copolymer of lactic and glycolic acids. The mass of polymer bonded to the devices and the extent of physical bonding were again regulated by the time an atomized dispersion of the bonding polymer was sprayed over the PGA fibres (Table 2; Figure 6). Once again, bonding increased the compression resistance of devices formed into a tubular structure (Figure 7A). However, these devices were not able to resist the same compressional forces as PLLA bonded devices. Tubes bonded with PLLA were capable of resisting forces up to 200mN, while tubes bonded with PLGA were only capable of resisting forces slightly greater than 50 rnN. The difference between devices stabilized with PLLA and PLGA was even more striking when the devices were tested after immersion in phosphate-buffered saline for 24h. PLGA bonded tubes, in contrast to PLLA bonded tubes, were significantly weakened by this treatment
(Figure 7t3).
Biomaterials 1996, Vol. 17 No. 2
132
The Biomaterials Silver Jubilee Compendium Stabilized PGA tubes" D.J. Mooney et ai.
118
Figure 1
P h o t o m i c r o g r a p h s of the exterior surface of PGA m e s h e s formed into tubular structures and sprayed with solutions
containing (A) 1%, (B) 5%, (C) 10% and (D) 15% PLLA. The spraying time was varied to yield an approximately constant mass of sprayed PLLA in all conditions. The original magnifications and size bars are shown in the photomicrographs.
Table 1 PGA mesh sprayed with solutions of varying PLLA concentration .
.
.
.
-
PLLA concentration (w/v)
Spraying time (s)
Mass of PLLA on device* (% initial PGA mass)
1 5 10 15
150 30 15 10
115+20 168 • 16 145 _L. 12 108 4- 73
*VaLues represent the mean 4-s.d. of three devices,
Tube degradation in vitro The time course for erosion of the tubes was determined by quantitating the mass loss and monomer release from tubes immersed in a pH balanced, isotonic saline solution. Devices bonded with PLGA were completely degraded by 11 weeks, while devices bonded with PLLA only lost 30% of their mass after 10 weeks (Figure 8A). The degradation of the PLLA bonded tubes was solely due to erosion of the PGA fibres, as glycolic acid was released from the Biomaterials 1996, Vol. 17 No. 2
tubes, but virtually no lactic acid was released over this time from the tubes (Figure 8B). PLLA degrades slowly, and no significant loss of PLLA mass is expected until 1-2 years. Erosion of tubes bonded with PLGA was due to erosion of both the PLGA fibres and the PLGA, as both glycolic acid and lactic acid were released from the tubes over this time flame (Figure
8c).
Compression resistance in vivo To confirm that stabilized tubes were capable of resisting compressional forces in vivo as well as in vitro, devices bonded with PLLA (5% PLLA; 30s spraying time) were implanted into the omentum of laboratory rats. The initial (3 day) host response was characterized by fibrin deposition and scattered inflammatory cells throughout the devices. A mature fibrovascular tissue was evident throughout the devices by 7 days, and the devices maintained their tubular structure with a central lumen for the 18 day duration of the experiment (Figure 9A). The invading fibroblasts and the newly deposited matrix were aligned with the lumens of the tubes (Figure 9B).
133
The Biomaterials Silver Jubilee Compendium Stabilized PGA tubes: D.J. Mooney et al.
119
Cell adhesion and organization in vitro on bonded tubes
Figure 2 A photomicrograph of the interior surface of PGA mesh formed into a tubular structure and sprayed with a solution of 5% PLLA for 30s. The interior surface, in contrast to the exterior surface (see Figure 1), was largely unaffected by this process. The original magnification and size bar are shown in the photomicrograph.
PLLA bonded tubes (5% PLLA; 30s spraying time) were subsequently seeded with smooth muscle cells and endothelial cells to investigate the suitability of these devices to serve as cell delivery vehicles. Blood vessels are largely comprised of these two cell types. The smooth muscle cells adhered to the polymer fibres (Figure I OA and B), and proliferated to fill the void space present between polymer fibres (Figure lOB). Endothelial cells also adhered to the devices, and over time formed a lining on the interior section of the devices (Figure I OA and C). Immunohistochemical staining for desmin confirmed that the cells filling the interstices between polymer fibres were smooth muscle cells, and staining for Factor 8 confirmed that the cells lining the luminal surface were endothelial in nature (not shown). This organization of the muscle and endothelial cells is similar to that observed in blood vessels.
DISCUSSION
Figure 3 Representative strain diagrams of tubes formed from the PGA mesh after spraying with a solution containing r-I, 1%; B, 5%; O, 10%; and O, 15% PLLA. Devices were subjected to a compressive force of 200 mN applied in a direction perpendicular to the axis of the device lumen starting at 0min. The force was removed at 10 min, and the change in the diameter of the tube (parallel to the direction of force application) was monitored both during and after the time of force application, and normalized to the initial diameter.
Table 2 Spraying PGA scaffolds for various times with a 5% solution of PLLA or PLGA Spraying time (s)
Mass of PLLA on device* (% initial PGA mass)
Mass of PLGA on device* (% initial PGA mass)
10 20 30 60
43:t: 160 • 165 + 390 +
54+9 59 + 40 140 + 10 313 • 51
11 55 22 37
*Values represent the mean • s.d. of three devices.
Three-dimensional tubes can be formed from PGA fibre scaffolds by physically bonding adjacent fibres. The compression resistance and degradation rate of these devices were controlled by the pattern and extent of physical bonding, and the type of polymer utilized to bond the PGA fibres. Fibrovascular tissue invaded the devices following implantation, leading to the formation of a tubular tissue with a central lumen. The potential of these devices to engineer tissues was exhibited by the finding that endothelial cells and smooth muscle cells adhered to the devices and formed a new tissue in vitro with appropriate tissue organization. The compression resistance of devices was monitored by applying a constant force on the tubes. The resulting changes in the device diameters were partially elastic, as indicated by the partial decompression following removal of the applied force. The irreversible changes in the device diameters were likely caused by both crushing and bending of fibres, and by rearrangement of fibres. Contact between the compression tip and the tubes was not analysed, and will likely change as the tubes compress and fibres rearrange. For this reason, results were reported for compressional forces, not stresses. Calculation of stresses using the entire contact area of the compression probe would give the most conservative estimate of mechanical moduli. Tubes which were bonded with PLLA were more resistant to compressional forces than tubes bonded with PLGA. This finding is not surprising, as crystalline PLLA is typically much stiffer than amorphous PLGA 4. Additionally, while the compression resistance of PLLA bonded devices was not greatly changed after exposure to an aqueous environment, PLGA bonded devices were markedly weakened after the same treatment. PLGA is more hydrophilic than PLLA 4 due to the presence of the glycolic acid residues, and the absorbed water likely acts as a plasticizer, weakening Biomaterials 1996, Vol. 17 No. 2
134
The Biomaterials Silver Jubilee Compendium 120
Stabilized PGA tubes: D.J. Mooney et ai.
Figure 4 Photomicrographs of the exterior surface of PGA meshes formed into tubular structures and sprayed with solutions containing 5% PLLA for (A) 10, (B) 20, (C) 30 and (D) 60 s. The original magnifications and size bars are shown in the photomicrographs,
the PLGA. The PLLA bonded devices were slightly weakened after this treatment, indicating that the PLLA was also somewhat plasticized. The erosion of the devices was also dependent on the polymer utilized for bonding. PLLA is hydrolysed very slowly, and virtually no lactic acid release was observed over the 10 weeks of the erosion study. The erosion of PLLA bonded devices was entirely due to hydrolysis of the glycolic acid bonds in the fibres. In contrast, both the PGA fibres and the PLGA used to bond the fibres eroded completely over 11 weeks. The release of glycolic acid from these devices occurred more rapidly than the release of lactic acid. This was likely caused by the more rapid erosion of the PGA fibres, followed by the slower release of both lactic acid and glycolic acid from the PLGA. Biodegradable devices are attractive for cell transplantation and tissue engineering since they can be designed to erode once tissue development is complete, leaving a completely natural tissue. The approach described in this report to mechanically stabilize fibre-based scaffolds was performed with PGA, PLGA and PLLA because of the long history of these polymers in medical devices, and the range of degradation rates that can be obtained with this class Biomaterials 1996, Vol. ] 7 No. 2
of polymers (Figure 8). However, this technique could potentially be used with a variety of other polymers, both erodible and non-erodible, for medical or nonmedical applications. Various approaches have previously been taken to mechanically stabilize structures formed from PGA fibres. PGA fibres can be physically bonded with a second polymer in a similar manner as described here by simply dipping the PGA scaffold into a solution of PLLA dissolved in chloroform, and allowing the chloroform to evaporate ~. Alternatively, a thermal processing technique that results in temporary melting and subsequent bonding of PGA fibres has been reported ~2. The bonding approach described in this report is simple, permits a variety of bonding polymers to be utilized and allows the fabrication of various threedimensional scaffolds. It also results in bonding only of the outermost fibres of the device (Figure 2), in contrast to the other methods. This preserves the desirable features of the PGA mesh (high porosity, high surface area/polymer mass ratio) throughout the interior sections. This approach also allows both the extent and pattern of bonding to be easily controlled. Extensive coating and bonding of fibres resulted when the polymer concentration in the atomized solution
The Biomaterials Silver Jubilee Compendium
135
Stabilized PGA tubes: D.J. Mooney et al.
A
100 80
k,,
q)
e9~
|
u >
sec
30 20
sec sec
10
sec
B
100 ~
~
Control
80
!._
(D
60
E m N9
40
==,,
Pre-wet
60
E
E I,.
40
>
o.,.
.,m
C3
60
A
E q) N r .-a
121
20 a
0
0
20
--.
Time
~.
t0
..
. ....
(minutes)
i
20
0
0
....
1
i
,
5
~0
~s
Time
--"1
z0
(minutes)
Figure 5 Representative strain diagrams of (A) PGA tubes sprayed for various times with a 5% PLLA solution and subjected to a compressive force of 200 mN starting at 0 rain. The force was removed at 10min. The force application and the change in the diameter of the tube (normalized to the initial diameter) were monitored, as described in the legend for Figure 3, both during and after the time of force application. (B) Devices sprayed with a 5% PLLA solution for 30s and tested dry (Control) or after pre-wetting for 24 h in a saline solution (Pre-wet). The compressional force was again 200 mN.
Figure 6 Photomicrographs of the exterior surface of PGA meshes formed into tubular structures and sprayed with solutions containing 5% PLGA for (A) 10, (B) 20, (C) 30 and (D) 60s. The original magnifications and size bars are shown in the photomicrographs. Biomaterials 1996, Vol. 17 No. 2
136
The Biomaterials Silver Jubilee Compendium Stabilized PGA tubes: D.J. Mooney et al.
122 100
"-"
|
.u_>,.-c
80
60
sec
30
sec
10
100 b ~,
,,...o._r
Control
80
sec
60
3 ~
60
40
u ... c ";
|
g,
40
Pre-wet
g, 20
200 -
0
u
1'0
Time
2'0
(minutes)
0
1'0
TIME
(minutes)
20
Figure 7
Representative strain diagrams of (A) PGA tubes sprayed for various times with a 5% PLGA solution and subjected to a compressive force of 50 mN starting at 0 min. The force was removed at 10 minutes, and the change in the diameter of the tube (normalized to the initial diameter) was monitored both during and after the time of force application. (B) Devices sprayed with a 5% PLGA solution for 30s and tested dry (Control) or after pre-wetting for 24h in a saline solution (Pre-wet), The compressional force in this test was again 50 mN.
A
B 100
W W cu
Q
6
80
W 0
PLLA
80
(I}
m
60 o
E~
PLGA
'-"
0
e~
40 "" 20
0
20
4
6
10
8
cld
m
q)
"gae
a
100
0
12
/ 0
.ot,o .c,, 2
4
Time
(wk)
Time
6
8
10
12
(wk)
C (I)
w r (I) ~
Se
IX: =
_s
t.
9 ii
E E
o~ C ,-0 5
100
Olycollc
80
acid
60 4o
Lactic
acid
2O
0
2
4
Tlme
6
8
10
12
(wk)
The degradation of devices bonded by spraying with PLLA or PLGA (5% solution; spraying time .... 30s), as measured by (A) quantitating the change in device mass over time, or (B) the release of glycolic and lactic acids from PLLA bonded devices, or (C) PLGA bonded devices. Devices were incubated at 37"C under static conditions in buffered saline and removed at various times for analysis, Values in (A) represent the mean and standard deviation calculated from three samples. Figure 8
Biomaterials 1996, Vol. 17 No. 2
137
The Biomaterials Silver Jubilee Compendium Stabilized PGA tubes: D.J. Mooney et al.
123
Figure 9 (A) Low-power and (B) high-power photomicrographs of a histological section from a bonded tube (5% PLLA; 30s) implanted for 17 days in the omentum of a Lewis rat. These cross-sections of the implanted device were cut perpendicular to the axis of the tube's lumen. (A) The central lumen (I) is visible, along with numerous polymer fibres (arrows), the host omental tissue (o), and the ingrown fibroblasts and fibrous tissue they deposited. (B) The fibroblasts which invaded the device and the fibrous tissue deposited by these cells aligned in parallel with the central lumen. The original magnifications of these photomicrographs were (A) • 16 and ( B ) x 158.
was low (1-5%) (Figure 1A and B). Increasing the concentration of polymer in the atomized solution to 10% resulted in the formation of a relatively smooth film over the external surface of PGA meshes, and utilizing a 15% solution resulted in the formation of a fibrous, non-homogeneous film over the PGA meshes (Figure 1C and D). Increasing the polymer concentration raises the viscosity of this solution and this likely increases the droplet size which is formed during the atomization process. This will effect how these droplets penetrate the PGA mesh, how they aggregate on the PGA mesh, and the rate of solvent evaporation. All of these factors will affect the pattern of bonding. To engineer a tissue with a desired three-dimensional structure, the cell delivery device must maintain a preconfigured geometry in the face of external forces during the process of tissue development. While the magnitude of the compressive forces that are exerted on implanted devices by the surrounding tissue are unclear, they are significant and will vary depending on the implant site. The magnitude of forces utilized in the present study to quantitate the compression resistance of devices in vitro was 50-200mN. This
Figure 10 (A) Low-power photomicrograph of a histological section of a bonded tube (5% PLLA; 30s) seeded with smooth muscle cells and endothelial cells in vitro as described in the Methods section. This cross-section was cut perpendicular to the axis of the tube's lumen. Highpower photomicrographs of (B) an interior section of the device and (C) a section adjacent to the lumen. Smooth muscle cells readily adhered to polymer fibres (p) and filled the interstices between polymer fibres (A and B), while endothelial cells formed a lining on the luminal surface (A and C; arrows). The original magnifications of these photomicrographs were (A) • (B) • and (C) • 158.
results in pressures ranging from approximately 50 to 200mmHg (6.65-26.6kPa) (assuming complete and continuous contact between the TMA compression tip and the tube). These pressures are in the same range Biomaterials 1996, Vol. 17 No. 2
138
The Biomaterials Silver Jubilee Compendium
Stabilized PGA tubes: D.J. Mooney et ai.
124
observed in blood vessels. Devices which were stable to high forces (PLLA bonded devices) were also stable after implantation into the ornentum of laboratory rats. The omentum was chosen as the implant site because it is highly vascularized, easily accessed and manipulated surgically, and its anatomic location makes it a preferred site to engineer a variety of gastrointestinal tissues (e.g. small intestine). The compressional forces exerted by the surrounding tissue are likely not as great as other potential implant sites (e.g. popliteal space). The formation of fibrovascular tissue in implanted tubes was not surprising, as it is well documented that this type of ingrowth occurs in porous, synthetic materials13,14. The ingrowth and organization of the fibrovascular tissue will also exert compressional forces on the forming tissue, although the magnitude of these forces is unclear. It is anticipated that the ingrowing fibrovascular tissue would have eventually filled the central lumen of the implanted tubes since there was no epithelial cell lining of the lumen. An endothelial cell lining would likely prevent this outcome. While large diameter synthetic blood vessels (>5 m m diameter) have been successfully utilized for years, prosthetic small diameter blood vessels (P, Eq. (1) reduces to D2(t)= 2 # ( t - P). The values of/~ and P were estimated from the slope and intercept of the line; this method for estimating /~ and P from experimental data was described in more detail in a previous publication [12].
3. Results
3.1. Microfabrication and characterization of parallel ridges~grooves Parallel polyimide ridges were successfully patterned on clean glass surfaces (Fig. l a-d). The ridges were 2 gm wide and 400 gm long. These two parameters were kept constant throughout the study. The height (h) was varied--either 5 lain (Fig. l a) or 3 lain (Fig. l b-d)--by altering the speed of rotation during spin coating. The repeat spacing (s) between ridges (both 5 and 3 gm tall) was varied from 6 to 14 lain in increment of 2 gm (in Fig. lb-d, h = 3 gm, only 6, 10 and 14 gm spacings are shown). SEM examination confirmed the scale and reproducibility of the patterns.
3.2. Cell morphology on microfabricated surfaces Neutrophil morphology on surfaces patterned with ridges/grooves was observed using either light microscopy (LM) (Fig. 2) or SEM (Fig. 3). LM observations revealed that cells consistently appeared to be within the grooves, rather than on the top of ridges, regardless of surface chemistry (Fig. 2). Uncoated and Au-Pd-coated surfaces supported similar cell morphology in the patterned region (compare cells within the patterned region of Fig. 2a-b). In contrast, cells on titaniumcoated patterned surfaces appeared more spread and elongated along the ridges (Fig. 2c). SEM studies of fixed cells on patterned surface with different geometry were consistent with LM observations; images of cells on the patterned surface revealed that almost 99% of cells were inside the grooves, regardless of ridge height and spacing (Fig. 3a-f). Cells appeared more elongated and confined in narrower grooves (6 lam) than in wider ones (14 ~tm). For example, most cells contacted two sides of ridges in narrow grooves (Fig. 3a and d) but cells contacted only one
195
The Biomaterials Silver Jubilee C o m p e n d i u m J. Tan, W.M. Saltzman / Biomaterials 23 (2002) 3215-3225
3219
Fig. 2. Light microscope images of neutrophils on microfabricated substrates (h = 5 btm, s = 10 btm) with various surface chemistry.
ridge in wide grooves (Fig. 3c and f). There appeared to be more opportunities for the cell surface to interact with the top of ridges in narrow and shallow grooves (Fig. 3a). The extent of cell-surface contact might be important in modifying cell motility, as described in Discussion. 3.3. Cell motility on microfabricated surfaces
Typical paths of neutrophil movement on surfaces with various microgeometry (h = 5 or 3 lam and s = 6, 10, 141.tm) and chemistry are shown in Figs. 4 and 5. More than 95% of neutrophils moved in the direction of the long axis of ridges/grooves (that is, movement was predominantly in the x-direction with IAyl 2 min. The presence of titanium did not effect the direction of cell movement but significantly reduced cell motility and eliminated the biphasic dependence on spacing. On unpatterned, titanium-coated surfaces, cells were almost immobilized with/~ only about 0.08 x 10 .9 cm2/s (Table 2). On patterned surfaces, cells were able to move a short distance (Fig. 5), but the motility was substantially slower than motility observed on uncoated or Au-Pd-coated surface with the same microgeometry (Table 2, Fig. 7b). The effect of spacing on cell motility was not as significant as those on un-coated and Au-Pdcoated surfaces. The greatest motility (3.2 x 10 .9 cm2/s)
Our studies revealed that more than 95% of human neutrophils migrated along the major axis of surfaces with parallel ridges/grooves of various spacing, depth and chemistry; this phenomenon was termed "contact guidance" by Weiss [3]. It has been suggested that neutrophils probably do not form focal contacts with substrata during migration, instead, cell membrane extensions probably determine cell speed [13-16]. On surfaces with the regularly arranged pillars or holes employed in our previous studies, cell membrane extensions occurred freely in both the x- and y-directions; cell motion was observed as a 2-D persistent random walk [8]. However, when cells were moving inside grooves (anisotropic topography) as in the experiments performed here, cell migration was observed along one axis and a 1-D persistence random walk was appropriate to describe such motion. In this situation, membrane extension appeared to be inhibited in the direction perpendicular to grooves and favored in the direction parallel to grooves [17,18]. In addition to guidance of direction of movement, neutrophil motility was enhanced for cells moving inside the grooves with particular characteristics. Our observations are consistent with studies using other cell types [18]. It was also interesting to note that neutrophils did change their direction of movement, but the frequency of directional change was less than that on smooth
The Biomaterials Silver Jubilee Compendium
199
J. Tan, W.M. Saltzman / Biomaterials 23 (2002) 3215-3225
3223
Fig. 8. Comparison of neutrophil motility on surfaces with isotropic and anisotropic features: holes vs. ridges/grooves.
surfaces. The mechanism is not clear at present time; we speculate that it is related to the inefficiency of reorganization of cytoskeleton structure upon contact with an adhesive vertical wall [19]. The speed of movement of a cell population (i.e. the cell motility coefficient) depended on groove spacing (6-14gm) and depth (3 or 5 gm). When the depth of grooves was 5gm, the behavior of cell motility as a function of repeat spacing could be described as biphasic. This biphasic behavior is similar to that observed in our previous studies of neutrophil migration on surfaces patterned with holes, in which the maximal cell motility was also observed at a spacing of ~ 10 gm [16]. However, there were noticeable differences in the cellular response to spacing. On ridge-patterned surfaces, cell motility decreased significantly with a spacing change of _+2~tm from 10gm whereas no significant change was observed with further changes in spacing; on hole patterned surfaces, the variation in motility with spacing was less pronounced (Fig. 8). We speculate that the mechanisms for cell migration on the two patterned substrates must be different. On surfaces with holes, cells appear to use the holes as mechanical edges to
"grab and pull" and the frequency of cell interaction with the edges is directly related to density (spacing) of the holes. However, this mechanism may not apply to neutrophil migration on parallel grooved surfaces, in which mechanical ridges are only available in one direction. SEM images suggest that cell surfaces interact with material of the groove bottom, side wall, and ridge top. The accumulative effect of these interactions probably determines the rate of cell motility. When the spacing was wide, 12 or 14 gin, cells only contacted one groove. Once cells were attached to this groove, they did not appear to dissociate easily. As a result, 12 and 14 gm spaced ridges produce a similar degree of "contact enhancement" for cell migration. When the repeat spacing was narrow (6 or 8 lam), cells were squeezed inside grooves, contacting materials on both sides, which might lead to some degree of restriction upon the cytoskeleton rearrangements necessary for movement [19]. The combination of "contact enhancement" and "restriction" probably produced no further change in cell motility as the spacing was changed from 8 to 6gm. An optimal spacing that was close to a cell diameter (10 lain) was required to achieve maximal cell
The Biomaterials Silver Jubilee Compendium
200 3224
J. Tan, W.M. Saltzman / Biomaterials 23 (2002) 3215-3225
[]
[]
[]
[]
I
"11
II
II
l
I
I
.... I '
I
I
[]
[]
[]
I
II
II
II
I
I
I
I
I
I
[]
[]
[]
I
II
1[
II
I
I
I
I
I
!
[]
[]
[]
(b) r ' - i
J i
i i
i
(c) l
I
I
i
I
Y
(a) L D x
I
t
,
I
I
I
i
i
(d) l ....
I
Fig. 9. Topographical transformation of pillars into ridges/grooves: (a) regularly distributed pillars; (b) pillars become wider in the horizontal direction (x-direction); (c) pillars contact each other in the horizontal direction to form continuous ridges (d).
motility; we speculate that this spacing maximizes contact with grooves but minimizes restrictions. Neutrophil migration was sensitive to groove depth, as observed for other cells (Table 1) [18,20,21]. Neutrophil motility decreased with decreasing groove depth probably due to less "contact enhancement". However, the effect of spacing for 3 ~m high features was not a simple biphasic function of the depth of grooves. We speculate that this behavior may be related to cell interaction with the top of the ridges and such interactions could play a role in modifying cell motility. Cells have more opportunities to interact with the top surfaces of narrower and shallower grooves. As a result, cell motility was the greatest on the surface with narrowest groove (6 ~tm). Surface chemistry was another important factor influencing cell migration speed. Surface energy (hydrophilicity/hydrophobicity) is often used to characterize materials with respect to cell behavior, although no general relationships between surface energy and cell behavior have emerged [22-25]. Contact angle measurements demonstrated that polyimide was more hydrophobic than glass, suggesting the possibility that neutrophils might respond to chemical anisotropy with fast motility on polyimide-patterned surfaces. The deposition of a thin layer of Au-Pd alloy converted these glass/polyimide patterns to an isotropic surface chemistry. Neutrophil migration on Au-Pd-coated surfaces did not differ from those on un-coated surfaces, either smooth or patterned ones. Together, these results indicate that (1) the chemical heterogeneity of polyimide and glass had a negligible impact on cell motility; and (2) the directional movement and the increased motility of neutrophils were indeed caused by the physical-rather than the chemical--patterns in the materials. On titanium-coated surfaces, for which the surface energy was much higher than glass and Au-Pd, cell migration was significantly slower. Other studies have
shown that neutrophil adhesion to metals generally increases with increasing surface energy (decreasing water contact angle) [26]. Therefore, the slower movement of neturophil on titanium-coated surfaces was probably due to stronger adhesion between the cells and the material. Apparently, neutrophil movement was also dependent on surface chemical property. The heterogeneity of polyimide and glass had a small effect on cell migration, probably because cell motility on these two materials was similar, even though the surface energy, as well as cell adhesion, was significantly different as we have previously shown [8]. Because the surface energy and cell motility of Au-Pd was similar to that of plain glass, the dependence of cell motility coefficient on spacing was almost identical on Au-Pd-coated and uncoated substrates patterned with ridges/grooves. Even though cell motility coefficient was increased 30-40 times due to the presence of ridges/grooves on titaniumcoated substrates, the effect of spacing (microgeometry) was not great. These results indicate that both surface topography and chemistry play important roles in the regulation of cell migration, and the effect of microgeometry can be overridden by a strong chemical property, such as the high adhesiveness of titanium. We recognize that protein adsorption is a key mediator in the interaction of cells with surfaces, so we intentionally performed the experiments in serumfree medium. Therefore, the only proteins that were present during cell migration are either trace amounts left from the whole blood during the separation (we estimate this to be less than a few nano gram per square centimeter in each experiment) or protein produced and secreted by the cells. Future experiments should examine the relationship of protein adsorption in the changes that we observe. Previous studies showed that neutrophil migration was significantly hindered by the presence of isotropic pillars [8]. But we note that pillar connected along one
201
The Biomaterials Silver Jubilee Compendium J. Tan, W.M. Saltzman / Biomaterials 23 (2002) 3215-3225
direction become a ridge (Fig. 9). Such topographical manipulation produces a substantial change in cell migration: not only in direction but speed and persistence time were also demonstrated to change in this study. Microfabrication technology provides a powerful tool for achieving precise surface topography. In addition, surface chemistry of a substrate can also influence neutrophil motility; this property can also be manipulated by microfabrication techniques. Therefore, the combination of both surface chemistry and topography should be considered to control and optimize neutrophil migration when designing implantable materials.
Ackno wledgements
We thank the CNF staff for their help in microfabrication technology. We thank Dr. S.H. Kang and Prof. Christopher K. Ober's laboratory for the assistance in contact angle measurement. We thank Hong Shen for many useful discussions, Thomas M. Yung and Sheryl Parker for the help in tracking cell positions in motility study. This work was supported by a grant from the National Science Foundation (BES-9710313) to WMS and it was performed in part at CNF (a member of National Nanofabrication Users Network) which is supported by the National Science Foundation under Grant ECS-9319005, Cornell University and industrial affiliates.
References [1] Saltzman WM. Cell interactions with polymers. In: Lanza RP, Langer R, Vacanti J, editors. Principles of tissue engineering. San Diego: Academic Press, 2000. [2] Harrison RG. The cultivation of tissues in extraneous media as a method of morphogenetic study. Anat Rec 1912;6:181-93. [3] Weiss P. In vitro experiments on the factors determining the course of the outgrowing nerve fiber. J Expt Zool 1934;68:348-93. [4] Weiss P. Experiments on cell and axon orientation in vitro: the role of colloidal exudates in tissue organization. J Expt Zool 1945;100:353-86. [5] Weiss P, Taylor AC. Fish scales as substratum for uniform orientation of cells in virto. Anat Rec 1956;124:381. [6] Curtis A, Wilkinson C. Topographical control of cells. Biomaterials 1997;18:1573-83. [7] Brunette DM, Chehroudi B. The effects of the surface topography of micromachined titanium substrata on cell behavior in vitro and in vivo. J Biomech Eng 1999;121:49-57. [8] Tan J, Shen H, Carter KL, Saltzman WM. Controlling human polymorphonuclear leukocytes motility using microfabrication technology. J Biomed Mater Res 2000;51:694-702. [9] Wilkinson PC, Shields JM, Haston WS. Contact guidance of human neutrophil leukocytes. Exp Cell Res 1982;140:55-62. [10] Tan J, Saltzman WM. Influence of synthetic polymers on neutrophil migration in three-dimensional collagen gels. J Biomed Mater Res 1999;46:465-74.
3225
[11] Glauert AM. Fixation, dehydration and embedding of biological specimens. Amsterdam: North-Holland Publishing Company, 1975. [12] Parkhurst MR, Saltzman WM. Quantification of human neutrophil motility in three-dimensional collagen gels. Effect of collagen concentration. Biophys J 1992;61:306-15. [13] Lackie JM. Aspects of the behaviour of neturophil leukocytes. In: Bellairs R, Curtis A, Dunn G, editors. Cell behaviour. New York: Cambridge University Press, 1982. p. 319-48. [14] Brown AF. Neutrophil and monocyte behavior in three-dimensional collagen matrices. Scanning Electron Microsc 1984;2: 747-54. [15] Mandeville JT, Lawson MA, Maxfield FR. Dynamic imaging of neutrophil migration in three dimensions: mechanical interactions between cells and matrix. J Leukoc Biol 1997;61:188-200. [16] Tan J, Shen H, Saltzman WM. Micron-scale positioning of features influences the rate of polymorphonuclear leukocyte migration. Biophys J 2001;81:2569-79. [17] Dunn GA, Heath JP. A new hypothesis of contact guidance of tissue cells. Expt Cell Res 1976;101:1-14. [18] Dalton BA, Walboomers XF, Dziegielewski M, Evans MDM, Taylor S, Jansen JA, Steele JG. Modulation of epithelial tissue and cell migration by microgrooves. J Biomed Mater Res 2001;56:195-207. [19] Dunn GA. Contact guidance of cultured tissue cells: a survey of potentially relevant properties of the substratum. In: Bellairs R, Curtis A, Dunn GA, editors. Cell behaviour. New York: Cambridge University Press, 1982. [20] Clark P, Connolly P, Curtis AS, Dow JA, Wilkinson CD. Topographical control of cell behaviour: II. Multiple grooved substrata. Development 1990;108:635-44. [21] Rajnicek AM, Britland S, McCaig CD. Contact guidance of CNS neurites on grooved quartz: influence of groove dimensions, neuronal age and cell type. J Cell Sci 1997;110:2905-13. [22] Horbett TA, Waldburger JJ, Ratner BD, Hoffman AS. Cell adhesion to a series of hydrophilic-hydrophobic copolymers studied with a spinning disc apparatus. J Biomed Mater Res 1988;22:384-404. [23] Tamada Y, Ikada Y. Fibrobalst growth on polymer surfaces and biosynthesis of collagen. J Biomed Mater Res 1994;28: 783-9. [24] Hallab NJ, Bundy KJ, O'Connor K, Moses RL, Jacobs JJ. Evaluation of metallic and polymeric biomaterial surface energy and surface roughness characteristics for directed cell adhesion. Tissue Eng 2001;7:55-71. [25] Tegoulia VA, Cooper SL. Leukocyte adhesion on model surfaces under flow: effects of surface chemistry, protein adsorption, and shear rate. J Biomed Mater Res 1999;50:291-301. [26] Nygren H, Hrustic E, Karlsson C, Oster L. Respiratory burst response of peritoneal leukocytes adhering to titanium and stainless steel. J Biomed Mater Res 2001;57:238-47. [27] Brunette DM. Fibroblasts on micromachined substrate orient hierarchically to grooves of different dimensions. Exp Cell Res 1986;164:11-26. [28] Wood A. Contact guidance on microfabricated substrata: the response of teleost fin mesenchyme cells to repeating topographical patterns. J Cell Sci 1988;90:667-81. [29] Perizzolo D, Lacefield WR, Brunette DM. Interaction between topography and coating in the formation of bone nodules in culture for hydroxyapatite- and titanium-coated micromachined surfaces. J Biomed Mater Res 2001;56:494-503. [30] Wojciak-Stothard B, Curtis A, Monaghan W, Macdonald K, Wilkinson C. Guidance and activation of murine macrophages by nanometric scale topography. Exp Cell Res 1996;223: 426-35.
202
This Page Intentionally Left Blank
The Biomaterials Silver Jubilee Compendium
The B i o m a t e r i a l s S i l v e r J u b i l e e C o m p e n d i u m
203
Available online at www.sciencedirect.com
8ClENCE~DIRECT e
ELSEVIER
Biomaterials
Biomaterials 24 (2003) 893-900 www.elsevier.com/locate/biomaterials
Photopolymerized hyaluronic acid-based hydrogels and interpenetrating networks Yong Doo Park, Nicola Tirelli*, Jeffrey A. Hubbell Department of Materials, Institute for Biomedical Engineering, Swiss Federal Institute of Technology and University of Zi;trich, Moussonstrasse 18, CH-8044 Zi;trich, Switzerland
Received 3 May 2002; accepted 6 September 2002
Abstract
Hyaluronic acid (HA) was derivatized with methacrylic esters used for the preparation of hydrogels via photopolymerization. Poly(ethylene glycol) diacrylate (PEG-DA) with a molecular weight of 570 was also used as a comacromonomer to improve elastic modulus and swelling behavior. The hydrogels were readily degraded by hyaluronidase and their mechanical properties could be modulated by HA molecular weight and concentration of PEG-DA. The incorporation of RGD peptides allowed modulation of the HA properties from cell non-adhesive to adhesive. Human dermal fibroblasts were cultured on the RGD, RDG, and nonfunctionalized HA hydrogels for up to 7 d, showing adhesion and proliferation only with incorporated RGD. 9 2002 Elsevier Science Ltd. All rights reserved. Keywords." Hyaluronic acid; Hydrogels; Photopolymerization; RGD; Cell adhesion
I. Introduction
The physiological significance of hyaluronic acid (HA), the linear, very high molecular weight (up to 1-2 million Da) glycosaminoglycan copolymer of D-glucuronic acid and N-acetyl-D-glucosamine that is found in all connective tissues [1], is largely attributable to its unique viscoelastic properties [2]. HA is also known to play a role in promoting cell motility and proliferation [3]. At least three HA cell surface receptors, namely CD44, R H A M M , and ICAM-1 [4-6], have been identified and influence HA activity in processes such as morphogenesis, wound repair, inflammation, and metastasis [7-10]. Much of the attention devoted to HA in the biomaterials field has been motivated based upon its specific chemical properties: (a) it can be obtained in wide range of molecular weights (by controlled hyaluronidase degradation); (b) it is enzymatically remodeled in vivo and in culture in presence of selected cell types (e.g. chondrocytes); (c) it can be functionalized with reactive groups, undergo cross-linking reactions and *Corresponding author. Tel.: + 41-1-632-63-48; fax: + 41-1-632-1214. E-mail address:
[email protected] (N. Tirelli).
produce materials in the form of hydrogels; (d) in the native form, it is substantially non-adhesive to cells [11], but can be functionalized or blended with cell-adhesive materials to tailor its adhesive properties [12]. Unmodified [13-15] and derivatized [16,17] HAs have been used for a variety of clinical applications such as ocular surgery, viscosupplementation for arthritis, wound healing, and plastic surgery, where HA is generally used as anti-adhesive component. In many cases, HA has been chemically modified, exploiting the reactivity of its carboxy and hydroxy groups. The carboxy groups have been modified by esterification with various techniques [18] and hydrogels have been obtained via biscarbodiimide coupling and hydrazide cross-linking [19-22]. The hydroxyl group has also been used for bisepoxide or divinyl sulfone cross-linking [23]. The overall goal of our work was to produce materials, e.g. for cartilage repair, through the synthesis of HA derivatives that can be cured in situ and can then promote cell in-growth; the cross-linked materials should not swell substantially after curing, should possess mechanical properties that are useful in a surgical setting, should be selectively degradable by hyaluronidase, and finally should promote cell adhesion. A mild cross-linking technique is necessary for the in vivo application of such a process; we have chosen
0142-9612/02/$- see front matter 9 2002 Elsevier Science Ltd. All rights reserved. PII: SO 142-96 12(02)00420-9
The Biomaterials Silver Jubilee Compendium
204 894
Y.D. Park et al. / Biomaterials 24 (2003) 893-900
CH20H COONa
0
00
0 NHCOCH. HN
~
+
(EDC) -..--~]~ H~
o
\ -]
HN...
NH 2
0 HO _~
4-o'
CH=OH 0 NHCOCH=
!.
Fig. 1. Scheme of modification of HA using a methacrylating agent. N-(3-amonopropyl)-methacrylamide hydrochloride was mixed with HA in water and EDC was added for the coupling reaction.
photopolymerization, a technique previously used in our laboratory [24]. In this study, HA-based hydrogels were prepared via photopolymerization of pendent methacrylic esters, previously introduced through functionalization of the carboxylic groups (see Fig. 1). Poly(ethylene glycol) diacrylate (PEG-DA) was copolymerized with HA, in order to improve elastic modulus and swelling behavior. HA-based hydrogels were then made cell-adhesive, introducing integrin-binding peptide via Michael-type addition of a cysteine-containing RGD sequence onto the PEG acrylates [25,26]; in this way, only a fraction of the double bonds was functionalized with negligible effect on the mechanical properties, while at the same time a sufficient number of RGD groups was immobilized to influence cell adhesion. The presence of the RGD peptide showed a dramatic effect on fibroblast adhesion and proliferation on the gels.
2. Materials and methods
was then boiled and the white protein precipitate was filtered using 0.45 gm nylon filter. The MW of degraded HA was determined using gel permeation chromatography (GPC) using pullulan as a standard and phosphate buffered saline (PBS, 10 mM in normal saline, pH 7.4) as a mobile phase.
2.3. Methacrylation of HA To a solution of 0.25mmol of degraded or nondegraded HA (based on the mer MW) in 100 ml distilled water were added 0.096g of EDC (0.5mmol) and 0.089 g of N-3-aminopropyl methacrylamide (0.5 mmol). The reaction mixture was incubated for 2 h at pH 6.5. The same amounts of EDC and N-(3-aminopropyl) methacrylamide were added after 2 h and were further incubated for 2 more hours. The solution was filtered through a 0.45gm nylon filter, then dialyzed against 10 mM sodium chloride for 1 d and distilled water for 2 d, and finally lyophilized for 4d to give a white cake of solid methacrylated HA (HA-Ac). The degree of acrylation was examined using 1H-NMR.
2.1. Materials HA sodium salt (MW 1-2 million Da) and N-(3aminopropyl) methacrylamide were purchased from Genzyme Inc. and Polysciences Inc., respectively. Hyaluronidase and N-(3-dimethylpropyl)-N-ethylcarbodiimide hydrochloride (EDC) were from Sigma (St. Louis, MO). Poly(ethylene) glycol diacrylate (MW 570) and N-vinyl pyrrolidone were purchased from Merck.
2.2. Degradation of HA HA sodium salt was dissolved in 100 ml distilled water to a concentration of 0.5 mg/ml. After complete HA dissolution, 1000 U of hyaluronidase was added to the solution and incubated for 16 h. The reaction mixture
2.4. Introduction of RGD peptide into the polymerization mixture The peptides G C G Y G R G D S P G and GCGYGRDGSPG were synthesized using standard solid-phase synthesis. The peptides were dissolved in 10 mM HEPES saline buffer (10 mM in normal saline, pH 7.4). PEG-DA was then added in a 10:1 acrylate/thiol ratio. The reaction mixture was incubated for l h at room temperature, allowing Michael-type addition to take place, with the thiol on the cysteine residues in the peptide as the Michael donor and the acrylate groups on the PEG termini as the Michael acceptors. The partially peptide-functionalized PEG-DA was not isolated, but used for photopolymerization experiments according to the protocol below. Because of the high ratio of
205
The Biomaterials Silver Jubilee Compendium Y.D. Park et al. / Biomaterials 24 (2003) 893-900
acrylates to thiols, only a very small functionalized PEGs will be unreactive tionalization of the PEG diacrylate, peptide will be incorporated dangling functionalized PEG acrylate.
number of the due to difunci.e. almost all from a singly
2.5. Photopolymerization In a typical photopolymerization experiment, 10mg HA-Ac or native HA was dissolved in 0.1 ml of 10 mM HEPES saline buffer (10mM in normal saline, pH 8.0) containing 100mM triethanolamine, 1 mM eosin Y, and 1% w/v N-vinyl pyrrolidone (NVP). A Xenon arc lamp was utilized to provide illumination for l min at 480-520 nm and 75 W/cm 2. To modulate the mechanical properties of the HA-based hydrogel, a variable quantity (generally 0.006 or 0.012mmol) of PEG-DA or peptide-modified PEG-DA mixture was also added before photopolymerization.
2.6. Rheology of gel formation The mechanical properties of the HA-based hydrogels were measured on a 120CVOHR Bohlin Rheometer, using a parallel plate geometry; a quartz bottom plate allowed the use of an optical fiber to perform photopolymerization experiments. In a typical experiment, 50 ~tl of a solution containing the photopolymerization reagents was placed between the plates, at a distance set to 0.1 mm. The frequency of oscillation was set to 10 Hz. During the photopolymerization, the changes of elastic and viscous moduli and of the phase angle were monitored.
2. 7. Swelling of the hydrogels Hydrogels samples with different HA-Ac MW, acrylation degree, and concentration of PEG-DA were prepared in Eppendorf tubes. The wet weight was measured, incubating the gels in water overnight at room temperature for swelling. The swelling ratio was measured by comparing change of weight of hydrogel before and after incubation.
895
merization and subsequence initiation of the rheological study was kept short relative the kinetics of degradation. In a different set of experiments the degradation products were studied using GPC analysis. The gels were prepared and incubated in HEPES saline buffer (10mM in normal saline, pH 8.0) overnight for swelling, then placed in 5 ml tubes, containing 200~tl HEPES saline buffer (10 mM in normal saline, pH 8.0). Hyaluronidase (1000 U) was added to these solutions and incubated at 37~ for 2 d. Samples of the solutions were collected at regular intervals and analyzed.
2.9. Cellculture Three different types of hydrogel were prepared, namely one with no adhesion peptide (as a control), one with the RGD-containing peptide, and one with a nonadhesive peptide, namely with the inactive sequence R D G (as a second control). Hydrogels were incubated overnight to swell in PBS. Human dermal fibroblasts (5000 cells/well) were seeded on hydrogels in D M E M with 10% fetal bovine serum. Cells were cultured in the same medium in a humidified incubator (5% CO2, 37~ Cell adhesion was documented by photomicrography for periods up to 1 week. The proliferation of fibroblasts on the hydrogel was measured using WST-1 kit, in which the derivatives of tetrazolium are formed and converted to colored formazan by mitochondrial dehydrogenases (Roche, Indianapolis, IN). Briefly, fibroblasts (5000 cells/well) were seeded in a 96-well tissue culture plates, which were already covered with the different types of pre-swelled hydrogels. Cells were incubated for 3 d in DMEM with 10% fetal bovine serum at 37~ in a humidified incubator (5% CO2, 37~ After incubation, 10~tl of cell proliferation reagent WST-1 was added to each well. The reagent was also added to the wells containing hydrogel without cells for the control. The culture plates were incubated for 3 h in a humidified incubator. After incubation, the multiwell plates were shaken 1 min on a shaker. The formation of red formazan was measured at 420 nm using ELISA reader.
3. Results and discussion
2.8. Enzymatic degradation of hydrogels 3.1. Modification of HA In a rheological study, 100U of hyaluronidase was added to 100 ~tl of solution before polymerization. The solution was photopolymerized in the rheometer as already described at a pre-set temperature of 37~ After illumination, the evolution of the elastic and viscous moduli was monitored for 10min. The enzyme was incorporated before gelation to prohibit gradients of the enzymes within the hydrogel; the time between mixing of the enzyme with the HA-Ac solution and photopoly-
HA of MW 1-2 million Da was used as a starting material, and the molecular weight of this material was reduced enzymatically by treatment with hyaluronidase. Analysis by GPC demonstrated reduction to an MS of approx. 50kDa after 16h enzymatic treatment (see Fig. 2). Both native and enzymatically degraded HA were subsequently used for chemical modification. Both HAs
206
896
The Biomaterials Silver Jubilee C o m p e n d i u m Y.D. Park et al. / Biomaterials 24 (2003) 893-900
were modified by using N-(3-aminopropyl) methacrylamide as an acrylating agent and the water soluble EDC as a coupling agent, incubating the mixture for 4 h at room temperature. The degree of acrylation of roughly 10% was determined by 1H-NMR, comparing the singlet peak of methyl group in HA acetamide (2.7 ppm) and the multiplet peaks of the acrylic double bond (5.3 and 5.6 ppm). 3.2. Gel preparation and mechanical properties
In the photopolymerization experiments, eosin Y was used as a visible light sensitizer, triethanolamine was employed as an initiator, and PEG-DA and NVP were used as comacromonomer and comonomer, respectively, according to a technique developed in our laboratory for hydrogel synthesis [27]. The use of a low MW PEG-DA allows the material to be toughened and permits reduction of its hydrophilicity and degree of
Fig. 2. GPC results of HA before and after treatment with hyaluronidase. Hyaluronidase was added to the HA solution at a concentration of 10U (A), 50U (11), and 100U (O) per mg HA. The reaction mixture was incubated for 20 h.
swelling (PEG hydrophilicity increases with MW), without reduction in biocompatibility; NVP was incorporated for the primary reason of enhancing the rate of the gelation reaction. After 1 min of illumination with a Xenon arc lamp, clear and soft hydrogels were obtained. Both HA-Ac and native HA were used in gel formation generating respectively, a cross-linked copolymeric network and a pseudo-interpenetrating polymer network (pIPN: one domain is a network and the other a polymer dispersed in the network; a real IPN is constituted by two or more networks), in which the native HA is present within the network as a physically entrapped non-cross-linked polymer. The gel points (defined as crossing points between viscous and elastic modulus, see Fig. 3) were detected in both cases; the rate of gelation of HA-Ac and P E G D A was faster than was observed with the formation of the pIPN, i.e. with HA and PEGDA. Covalent incorporation of HA-Ac within the gel was demonstrated by the fact that the complex modulus reached a plateau after 1 min and stabilized at around 10kPa, whereas the pIPN gel containing native HA yielded a complex modulus of approximately 1 kPa after the same duration of irradiation. In exploration of hydrogel preparation, we focused our attention on some of the factors affecting mechanical strength and post-gelation swelling, namely MW of the HA-Ac and degree of acrylation of the derivatized HA, as well as concentration and composition of the comacromonomer mixture. As expected (see Fig. 4), the complex modulus increased as the number of reactive groups in multifunctional monomers were increased, i.e. with the degree of HA acrylation (0% vs. 10%), and the PEG-DA concentration (2% vs. 5% (w/v)); furthermore, higher HA-Ac MW (1-2 million Da) resulted in gels with a higher complex modulus than was observed in gels formed from lower M W HA-Ac (50kDa), namely 5 x 104 vs. 106 Pa. The equilibrium extent of swelling of the hydrogel may be of significant interest in two regards: first, the
Fig. 3. Viscous and elastic modulus evolution during photopolymerization. Panel A: native HA with 5% PEG-DA. Panel B: HA-Ac with 5% PEG-DA.
207
The Biomaterials Silver Jubilee Compendium Y.D. Park et al. / Biomaterials 24 (2003) 893-900
Fig. 4. Influence of the hydrogel composition on the mechanical properties. The complex modulus of the hydrogel was measured by changing the acrylation (solid bar; acrylated HA, hatched bar; native HA), molecular weight of HA (degHA; 50,000, and HA; 500,000) and PEG-DA concentration (2% and 5%).
mechanical and cell invasion characteristics of the hydrogel would be expected to relate to swelling, and second it may be desirable to limit the post-gelation swelling, to enable a surgeon to apply the material by photopolymerization in a size and shape that will be retained in vivo, after an opportunity for equilibration from the as-gelled state. It should be possible to manipulate both equilibrium swelling and post-gelation incremental swelling, in that they should be dependent upon cross-linking density and hydrophobicity, as well as the concentration of precursors in the macromonomer solution. The absolute level of swelling is in a general sense related to the complex modulus results: the higher the modulus, the higher the cross-linking density and the lower the swelling. An exception is that an increase in the MW of HA increases the cross-linking density, but also the internal osmotic pressure (the volume occupied is exponentially increasing with the MW [28,29]), and the second factor can overwhelm the first one. In fact, the lowest swelling was shown by the formulation having the higher cross-linking density compatible with a low MW of HA, that is highest PEG-DA content, lowest MW of HA-Ac (see Fig. 5). The final aspect of gel synthesis consisted of the introduction of a cell-adhesive peptide, comprising the RGD sequence, and this was accomplished by the means of a Michael-type addition reaction to attach the peptide to one end of a PEG-DA chain, the other acrylate terminus then remaining to participate in the photopolymerization reaction. The peptide was designed so as to contain a cysteine residue, bearing a thiol group that is able to rapidly react with the acrylic groups in PEGDA at pH 7.4. This scheme for incorporation of
897
Fig. 5. Influence of the hydrogel composition on the swelling behavior. The swelling of the hydrogel was measured as a function of the acrylation, molecular weight of HA (degHA; 50,000, and HA; 500,000) and PEG-DA concentration (2% and 5%).
adhesion peptides into the precursors of photopolymerizable gels has been described in detail elsewhere [25]; it has been previously demonstrated that the rate of the Michael-type addition reaction is substantially higher than the rate of the competitive disulfide bonding reaction under these reaction conditions. After incubation of the cysteine-containing peptide with an excess of PEG-DA (in order to obtain a mixture mostly constituted by PEG-DA with some singly coupled chains, with one acrylate group remaining), assuming the reaction to be quantitative approximately 10% of acrylic groups were functionalized with peptides. The remaining acrylate groups in the peptide-grafted PEG monoacrylate (the minor component) and the remaining PEG-DA (the major component) were then used for the photopolymerization reaction. The mechanical properties of the RGD-containing hydrogels were statistically indistinguishable from the hydrogels not containing the RGD-peptide (data not shown), presumably because such a small fraction of the acrylate groups had been consumed by the Michael-type addition reaction that was used to incorporate the peptide.
3.3. Deoradation of HA-based hydrogels The acrylation of HA and the presence of nondegradable PEG-DA introduce new features on the HA backbone, possibly interfering with the enzymatic degradation. The effect of hyaluronidase on the HAAc, measured by GPC, and on the formed hydrogels, monitored theologically and by characterization of the MW of released soluble products, was examined. The hyaluronidase degradation rate of the acrylated polymer
208
The Biomaterials Silver Jubilee Compendium
898
Y.D. Park et al. / Biomaterials 24 (2003) 893-900
(HA-Ac) was somewhat slower, by about 70%, than was observed with native HA; the derivatization may cause some steric hindrance to the enzyme. In the rheological characterization of the cross-linked gels, the change of complex modulus was monitored by incubating the hydrogels at 37~ with an excess of hyaluronidase (100U/50 ~tl solution) inside; the enzyme was mixed in the sample before polymerization, and it could safely be assumed to produce a negligible effect over the time scale of photopolymerization. In a typical experiment
10000
mmm mn mm
13..
v
(,t)
1000
3.4. Cell behavior
_= '--I
mi
0
E
mm:
x
E
(see Fig. 6), the complex modulus increased from roughly 200Pa for the precursor solution to 10kPa during a 1 min photopolymerization. After illumination for 4 min, the value started fluctuating because of loss of mechanical integrity, and decreased down to 500Pa. After 10 min, the value reached about 20 Pa indicating a complete degradation of the gel. GPC analysis was performed on the degradation products of HA-Ac hydrogels formed with and without the incorporation of PEG-DA in the precursor mixture, after incubation of the resulting gel with hyaluronidase (1000 U). The same retardation effect observed in the enzymatic degradation of HA-Ac seems to occur also in the case of the pIPN, but the somewhat slower degradation did not give products of appreciable different molecular weight (see Fig. 7).
9
mm
100
l~/llil
0
nm n
c)
mm 9 mUm
I
I
0
200
,
I
'
400
I
600
;
mUunm_ 9
I
800
'
I
1000
Time (sec) Fig. 6. Rheological study of the preparation and degradation of a hyaluronidase-containing hydrogel. Hyaluronidase (100 U) was added to the 100mg/ml HA-Ac solution containing the photopolymerizing reagents. The gel was obtained via illumination for 1 min.
Human dermal fibroblasts were cultured on the surface of hydrogels containing no peptide (as a control), the active RGD peptide, or the inactive peptide RDG (as an additional control, with similar overall physicochemical properties of the RGD peptide, but without any cell-binding bioactivity). After washing and swelling, fibroblasts were seeded on the three different hydrogel surface, respectively. Cells on the HA-Ac gel and RDG peptide-containing hydrogel did not spread at all after 1 or 2 d in culture. By contrast, cells on RGDcontaining hydrogel spread well, comparably to human dermal fibroblasts on tissue culture plates, and
Fig. 7. Molecular weight determination of the degradation products released from HA-based hydrogels. Preswelled hydrogels (wet weight 0.1 g) were incubated by adding 1000 U hyaluronidase in 200 gl HEPES buffer. After collecting the sample (50 gl), the molecular weight of the degraded product was measured by GPC.
209
The Biomaterials Silver Jubilee Compendium Y.D. Park et al. / Biomaterials 24 (2003) 893-900
899
Fig. 9. Proliferation of fibroblasts as determined by mitochondrial activity. Human dermal fibroblasts were cultured for 3d on four different substrates: (a) reagent only (blank), (b) HA-Ac hydrogels (negative control), (c) RDG-containing hydrogels (negative control), (d) RGD-containing hydrogels, and (e) tissue culture polystyrene (positive control) in 96 well plates. After adding 10 Ftl WST-1 reagent, culture plates were incubated for 3 h and formation of red formazan was measured at 420 nm using an ELISA plate reader.
Fig. 8. Fibroblasts cultured on different HA-based hydrogels. Human dermal fibroblasts were cultured for 7 d on three different substrates: acrylated HA-based hydrogel (a), RDG-modified HA-based hydrogel (b), and RGD-modified HA-based hydrogel (c).
useful for anti-adhesive treatments [11,30]. The cell proliferation on the RGD hydrogel was by contrast found to be around 70% of that observed on the positive control material, namely tissue culture-modified polystyrene (see Fig. 9). HA-based gels containing the inactive RDG peptide offered a much less favorable condition for cell proliferation, with about one-half of the proliferation observed with the active RGD peptide.
4. Conclusions
proliferated and eventually covered the surface of the hydrogel throughout a 7 d period of culture (Fig. 8). Cells on the RGD-containing hydrogel were stained with rhodamine-phalloidin to label F-actin stress fibers of the cytoskeleton, which would be expected if cell spreading morphology were normal. The observed stress fiber morphology was qualitatively consistent with that observed on tissue culture plates and was, by contrast, lacking in cells cultured on RDG-containing hydrogels (data not shown). Cell proliferation was measured after the third day of culture, using the cell proliferation assay reagent WST1, which reflects the mitochondrial dehydrogenase activity of the cell. Cells seeded on tissue culture plates or HA-Ac hydrogels (lacking peptides) were used as a positive and negative control, respectively. The HA hydrogel did not provide conditions that were suitable for cell proliferation. Indeed, HA is known from the literature to provide cell-resistant surfaces and to be
A process for the in situ and possibly in vivo preparation of hydrogels based on HA has been developed. The rationale for the use of this natural polymer was its well-known biocompatibility and very low protein absorption, which make it an excellent antiadhesive material and a kind of natural and enzymatically degradable analog of PEG. The goals that have been achieved in this study are that (a) the functionalization with methacrylic groups and an appropriate formulation of the comacromonomer mixture did produce hydrogels, using photopolymerization; (b) with this technique, a polymer solution can be applied to any shape and form according to a defect region and after a short illumination can produce a gel with negligible post-gelation swelling and interesting mechanical properties; (c) the gels obtained with this technique were demonstrated to be still degradable by hyaluronidase; in other words, HA was an essential part of the network. These HA-based hydrogels showed the
210 900
The Biomaterials Silver Jubilee Compendium Y.D. Park et al. / Biomaterials 24 (2003) 893-900
typical anti-adhesive properties of HA materials; in our case, fibroblasts on the gel surface did not spread at all and their proliferation was as low as control, blank values. It seems that H A alone cannot provide a sufficient signal for cell adhesion and spreading, although may cell types do possess a HA receptor [31]. Finally, the introduction of an R G D peptide showed dramatic changes in terms of cell adhesion and proliferation. Cells cultured on the RGD-modified hydrogel proliferated and grew to confluence, whereas cells on the R D G control hydrogel did not show any sign of spreading over long durations in culture and in the presence of serum proteins.
References [1] Laurent TC, Laurent UB, Fraser JR. The structure and function of hyaluronan: an overview. Immunol Cell Biol 1996;74:A1. [2] Lee B, Litt M, Buchsbaum G. Rheology of the vitreous body: Part 3. Concentration of electrolytes, collagen and hyaluronic acid. Biorheology 1994;31:339-51. [3] Trochon V, Mabilat C, Bertrand P, Legrand Y, Smadja-Joffe F, Soria C, Delpech B, Lu H. Evidence of involvement of CD44 in endothelial cell proliferation, migration and angiogenesis in vitro. Int J Cancer 1996;66:664-8. [4] Entwistle J, Hall CL, Turley EA. HA receptors: regulators of signalling to the cytoskeleton. J Cell Biochem 1996;61:569-77. [5] Yang B, Zhang L, Turley EA. Identification of two hyaluronanbinding domains in the hyaluronan receptor RHAMM. J Biol Chem 1993;268:8617-23. [6] Lesley J, Hyman R, Kincade PW. CD44 and its interaction with extracellular matrix. Adv Immunol 1993;54:271-335. [7] Chen WY, Abatangelo G. Functions of hyaluronan in wound repair. Wound Repair Regen 1999;7:79-89. [8] Menzel EJ, Farr C. Hyaluronidase and its substrate hyaluronan: biochemistry, biological activities and therapeutic uses. Cancer Lett 1998;131:3-11. [9] Lesley J, Hyman R, English N, Catterall JB, Turner GA. CD44 in inflammation and metastasis. Glycoconj J 1997;14:611-22. [10] King SR, Hickerson WL, Proctor KG. Beneficial actions of exogenous hyaluronic acid on wound healing. Surgery 1991;109: 76-84. [11] Pavesio A, Renier D, Cassinelli C, Morra M. Anti-adhesive surfaces through hyaluronan coatings. Med Device Technol 1997; 8(20-1):24-7. [12] Liu LS, Thompson AY, Heidaran MA, Poser JW, Spiro RC. An osteoconductive collagen/hyaluronate matrix for bone regeneration. Biomaterials 1999;20:1097-108. [13] Balazs EA, Denlinger JL. Clinical uses of hyaluronan. Ciba Found Symp 1989;143:265-75.
[14] Wen DY. Intra-articular hyaluronic acid injections for knee osteoarthritis. Am Fam Physician 2000;62:565-70. 572. [15] Rosier RN, O'Keefe RJ. Hyaluronic acid therapy. Instr Course Lect 2000;49:495-502. [16] Vercruysse KP, Prestwich GD. Hyaluronate derivatives in drug delivery. Crit Rev Ther Drug Carrier Syst 1998;15:513-55. [17] Radice M, Brun P, Cortivo R, Scapinelli R, Battaliard C, Abatangelo G. Hyaluronan-based biopolymers as delivery vehicles for bone-marrow-derived mesenchymal progenitors. J Biomed Mater Res 2000;50:101-9. [18] Campoccia D, Doherty P, Radice M, Brun P, Abatangelo G, Williams DF. Semisynthetic resorbable materials from hyaluronan esterification. Biomaterials 1998;19:2101-27. [19] Prestwich GD, Marecak DM, Marecek JF, Vercruysse KP, Ziebell MR. Controlled chemical modification of hyaluronic acid: synthesis, applications, and biodegradation of hsdrazide derivatives. J Controlled Rel 1998;53:93-103. [20] Benedetti L, Cortivo R, Berti T, Berti A, Pea F, Mazzo M, Moras M, Abatangelo G. Biocompatibility and biodegradation of different hyaluronan derivatives (Hyaff) implanted in rats. Biomaterials 1993;14:1154-60. [21] Kuo JW, Swann DA, Prestwich GD. Chemical modification of hyaluronic acid by carbodiimides. Bioconj Chem 1991;2:232-41. [22] Tomihata K, Ikada Y. Crosslinking of hyaluronic acid with water-soluble carbodiimide. J Biomed Mater Res 1997;37:243-51. [23] Larsen NE, Pollak CT, Reiner K, Leshchiner E, Balazs EA. Hylan gel biomaterial: dermal and immunologic compatibility. J Biomed Mater Res 1993;27:1129-34. [24] Hill-West JL, Chowdhury SM, Slepian MJ, Hubbell JA. Inhibition of thrombosis and intimal thickening by in situ photopolymerization of thin hydrogel barriers. Proc Natl Acad Sci USA 1994;91:5967-71. [25] Elbert DL, Hubbell JA. Conjugate addition reactions combined with free-radical cross-linking for the design of materials for tissue engineering. Biomacromolecules 2001;2:430-41. [26] Lutolf M, Tirelli N, Cerritelli S, Cavalli L, Hubbell JA. Systematic modulation of Michael-type reacti,~ity of thiols through the use of charged amino acids. Bioconi Chem 2001; 12:1051-6. [27] Hubbell JA. Bioactive biomaterials. Curr Opin Biotechnol 1999; 10:123-9. [28] Kobayashi Y, Okamoto A, Nishinari K. Viscoelasticity of hyaluronic acid with different molecular weights. Biorheology 1994;31:235-44. [29] Bothner H, Wik O. Rheology of hyaluronate. Acta Otolaryngol Suppl 1987;442:25-30. [30] Schier F, Danzer E, Bondartschuk M. Hyaluronate, tetrachlorodecaoxide, and galactolipid prevent adhesions after implantation of Gore-Tex and dura mater into the abdominal wall in rats. Pediatr Surg Int 1999;15:255-9. [31] Hu M, Sabelman EE, Lai S, Timek EK, Zhang F, Hentz VR, Lineaweaver WC. Polypeptide resurfacing method improves fibroblast's adhesion to hyaluronan strands. J Biomed Mater Res 1999;47:79-84.
The Biomaterials Silver J u b i l e e C o m p e n d i u m
211
Available online at www.sciencedirect.com
SCIENCE~DIRECT e
ELSEVIER
Biomaterials
Biomaterials 24 (2003) 2309-2316 www.elsevier.com/locate/biomaterials
Cell sheet engineering for myocardial tissue reconstruction Tatsuya Shimizu, Masayuki Yamato, Akihiko Kikuchi, Teruo Okano* Institute of Advanced Biomedical Engineering and Science, Tokyo Women's Medical University, 8-1 Kawada-cho, Shinjuku-ku, Tokyo 162-8666, Japan
Received 28 October 2002; accepted 9 December 2002
Abstract
Myocardial tissue engineering has now emerged as one of the most promising treatments for the patients suffering from severe heart failure. Tissue engineering has currently been based on the technology using three-dimensional (3-D) biodegradable scaffolds as alternatives for extracellular matrix. According to this most popular technique, several types of 3-D myocardial tissues have been successfully engineered by seeding cardiomyocytes into poly(glycolic acid), gelatin, alginate or collagen scaffolds. However, insufficient cell migration into the scaffolds and inflammatory reaction due to scaffold biodegradation remain problems to be solved. In contrast to these technologies, we now propose novel tissue engineering methodology layering cell sheets to construct 3-D functional tissues without any artificial scaffolds. Confluent cells on temperature-responsive culture surfaces can be harvested as a viable contiguous cell sheet only by lowering temperature without any enzymatic digestions. Electrical communications are established between layered cardiomyocyte sheets, resulting in simultaneous beating 3-D myocardial tissues. Layered cardiomyocyte sheets in vivo present long survival, macroscopic pulsation and characteristic structures of native heart tissue. Cell sheet engineering should have enormous potential for fabricating clinically applicable myocardial tissues and should promote tissue engineering research fields. 9 2003 Elsevier Science Ltd. All rights reserved. Keywords." Myocardial tissue engineering; Cell sheet; Cardiac myocyte; Transplantation; Temperature-responsive culture surface
1. Introduction
Recently, alternative treatments for cardiac transplantation have been strongly requested to repair damaged heart tissue, because the utility of heart transplantation is limited by donor shortage. Cell therapy is now considered to be one of the most effective treatments for impaired heart tissue [1,2]. Direct transplantation of cell suspension has been researched since the early 1990s [3]. In these studies, survival of transplanted cells, integration of native and grafted cells, and improvement of host cardiac function have been reported. It is a critical point how to isolate and expand clinically transplantable myocardial cell source. Autologous myoblast transplantation has been performed clinically and the contraction and viability of grafted myoblasts have been confirmed [4]. Multipotent bone marrow cells or embryonic stem cells have been
*Corresponding author. Tel.: + 81-3-3353-8111x30234; fax: + 81-33359-6046. E-mail address."
[email protected] (T. Okano).
now aggressively investigated as possible candidates for human implantable myocardial cell source [5-8]. In direct injection of dissociated cells, it is difficult to control shape, size and location of the grafted cells. Additionally, isolated cell transplantation is not enough for replacing congenital defects. To overcome these problems, research on fabricating three-dimensional (3D) cardiac grafts by tissue engineering technology has also now begun [9]. Tissue engineering has currently been based on the concepts that 3-D biodegradable scaffolds are useful as alternatives for extracellular matrix (ECM) and that seeded cells reform their native structure in according to scaffold biodegradation [10]. This context has been used for every type of tissue. In myocardial tissue engineering, poly(glycolic acid) (PGA), gelatin and alginate have been used as prefabricated biodegradable scaffolds. Papadaki et al. engineered 3-D cardiac constructs by using PGA scaffolds processed into porous meshes and rotating bioreactors [11]. Li et al. have demonstrated that transplantation of tissue-engineered cardiac grafts using biodegradable gelatin sponges replaced myocardial scar and right ventricular outflow track defect [12,13].
0142-9612/03/S-see front matter 9 2003 Elsevier Science Ltd. All rights reserved. doi: 10.1016/S0142-9612(03)00110-8
212 2310
The Biomaterials Silver Jubilee Compendium T. Shimizu et al. / Biomaterials 24 (2003) 2309-2316
Furthermore, Leor et al. reported that bioengineered heart grafts using porous alginate scaffolds attenuated left ventricular dilatation and heart function deterioration in myocardial infarction model [14]. As the technique premixing cells and ECM alternatives instead of seeding cells into preformed scaffolds, Zimmermann et al. engineered 3-D heart tissue by gelling the mixture of cardiomyocytes and collagen solution [15,16]. The construct has allowed direct measurement of isometric contractile force as heart tissue model. In spite of these desirable results, insufficient cell migration into scaffolds and inflammatory reaction due to scaffold biodegradation remain problems to be solved [13,14]. In native myocardial tissue, cells are considerably dense (Fig. 1A) in comparison with other tissues including cartilage, vascular, and heart valve, which are cell-sparse tissues and have been successfully engineered by using biodegradable scaffolds (Fig. 1B). Cardiomyocytes are also tightly interconnected with gap junctions, which mediated the reciprocal exchange of small molecules and ions resulting in electrically synchronous beating [17]. In myocardial tissue engineering, biodegradable scaffolds themselves attenuate cell-to-cell connections and scaffold biodegradation leads to fibrous tissues containing excessive amount of ECM, which is shown in pathological states including ischemic heart disease or dilated cardiomyopathy. Investigators are now trying to fabricate more porous structure of biodegradable scaffolds and to develop new techniques seeding more cells into the scaffolds. In particular, structural balance between cells and ECM should be controlled to fabricate native heart-like tissues. By contrast, we now propose novel tissue engineering methodology that is to construct 3-D functional tissues by layering 2-D cell sheets without any biodegradable alternatives for ECM. To obtain viable cell sheets, we have exploited intelligent culture surfaces, from which cultured cells detach as a cell sheet simply by reducing temperature. In this paper, we present the new technology "cell sheet engineering" and its application to myocardial tissue reconstruction.
2. Temperature-responsive culture surfaces
Temperature-responsive culture surfaces were developed among the research to control cell adhesion to biomaterials. Cells adhere to culture surfaces via membrane receptors and cell adhesive proteins, including fibronectin, that reside in serum or are secreted from the cells in culture (Fig. 2A). The interaction between adhesive proteins and culture surfaces depends on the wettability of the surface. Normal tissue culture polystyrene (TCPS) dishes are hydrophobic and absorb ECM proteins resulting in cell attachment and proliferation. To harvest cells from the surfaces, enzymatic
digestion including trypsin and dispase are usually utilized. In that case, both adhesive proteins and membrane receptors are disrupted, then cells detach with considerable damages (Fig. 2B). On the other hand, we graft temperature-responsive polymer, poly(N-isopropylacrylamide)(PIPAAm) to TCPS dishes covalently by electron beam. The surfaces are hydrophobic and cells adhere and proliferate under culture condition at 37~ By lowering temperature below 32~ the surfaces change reversibly to hydrophilic and not cell adhesive due to rapid hydration and swelling of the grafted PIPAAm. This unique surface change allows cultured cells to detach spontaneously from these grafted surfaces simply by lowering temperature [18]. As against using enzymatic digestion, only the interaction between adhesive proteins and material surfaces is released and cells detach together with intact membrane proteins and adhesive proteins (Fig. 2C) [19]. As a result, cells recovered by using PIPAAm-grafted surfaces maintain their differentiated functions more strongly than the cells recovered by protease digestion [20]. For example, trypsin-treated hepatocytes decrease albumin production, on the other hand, those cells harvested from PIPAAm-grafted surface preserve albumin secretion [21]. In addition to the passive mechanism of the surface change from hydrophobic to hydrophilic, cell-mediated active processes have been ascertained as cell detachment mechanisms [22]. Sodium azide, an ATP synthesis inhibitor, considerably retarded cell release from PIPAAm-grafted surfaces, indicating that energy-dependent metabolic process is one of major mechanisms. The active processes are also mediated by intracellular signal transduction, including tyrosine phosphorylation and cytoskeltal reorganization and lead to the cell morphological change from spread to round after surface property change [23].
3. Cell sheet engineering
When cells are cultured confluently, they connect to each other via cell-to-cell junction proteins and ECM (Fig. 3A). With enzymatic digestions, these proteins are disrupted and each cell is released separately (Fig. 3B). In the case using PIPAAm-grafted surfaces, cell-to-cell connections are not disrupted and cells are harvested as a contiguous cell sheet by decreasing temperature (Fig. 3C). Furthermore, adhesive proteins underneath cell sheets are also maintained and they play a desirable role as an adhesive agent in transferring cell sheets onto other culture materials or other cell sheets [24]. These viable cell sheets are composed of cells and biological ECM without any artificial scaffolds. Various types of cell sheets have been successfully lifted up and transferred on other surfaces [25-32].
The Biomaterials Silver Jubilee Compendium T. Shimizu et al. / Biomaterials 24 (2003) 2309-2316
213 2311
Fig. 1. Histological comparison between cell-dense and cell-sparse tissues. Hematoxilin and eosin staining shows that cells are dense and tightly connected in myocardial tissue (A) On the other hand, cartilage tissue includes sparse cells and large amount of ECM (B).
Fig. 2. Cell harvest mechanism by using temperature-responsive culture surfaces. (A) Cells attach to hydrophobic culture surfaces via cell membrane proteins and ECM, which reside in serum or are secreted from the cells. (B) When enzymatic digestion is used, both membrane and ECM proteins are disrupted, resulting in cell detachment. (C) When cells are cultured on temperature-responsive culture surfaces, the interconnection between ECM and hydrophilic culture surfaces is released only by lowering temperature. Then the cells detach together with intact proteins.
Fig. 3. Cell sheet release from temperature-responsive culture surfaces. (A) When cells are cultured confluently, the cells connect to each other via cell-to-cell junction proteins. (B) When harvested by protease treatments, cell-to-cell connections are disrupted and cells are released separately. (C) When PIPAAm-grafted surfaces are used, cell-to-cell connections are completely preserved and the cells are released as a contiguous cell sheet. ECM retained underneath the cell sheets play a role as a adhesive agent.
214
2312
The B i o m a t e r i a l s S i l v e r J u b i l e e C o m p e n d i u m T. Shimizu et al. /Biomaterials 24 (2003) 2309-2316
As cell sheet manipulation, two techniques have been performed according to cell types and objects. One is to manipulate cell sheets directly with forceps or pipetting after the sheets are completely harvested resulting in proportionally shrunk and thicker constructs due to active cytoskeletal reorganization. As indicated by synchronized beating of shrunk cardiomyocyte sheets, cell-to-cell connections are preserved after this procedure [31]. The other is to use support membranes including a hydrophilically modified poly(vinylidene difluoride)(PVDF) membrane for preserving cell sheet morphology without any shrinkage. Before cell sheets release, support membranes are placed over the confluent cells. Then the cell sheets physically attached to the support membranes are harvested from PIPAAmgrafted surfaces below 32~ and transferred onto other surfaces. Incubation at 37~ causes reattachment of the cell sheets to new surfaces via remaining adhesive proteins. Finally, only the support membranes are removed. The latter technique has realized the cell sheet manipulation preserving their structure and function [26-30]. These cell sheet manipulation techniques without using any biodegradable scaffolds have been applied to tissue engineering in three types of contexts (Fig. 4). First is transplanting single cell sheet for skin and
cornea reconstruction. Advantages of skin epithelial cell sheets harvested by using PIPAAm-grafted surfaces have been confirmed in comparison with those harvested by dispase treatments. E-cadherin, which is an essential protein for skin cell-to-cell junctions, and laminin 5, which is a major component of epithelial basement membranes, were retained in skin cell sheets released from PIPAAm-grafted surfaces [27]. It should attenuate the risk of infection after artificial skin transplantation. Second is to layer same cell sheets for reconstructing homogeneous tissues including myocardium. Third is to layer several types of cell sheets for fabricating laminar structures including liver, kidney and vascular. Layered co-culture comprising a hepatocyte sheet and an endothelial cell sheet has revealed the differentiated cell shape and extensive albumin expression of hepatocytes, which have never been seen in hepatocyte mono-culture [32]. We have been now applied these technologies "cell sheet engineering" to reconstructing various types of tissues. Among them, myocardial tissue engineering based on the second context is described below.
4. Myocardial tissue reconstruction by layering cardiomyocyte sheets [28,30,31] Cardiomyocytes are tightly interconnected with gap junctions and pulsate simultaneously in native heart tissue. It is also well-known that confluent cultured cardiomyocytes on culture surfaces connect via gap junctions and beat simultaneously [33]. Therefore, in myocardial tissue engineering by layering cell sheets, it is a crucial point whether electrical and morphological communications are established between bilayer cell sheets. Chick embryo or neonatal rat cardiomyocyte sheets released from PIPAAm-grafted surfaces presented synchronized pulsation. To examine the electrical communication, two cardiomyocyte sheets were overlaid partially as schematically illustrated in Fig. 5. Two
Fig. 4. Three contexts in cell sheet engineering. (A) Single cell sheet is useful for skin or cornea transplantation. (B) Same cell sheets are layered to reconstruct homogeneous 3-D tissues including myocardium. (C) Several types of cell sheets are co-layered to fabricate laminar structures including liver and kidney.
Fig. 5. Schematic illustration of electrical analysis of layered cardiomyocyte sheets. To examine the electrical synchronization, two cardiomyocyte sheets (A, B) are overlaid partially. Two electrodes are set over monolayer parts of both cell sheets to detect the electrical potentials separately.
215
The Biomaterials Silver Jubilee Compendium
2313
T. Shimizu et al. / Biomaterials 24 (2003) 2309-2316
electrodes were set over monolayer parts of both cell sheets. Detected electrical potentials of the two sheets completely synchronized (Fig. 6). Furthermore, electrical stimulation to the single-layer region of one sheet was transmitted to the other cell sheet and the two cell sheets pulsated simultaneously. Histological analysis showed that bilayer cardiomyocyte sheets contacted intimately resulting in homogeneous tissue. Cell-to-cell connections including desmosomes and intercalated disks were confirmed by transmission electron microscopic images. These data indicate that electrical and morphological communications are established between layered cardiomyocyte sheets. Under conventional culture conditions, cardiac myocytes are fixed to rigid material surfaces and their motion is highly limited. To minimize the interaction between cell sheets and culture materials, the sheets were overlaid on several types of materials including polyethylene meshes, elastic polyurethane meshes or framelike collagen membranes. In any cases, the constructs pulsated simultaneously with higher amplitude than the cells fixed on rigid culture surfaces. When cardiomyocyte sheets were layered on frame-like collagen mem-
branes, the center part of them is free from any culture materials. In result, 4-layer cardiac constructs on the frame-like collagen membranes pulsated spontaneously in macroscopic view. To examine in vivo survival and function of layered cardiomyocyte sheets, the constructs were transplanted into dorsal subcutaneous tissues of nude rats. Surface electrograms originating from transplanted constructs were detected independently from host electrocardiograms, in the earliest case, at 2 weeks after the operation (Fig. 7). When transplantation sites were opened, macroscopic simultaneous graft beatings were observed at the earliest period, 3 days after the transplantation. Furthermore, graft survival was confirmed at least up to 1 year. Morphological analysis demonstrated that neovascularizations occurred in a few days and that vascular network was organized within a week (Fig. 8A). Cross-sectional views revealed stratified celldense myocardial tissues (Fig. 8B), well-differentiated sarcomeres and diffuse formation of gap junctions. In comparison between 2-layer and 4-layer cardiac tissue grafts, fractional shortening increased depending on the number of layered cell sheets. Thus, the basic technology has been established to fabricate electrically communicative, pulsatile myocardial tissues by using cell sheets both in vitro and in vivo.
Sheet A
5. Future perspectives Sheet B
1 sec
Fig. 6. Synchronization of layered cardiomyocyte sheets. Representative tracings of electrical potentials of sheet A and sheet B show complete synchronization.
Recently, research on myocardial tissue engineering has been accelerated to develop further advanced therapy for severe heart failure. Transplantation of layered cardiomyocyte sheets on the myocardial scar may be more beneficial than that of bioengineered heart tissue including biodegradable scaffolds in the point of scaffold-mediated disadvantages. However, there are several common problems in myocardial tissue
1 sec
Fig. 7. Skin surface electrogram of transplanted cardiomyocyte sheets. Representative tracings of the host electrocardiogram (upper) and the electrical potential detected via the electrode set at the skin just above the transplanted heart graft (lower) are shown. Skin surface electrogram originating from the graft is detected independently from host electrocardiogram.
216 2314
The Biomaterials Silver Jubilee Compendium T. Shimizu et al. / Biomaterials 24 (2003) 2309-2316
Fig. 8. (A) Macroscopic view of the transplanted cardiac graft. Multiple neovascularization is shown in the square-designed cardiac graft transplanted into dorsal subcutaneous tissue. (B) Azan staining shows a stratified cardiac tissue graft including elongated cardiomyocytes and microvasculars (arrows).
Fig. 9. Schematic illustration of myocardial reconstruction based on cell sheet engineering. We now propose the application of "cell sheet engineering" to myocardial tissue reconstruction. Cell sourcing remains a crucial problem. Neovascularization for oxygen and nutrition supply is also critical to fabricate human applicable myocardial tissue. Growth factors, gene delivery and the utility of gene-modified cells or endothelial cells may be helpful. Mechanical load by using bioreactors should strengthen the engineered myocardial tissues. Transplantation of engineered tissue into myocardial infarction model is now in progress.
engineering. As described in Section 1, myocardial cell sourcing remains a crucial problem. Further advance in stem cell biology for cardiomyocytes will be needed to realize clinical application of bioengineered myocardial tissues. Vascular reconstruction is also one of the most critical issues in myocardial tissue engineering. Sufficient supply of oxygen and nutrition is required for functionally beating heart tissue. It has been reported that cells are dense in the graft periphery, but sparse in the interior part due to insufficient oxygen perfusion in scaffoldbased heart tissue grafts [34]. Although, in our studies, multiple neovascularization arose in transplanted cardiac grafts in a few days, primary insufficient oxygen and nutrition permeation also limit the number of transplanted cardiomyocyte sheets. Hence, new methods to accelerate blood vessel formation are now requested to engineer larger or thicker constructs for heart tissue
repair. As examined in isolated cell injection, genemodified cells may be also applicable for engineering more vascularized heart tissues [35]. Using cell sheet technology, it has been reported that a single layer of endothelial cell sheet enhances the capillary formation in vivo [36]. Therefore, heterogenous layering of endothelial cell sheets between cardiomyocyte sheets may promote neovascularization. Further research and development will be needed to engineer vascular networks sufficient for fabricating clinically applicable heart tissues. In native heart, cardiomyocytes are gradually elongated and hypertrophied by mechanical load increase in accordance with the growth of the body. Therefore, some investigators have attempted to strengthen bioengineered heart tissues by using mechanical devices. Carrier et al. used a rotating bioreactor for culturing cardiomyocytes on PGA scaffolds [37]. Fink et al.
217
The Biomaterials Silver Jubilee Compendium T. Shimizu et al. / Biomaterials 24 (2003) 2309-2316
clearly demonstrated that application of stretch devices to engineering heart tissues strengthened the contraction power and oriented the cells unidirectionally [38]. We are now trying to stretch layered cardiomyocyte sheets to fabricate more powerful cardiac constructs in vitro. Finally, our concept of myocardial tissue engineering is schematically illustrated in Fig. 9. Although further interdisciplinary research will be needed to clear the existing several problems, cell sheet engineering should have enormous potential for constructing clinically applicable heart grafts and should promote tissue engineering research fields.
[12]
[13]
[14]
[15]
Acknowledgements [16]
The present work was supported by the Japan Society for the Promotion of Science, Grant-in-Aid for Scientific Research (A) (13308055) and Grant-in-Aid for Encouragement of Young Scientists (13780693). It was also supported in part by the Open Research Grant from the Japan Research Promotion Society for Cardiovascular Diseases.
[17]
[18]
[19]
References [1] Fuchs JR, Nasseri BA, Vacanti JP. Tissue engineering: a 21st century solution to surgical reconstruction. Ann Thorac Surg 2001;72:577-91. [2] Mann BK, West JL. Tissue engineering in the cardiovascular system: progress toward a tissue engineered heart. Anat Rec 2001;263:367-71. [3] Reinlib L, Field L. Cell transplantation as future therapy for cardiovascular disease. A workshop of the National Heart Lung and Blood Institute. Circulation 2000;101:E182-7. [4] Menasche P, Hagege AA, Scorsin M, Pouzet B, Desnos M, Duboc D, Schwartz K, Vilquin JT, Marolleau JP. Myoblast transplantation for heart failure. Lancet 2001;357:279-80. [5] Kehat I, Kenyagin-Karsenti D, Snir M, Segev H, Amit M, Gepstein A, Livne E, Binah O, Itskovitz-Eldor J, Gepstein L. Human embryonic stem cells can differentiate into myocytes with structural and functional properties of cardiomyocytes. J Clin Invest 2001;108:407-14. [6] Min JY, Yang Y, Converso KL, Liu L, Huang Q, Morgan JP, Xiao YF. Transplantation of embryonic stem cells improves cardiac function in postinfarcted rats. J Appl Physiol 2002;92:288-96. [7] Makino S, Fukuda K, Miyoshi S, Konishi F, Kodama H, Pan J, Sano M, Takahashi T, Hori S, Abe H, Hata J, Umezawa A, Ogawa S. Cardiomyocytes can be generated from marrow stromal cells in vitro. J Clin Invest 1999;103:697-705. [8] Orlic D, Kajstura J, Chimenti S, Jakoniuk I, Anderson SM, Li B, Pickel J, McKay R, Nadal-Ginard B, Bodine DM, Leri A, Anversa P. Bone marrow cells regenerate infarcted myocardium. Nature 2001;410:701-5. [9] Akins RE. Can tissue engineering mend broken hearts? Circ Res 2002;90:120-2. [10] Langer R, Vacanti JP. Tissue engineering. Science 1993;260:920-6. [11] Papadaki M, Bursac N, Langer R, Merok J, Vunjak-Novakovic G, Freed LE. Tissue engineering of functional cardiac muscle:
[20]
[21]
[22]
[23]
[24]
[25]
[26]
[27]
2315
molecular, structural, and electrophysiological studies. Am J Physiol Heart Circ Physiol 2001 ;280:H 168-78. Li RK, Jia ZQ, Weisel RD, Mickle DA, Choi A, Yau TM. Survival and function of bioengineered cardiac grafts. Circulation 1999;100:II63-9. Sakai T, Li RK, Weisel RD, Mickle DA, Kim ET, Jia ZQ, Yau TM. The fate of a tissue-engineered cardiac graft in the right ventricular outflow tract of the rat. J Thorac Cardiovasc Surg 2001;121:932-42. Leor J, Aboulafia-Etzion S, Dar A, Shapiro L, Barbash IM, Battler A, Granot Y, Cohen S. Bioengineered cardiac grafts: a new approach to repair the infarcted myocardium. Circulation 2000; 102:III56-61. Eschenhagen T, Fink C, Remmers U, Scholz H, Wattchow J, Weil J, Zimmermann W, Dohmen HH, Schafer H, Bishopric N, Wakatsuki T, Elson EL. Three-dimensional reconstitution of embryonic cardiomyocytes in a collagen matrix: a new heart muscle model system. FASEB J 1997;11:683-94. Zimmermann WH, Schneiderbanger K, Schubert P, Didie M, Munzel F, Heubach JF, Kostin S, Neuhuber WL, Eschenhagen T. Tissue engineering of a differentiated cardiac muscle construct. Circ Res 2002;90:223-30. Luque EA, Veenstra RD, Beyer EC, Lemanski LF. Localization and distribution of gap junctions in normal and cardiomyopathic hamster heart. J Morphol 1994;222:203-13. Yamada N, Okano T, Sakai H, Karikusa F, Sawasaki Y, Sakurai Y. Thermo-responsive polymeric surfaces; control of attachment and detachment of cultured cells. Makromol Chem Rapid Commun 1990;11:571-6. Yamato M, Konno C, Kushida A, Hirose M, Utsumi M, Kikuchi A, Okano T. Release of adsorbed fibronectin from temperatureresponsive culture surfaces requires cellular activity. Biomaterials 2000;21:981-6. Nakajima K, Honda S, Nakamura Y, Lopez-Redondo F, Kohsaka S, Yamato M, Kikuchi A, Okano T. Intact microglia are cultured and non-invasively harvested without pathological activation using a novel cultured cell recovery method. Biomaterials 2001;22:1213-23. Okano T, Yamada N, Sakai H, Sakurai Y. A novel recovery system for cultured cells using plasma-treated polystyrene dishes grafted with poly(N-isopropylacrylamide). J Biomed Mater Res 1993;27:1243-51. Okano T, Yamada N, Okuhara M, Sakai H, Sakurai Y. Mechanism of cell detachment from temperature-modulated, hydrophilic-hydrophobic polymer surfaces. Biomaterials 1995;16:297-303. Yamato M, Okuhara M, Karikusa F, Kikuchi A, Sakurai Y, Okano T. Signal transduction and cytoskeletal reorganization are required for cell detachment from cell culture surfaces grafted with a temperature-responsive polymer. J Biomed Mater Res 1999;44:44-52. Kushida A, Yamato M, Konno C, Kikuchi A, Sakurai Y, Okano T. Temperature-responsive culture dishes allow nonenzymatic harvest of differentiated Madin-Darby canine kidney (MDCK) cell sheets. J Biomed Mater Res 2000;51:216-23. Kikuchi A, Okuhara M, Karikusa F, Sakurai Y, Okano T. Twodimensional manipulation of confluently cultured vascular endothelial cells using temperature-responsive poly(N-isopropylacrylamide)-grafted surfaces. J Biomater Sci Polym Ed 1998;9:1331-48. Hirose M, Kwon OH, Yamato M, Kikuchi A, Okano T. Creation of designed shape cell sheets that are noninvasively harvested and moved onto anothersurfaces. Biomacromolecules 2000;1: 377-81. Yamato M, Utsumi M, Kushida A, Konno C, Kikuchi A, Okano T. Thermo-responsive culture dishes allow the intact harvest of
218 2316
[28]
[29]
[30]
[31]
[32]
The Biomaterials Silver Jubilee Compendium T. Shimizu et al. / Biomaterials 24 (2003) 2309-2316
multilayered keratinocyte sheets without dispase by reducing temperature. Tissue Eng 2001;7:473-80. Shimizu T, Yamato M, Kikuchi A, Okano T. Two-dimensional manipulation of cardiac myocyte sheets utilizing temperatureresponsive culture dishes augments the pulsatile amplitude. Tissue Eng 2001;7:141-51. Kushida A, Yamato M, Kikuchi A, Okano T. Two-dimensional manipulation of differentiated Madin-Darby canine kidney (MDCK) cell sheets: the noninvasive harvest from temperatureresponsive culture dishes and transfer to other surfaces. J Biomed Mater Res 2001;54:37-46. Shimizu T, Yamato M, Akutsu T, Shibata T, Isoi Y, Kikuchi A, Umezu M, Okano T. Electrically communicating three-dimensional cardiac tissue mimic fabricated by layered cultured cardiomyocyte sheets. J Biomed Mater Res 2002;60:110-7. Shimizu T, Yamato M, Isoi Y, Akutsu T, Setomaru T, Abe K, Kikuchi A, Umezu M, Okano T. Fabrication of pulsatile cardiac tissue grafts using a novel 3-dimensional cell sheet manipulation technique and temperature-responsive cell culture surfaces. Circ Res 2002;90:e40-8. Harimoto M, Yamato M, Hirose M, Takahashi C, Isoi Y, Kikuchi A, Okano T. Novel approach for achieving doublelayered cell sheets co-culture: overlaying endothelial cell sheets onto monolayer hepatocytes utilizing temperature-responsive culture dishes. J Biomed Mater Res 2002;62:464-70.
[33] Oyamada M, Kimura H, Oyamada Y, Miyamoto A, Ohshika H, Mori M. The expression, phosphorylation, and localization of connexin 43 and gap-junctional intercellular communication during the establishment of a synchronized contraction of cultured neonatal rat cardiac myocytes. Exp Cell Res 1994;212:351-8. [34] Bursac N, Papadaki M, Cohen RJ, Schoen FJ. Eisenberg SR, Carrier R, Vunjak-Novakovic G, Freed LE. Cardiac muscle tissue engineering: toward an in vitro model for electrophysiological studies. Am J Physiol 1999;277(2 Part 2):H433-44. [35] Suzuki K, Murtuza B, Smolenski RT, Sammut IA, Suzuki N, Kaneda Y, Yacoub MH. Cell transplantation for the treatment of acute myocardial infarction using vascular endothelial growth factor-expressing skeletal myoblasts. Circulation 2001;104: I207-12. [36] Soejima K, Negishi N, Nozaki M, Sasaki K. Effect of cultured endothelial cells on angiogenesis in vivo. Plast Reconstr Surg 1998;101:1552-60. [37] Carrier RL, Papadaki M, Rupnick M, Schoen FJ, Bursac N, Langer R, Freed LE, Vunjak-Novakovic G. Cardiac tissue engineering: cell seeding, cultivation parameters, and tissue construct characterization. Biotechnol Bioeng 1999:64:580-9. [38] Fink C, Ergun S, Kralisch D, Remmers U, Weil J, Eschenhagen T. Chronic stretch of engineered heart tissue induces hypertrophy and functional improvement. FASEB J 2000;14:669-79.
219
The Biomaterials Silver Jubilee Compendium Available online at www.sciencedirect.com
8CIENCB~DIRECT e
ELSEVIER
Biomateriais
Biomaterials 25 (2004) 5681-5703 www.elsevier.com/locate/biomaterials
Review
Biomaterial-associated thrombosis" roles of coagulation factors, complement, platelets and leukocytes Maud B. Gorbet, Michael V. Sefton* Department of Chemical Engineering and Applied Chemistry, Institute of Biomaterials and Biomedical Engineering, University of Toronto, 4 Taddle Creek Road, Room 407D, Toronto, Ont., Canada M5S 3G9 Received 3 September 2003; accepted 19 January 2004
Abstract
Our failure to produce truly non-thrombogenic materials may reflect a failure to fully understand the mechanisms of biomaterialassociated thrombosis. The community has focused on minimizing coagulation or minimizing platelet adhesion and activation. We have infrequently considered the interactions between the two although we are generally familiar with these interactions. However, we have rarely considered in the context of biomaterial-associated thrombosis the other major players in blood: complement and leukocytes. Biomaterials are known agonists of complement and leukocyte activation, but this is frequently studied only in the context of inflammation. For us, thrombosis is a special case of inflammation. Here we summarize current perspectives on all four of these components in thrombosis and with biomaterials and cardiovascular devices. We also briefly highlight a few features of biomaterial-associated thrombosis that are not often considered in the biomaterials literature: 9 9 9 9
The importance of tissue factor and the extrinsic coagulation system. Complement activation as a prelude to platelet activation and its role in thrombosis. The role of leukocytes in thrombin formation. The differing time scales of these contributions.
9 2004 Elsevier Ltd. All rights reserved. Keywords." Leukocytes; Tissue Factor; CD1 l b; Platelets; Biomaterials; Complement activation; Thrombosis; Coagulation
Contents
1.
Introduction
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
2.
Coagulation cascade . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.1. The intrinsic pathway . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2. The extrinsic pathway . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.3. Physiologic inhibitors of coagulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.4. Biomaterials and coagulation pathways . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.5. Anticoagulants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
5683 5683 5684 5684 5685 5686
3.
Complement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.1. Classical pathway . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2. Alternative pathway . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.3. Regulatory molecules of complement activation . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.4. Interactions of complement and coagulation cascade . . . . . . . . . . . . . . . . . . . . . . . . . 3.5. Complement activation and biomaterials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
5686 5686 5687 5688 5688 5688
*Corresponding author. Tel.: + 1-416-978-3088; fax: + 1-416-978-4317. E-mail address."
[email protected] (M.V. Sefton). 0142-9612/$- see front matter 9 2004 Elsevier Ltd. All rights reserved. doi: 10.1016/j.biomaterials.2004.01.023
5682
220
The Biomaterials Silver Jubilee Compendium
5682
M.B. Gorbet, M.V. Sefton / Biomaterials 25 (2004) 5681-5703
4. Platelets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.1. Platelet biology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.2. Platelets and biomaterials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
5689 5689 5690
5. Leukocytes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.1. Leukocyte biology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.2. Leukocyte activation and biomaterials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.3. Leukocytes, platelets and coagulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
5692 5692 5693 5694
6. Other important factors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.1. Flow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.2. Endotoxin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
5694 5694 5695
7. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
5696
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
5697
I. Introduction
Biocompatibility is defined as "the ability of a material to perform with an appropriate host response in a specific application". Biocompatibility of blood contacting devices relates mainly to the thrombotic response induced by the materials. Although no material has been found truly biocompatible, many cardiovascular devices function with low or acceptable risks of complications [1]. Hemolytic, toxic and immunologic reactions have usually been dealt with earlier in the development of a material to be used for cardiovascular devices and are rarely an issue with their use; an exception may be immunological reactions to tissue engineered constructs. However, thrombotic and thromboembolic complications, as well as bleeding risks associated with the necessary anticoagulant therapy remain of serious concern with cardiovascular devices. Clinical manifestations of the bioincompatibility of cardiovascular devices are numerous: sudden and complete obstruction of stents within weeks [2]; acute and subacute thrombotic occlusion in medium sized grafts (4-6 mm) [3]; embolic complications with artificial hearts [4], catheters [5] and prosthetic valves [6,7]; thrombotic complications during cardiopulmonary bypass [3] and angioplasty [8]. Larger vascular grafts remain thrombogenic for many years, but fewer thrombotic complications are observed as high flows disperse the clotting factors. However, occasional embolic episodes occur as high flows dislodge the thrombotic deposits. Even if the risk of thrombotic complication appears to be low (varying between 2% and 10% depending on the device), they may have fatal outcomes and the cost associated with the follow-up intervention is not negligible. Furthermore, these thrombotic complications with cardiovascular devices occur despite the use of antiplatelet and anticoagulant therapies reinforcing the inherent thrombogenicity of the materials. Material thrombogenicity is further illustrated by the acute failure of small diameter vascular
grafts despite the strong anticoagulant regimen. Many years of intensive research on biomaterials have not yet produced a material, which has proven suitable for this last application. To improve the blood compatibility of cardiovascular devices, surface modifications, such as attachment of antithrombotic agents or the immobilization of polyethylene oxide (PEO) have been considered but their success has been limited. Treating surfaces with PEO to reduce protein adsorption and prevent platelet adhesion has remained unproven in terms of thrombogenicity. Different heparin coatings have been developed and some have actually been able to reach the commercial stage in cardiopulmonary bypass [9] and in coronary stents. However, reports on the improvement of in vivo biocompatibility have been mixed [10-16]. Heparinized cardiopulmonary bypass circuits appear to partially reduce the inflammatory response associated with cardiopulmonary bypass [17,18]. But to date, heparin coatings have not yet been shown to significantly reduce the number of postoperative complications, improve patient outcome, or reduce hospital stay [16,19,20]. This illustrates another limit of our understanding of bloodmaterial interactions: we do not know how much of an inflammatory and thrombosis response is tolerable or whether any of the changes in normal hemostasis induced by the device result in harmful consequences. Since biomaterial strategies have not resolved the problem, different pharmacological approaches are being investigated. Complement inhibition with the use of sCR1 [21] or anti-C5a antibody [22], serine protease inhibitors such as aprotinin [23,24], platelet receptor antagonists such anti-GPIIb/IIIa [25] and cytokine antibody [26] are newer approaches to reduce thrombotic complications with cardiopulmonary bypass. While the results are promising--a partial reduction of the inflammatory or the thrombotic response to cardiovascular devices--it is as yet not possible to make conclusions with respect to the overall improvement of device biocompatibility. Unfortunately, most agents
221
The Biomaterials Silver Jubilee Compendium
5683
M . B . Gorbet, M . V . S e f t o n / B i o m a t e r i a l s 25 ( 2 0 0 4 ) 5 6 8 1 - 5 7 0 3
(Fig. 2). Thrombin is formed following a cascade of reactions where an inactive factor becomes enzymatically active following surface contact or after proteolytic cleavage by other enzymes; the newly activated enzyme then activates another inactive precursor factor. Initiation of clotting occurs either by surface-mediated reactions, or through tissue factor (TF) expression by cells. The two systems converge into the common pathway resulting in the formation of fibrin clot upon action of thrombin on fibrinogen. Then, Factor XIII, activated by thrombin, crosslinks and stabilizes the fibrin clot into an insoluble fibrin gel. Fig. 1. Overview of blood-material interactions showing the components relevant to thrombosis. While complement and leukocytes are normally considered under "inflammation" we consider them as participants in thrombosis along with platelets and coagulation.
affect only one of the players in the blood compatibility response and this is not likely to be sufficient to result in clinical benefits. On the other hand, these inhibitors and antibodies provide valuable information on the mechanisms involved in the thrombotic complications associated with cardiovascular devices. Under normal conditions, blood contacts an endothelium with anticoagulant and antithrombotic properties. The use of a cardiovascular device represents the introduction of a foreign surface in the circulation, without the properties of the endothelium. Bloodmaterial interactions trigger a complex series of events including protein adsorption, platelet and leukocyte activation/adhesion, and the activation of complement and coagulation; they are highly interlinked (Fig. 1). This review outlines the current state of understanding of these phenomena with particular reference to the biomaterial or cardiovascular device as an agonist of these thrombotic reactions. This review is not, however, a catalog of surface modification or biomaterial design strategies that have been used in an attempt to control this aspect of the host response. Rather the focus is on the mechanism of the thrombotic response to a biomaterial, how each component is thought to interact with a biomaterial and how these may interact to produce the observed thrombosis. We are particularly interested in the role of complement and leukocytes, which are not often considered by the biomaterials community to be contributors to thrombosis. Our approach is to treat thrombosis as a special case of inflammation.
2. Coagulation cascade Blood coagulation involves a series of proteolytic reactions resulting in the formation of a fibrin clot
2.1. The intrinsic pathway The classic picture (Fig. 2) shows the intrinsic pathway being initiated by contact activation of high molecular weight kininogen (HMWK), prekallikrein and Factor XII: it is commonly said that these molecules require contact with (negatively charged) surfaces for zymogen activation in vitro [27]. Factor XII is activated by adsorption, FXIIa converts prekallikrein into kallikrein and with H M W K as a cofactor activates Factor XI to Factor FXIa. Factor XIa activates Factor IX to Factor IXa. Following a cascade of reactions involving among others the intrinsic tenase complex (Factor IXaFactor VIIIa), prothrombin is cleaved into thrombin. The importance of the intrinsic pathway to normal blood coagulation remains speculative, as the occurrence of negatively charged surfaces in viva is limited.
EXTRINSIC SYSTEM
INTRINSIC SYSTEM Surface Contact
Factor :X]:I~--~:X~ o Factor XI ~-~--~XIo Factor
~ FactorV]~
',~,,,%/..Je_.le_.ts_ __ _ I . . . . Factor X
COMMON PATHWAY Prothrombin
- Factor
YJIo --~
IXo++ IX- C!++= i~/"a
F••lat ~Xo
Co~
Tissue/
-z.
,,
~247 ctor V
elets
Fibrinogen (monomer)
Factor X F~zto r XI]]
Thrombin
L
l(lllo Fibrifl (polymer)
- Fibrin
(slable polymer
)
Fig. 2. Simplified blood-coagulation cascade as viewed by biomaterial textbooks [117] and older hematology books. The intrinsic system, starting with Factor XII, is shown as a linear cascade of zymogen activation steps in parallel with the extrinsic system that involves TF. TF and platelets are shown as 'cofactors' of the process (similar to Ca +2) rather than as central participants. The intrinsic and extrinsic systems converge on the common pathway to produce thrombin and fibrin [used with copyright approval from Elsevier/Academic Press].
222
The Biomaterials Silver Jubilee Compendium
5684
M.B. Gorbet, M.V. Sefton / Biomaterials 25 (2004) 5681-5703
Collagen present in the subendothelium after vessel injury could be the surface required for this reaction [28]. Under physiologic conditions, the lack of relevance of the contact activation system is consistent with the fact that deficiencies of the contact proteins, HMWK, prekallikrein and Factor XII, have not been associated with abnormal bleeding [1,27,29].
2.2. The extrinsic pathway The hematology textbooks center their view of the blood-coagulation cascade, on the TF-dependent pathway (Fig. 3); this perspective has replaced the older perspective that is still used in biomaterials textbooks. The physiological initiator of coagulation is TF, which is expressed on damaged cells at the site of vascular injury. Plasma Factor VII (FVII) binds to TF on the cell membranes and requires activation to FVIIa to form the extrinsic tenase complex: TF-VIIa complex. FVII is activated by trace amounts of thrombin, FIXa, FXa and TF-VIIa complex, the latter being more likely physiologically relevant as small amounts of TF-FVIIa are present extravascularly in vivo [30,31]. Picomolar concentrations of FVIIa circulate normally in blood and are also thought to serve as a primer in the initiation of the coagulation cascade by allowing direct formation of TF-FVIIa complex upon TF exposure [32]. TF-FVIIa complex on cell membranes cleaves Factor X into Factor Xa in the presence of calcium. The prothrombinase complex can then assemble on the membrane and generate thrombin (common pathway of the coagulation cascade). FX is not the only physiological substrate of the TF-FVIIa complex. The TF-FVIIa complex also activates FIX [33].
Extrinsic tenase . ,,.
,~.~
"/
TF § VII
~~176 TF-VII
~.~c,
II . . . . . .
. . . . . . .
PROTHROMBIN Fig. 3. Revised (simplified) blood-coagulation cascade as given in a standard hematology textbook [30]. Unlike Fig. 2, there is no intrinsic system or Factor XII. The cascade is not linear with several feedback loops and most importantly it begins with TF and the tenase complex. Platelet involvement is still shown as a 'cofactor' and the reactions past thrombin are not shown [used with copyright approval from McGrawHill].
The extrinsic and intrinsic pathways are not independent of each other. When coagulation is initiated by a TF-dependent pathway, the intrinsic tenase remains important, since production of FXa by FIXa-FVIIIa complex has been shown to significantly contribute to thrombin generation [34]. It appears that extrinsic tenase TF-FVIIa is responsible for the onset of coagulation while the intrinsic tenase is the major player in the propagation phase [35]. The activation of FX by FIXa is all the more important since tissue factor pathway inhibitor (TFPI) will reduce the production of FXa by TF-VIIa complex [30,31].
2.3. Physiologic inhibitors of coagulation Most activated coagulation factors are serine proteases. Plasma contains several protease inhibitors, such as C~l-protease inhibitor, c~2-macroglobulin, heparin cofactor II and antithrombin III, to modulate and inhibit their activity. Antithrombin III is the most important [36]: it neutralizes its target enzymes, FXa and thrombin, by forming a complex with the enzyme in which the enzyme's active site is blocked. In the absence of heparin, complex formation occurs at a relatively slow rate. However, in the presence of the polysaccharide heparin or naturally occurring heparan sulfate (on the endothelium), inhibition rates rise significantly. Antithrombin is also able to inhibit FIXa, FXIa and FXIIa [37]. While TF-VIIa is not efficiently inhibited by antithrombin, it has its own inhibitor: the TFPI. Plasma TFPI is not the major intravascular pool of TFPI. A larger pool exists on the luminal surface of the vascular endothelium, which is released by a bolus injection of heparin. Platelets also carry 10% of the total TFPI in blood and release their TFPI following stimulation by thrombin and other agonists [38]. Activated monocytes may also release TFPI [39]. TFPI has two inhibitory sites, one for FXa and one for TF-VIIa complex. Inactivation of FXa through binding to TFPI in solution is required for inhibition of the complex TF-VIIa. The initial binding of FXa to TFPI is relatively slow [40] and may not be able to prevent thrombosis when TF-FVIIa complex are being formed at a high rate [31]. Leukocyte elastase also cleaves TFPI and impairs its ability to inhibit both FXa and TF-VIIa complex. Upon exposure of a complex FXa/TFPI/TFVIIa to elastase, FXa and TF-VIIa activities are restored [41]. The endothelium also participates in the regulation of thrombosis via thrombomodulin. Thrombin binds to thrombomodulin on endothelial cells and this complex activates protein C [42]. Protein C is a vitamin Kdependent protein and activated protein C inactivates FVa and FVIIIa. Protein S, also a vitamin K-dependent protein, is a necessary cofactor for activated protein C.
223
The Biomaterials Silver Jubilee Compendium M.B. Gorbet, M.V. Sefton / Biomaterials 25 (2004) 5681-5703
More than half of the protein S in plasma is bound to the C4b-binding protein (a regulatory protein of the complement system) and is not functionally active.
2.4. Biomaterials and coayulation pathways While under physiological conditions the role of FXII activation is questionable, in the presence of a cardiovascular device activation of FXII may occur. The artificial surface, with its adsorbed protein layer, then represents the required ("negatively charged") surface. Protein adsorption is the first event in blood-material interactions, and adsorption of the contact phase proteins may result in activation of the intrinsic cascade. Earlier studies had focused on protein adsorption on glass or biomaterial surfaces with isolated protein solutions or diluted plasma, and showed how fibrinogen is replaced over time by H M W K (the Vroman effect), suggesting a possible role of the intrinsic pathway in material thrombosis [43]. Recent studies using whole blood or plasma have provided new insights on the adsorption and activation of contact phase proteins. FXII adsorption has been observed in moderate amounts on materials used in vascular grafts [44] and hemodialysers [45]; however, it was not found in its activated form [46]. Although H M W K and prekallikrein may be adsorbed on the material surface, the lack of FXIIa on the material surface will stall the initiation of coagulation through the intrinsic pathway. In some instances, in vitro activation of FXII and kallikrein has been reported with biomaterials [47-49]; however, no test was performed to determine if such activation resulted in any significant activation of coagulation. Other studies have actually shown that only minute amounts of thrombin or thrombin-antithrombin III complex (TAT) are generated when biomaterials are incubated with undiluted plasma alone [50-52]. Furthermore, higher levels of adsorbed kallikrein and Factor XII on biomaterials do not correlate with TAT formation [50], suggesting that the contact phase proteins, by themselves, play little role in the activation of coagulation. In fact, a study by Hong et al. [51] suggested that the presence of leukocytes is required for activation of the coagulation cascade; the requirement of a TF-dependent pathway of initiation of coagulation may thus also apply to biomaterials. In vivo, a small increase of FXIIa is observed during cardiopulmonary bypass [53,54], but it appears to be in response to the surgical intervention and the establishment of extracorporeal circulation (i.e., exposure to the biomaterials of the circuit) does not further increase FXIIa levels [54]. Furthermore, no significant correlation has been observed between FXIIa and thrombin generation [54]. In vivo results with hemodialysis also failed to show any significant increase of FXIIa [55]. That a patient with a severe FXII deficiency showed levels of thrombin
5685
generation during cardiopulmonary bypass similar to normal patients [56] casts further doubt on the role of Factor XII in the initiation of coagulation with biomaterials. Taken together, these studies do not support the view that activation of the contact phase proteins is important in the activation of coagulation by biomaterials. While the TF pathway of blood coagulation has become the focus in hematology and has led to the revised version of the coagulation cascade (compare Figs. 2 and 3), the biomaterials community has been slow to recognize its importance. The textbooks refer to the older model of coagulation with the separated intrinsic and extrinsic pathways with the thought that the extrinsic pathway is not directly related to bloodmaterial interactions [1,57-59]. However, blood contact with a material represents a potential stimulus to induce TF expression by monocytes, resulting in blood coagulation through the extrinsic system. Indeed, monocyte TF expression has been observed in vivo during or after cardiopulmonary bypass [60-62]. Further details of TF expression in the presence of biomaterials are given elsewhere [63,64] and summarized below in the section on leukocytes. The role of leukocytes (most likely due to TF on monocytes) in activation of the coagulation by biomaterials was highlighted in the study by Hong et al. [51] that was referred to earlier. Thrombin-antithrombin formation (TAT) on PVC was found to be negligible in both plasma and platelet-rich plasma, while significant levels of TAT were observed in whole blood; i.e., only in the presence of leukocytes was there significant thrombin formation. More research is needed, however, to define the relative importance of the extrinsic and intrinsic pathways of coagulation in the overall picture of thrombosis on biomaterials. The time scales associated with initiation of the coagulation cascade by contact phase activation and with TF expression will have an impact on their relative importance. As it is part of the protein adsorption "reaction", contact phase activation occurs during the first few seconds/minutes of blood contact with a material. On the other hand, to be expressed on monocytes, TF requires synthesis and thus a minimum of 60rain (in a normal patient) would elapse before this pathway could significantly contribute to thrombin generation. Thrombin generation by FXIIa on materials is also dependent on flow [65], because of the effects of flow on mass transfer as well as direct effects on platelet/leukocyte phenotype. One manifestation of this is the relevant time scales of the coagulation reactions will vary with the flow situation. Thus, the relative roles of the intrinsic and extrinsic pathways in thrombin generation will likely depend on both the flow situation and the relevant time frame. This adds another level of complexity to the understanding/
224 5686
The Biomaterials Silver Jubilee Compendium M.B. Gorbet, M.V. Sefton / Biomaterials 25 (2004) 5681-5703
characterization of thrombin generation with cardiovascular devices.
2.5. Anticoagulants To prevent the formation of thrombi during cardiopulmonary bypass, hemodialysis and angioplasty, anticoagulants are routinely administered. Heparin is the most common [37]. Heparin binds to antithrombin III (ATIII) via a unique pentasaccharide sequence, and causes a conformational change in the reactive center of ATIII, thereby accelerating the rate of ATIIImediated inactivation of several clotting enzymes (thrombin, FXIa, FXa and FIXa). Heparin promotes the formation of the complex thrombin-ATIII (TAT) by binding to both proteins. On the other hand, to inactivate FXa, heparin needs to bind only to ATIII. ATIII forms a 1:1 irreversible complex with coagulation enzymes, the heparin then dissociates and can be reused. A limitation of heparin is that it is unable to inactivate thrombin bound to fibrin or to surfaces, or to inhibit FXa within the prothrombinase complex [66]. Heparin can also cause thrombocytopenia. Heparin binds to platelet factor 4 (PF4) and in some patients, antibodies will develop against this heparin-PF4 complex. The antibody-PF4-heparin complex then binds to platelets and induces platelet activation, aggregation and activates the blood-coagulation pathways, resulting in both a loss of circulating platelets and a thrombotic state [67]. Following implantation of cardiovascular devices, such as vascular grafts and artificial valves, anticoagulants such as warfarin is used. It can be taken orally and it interferes with the vitamin K cycle, thus impairing the biological function of vitamin K-dependent coagulation proteins (prothrombin, FVII, FIX and FX) [37]. For stents, anticoagulants are not needed except during placement. Aspirin and Plavix (clopidogrel, antiplatelet aggregation) are used after placement (aspirin forever; Plavix for 1 month or 1 year) to control thrombosis, while heparin and a GPIIb/IIIa antagonist (e.g., ReoPro) is only needed perioperatively. Other anticoagulants of interest for use of cardiovascular devices [66] are the tick anticoagulant peptide (TAP) and antistatin which binds to FXa even within the prothrombinase complex; hirudin, a leech-derived protein, a potent thrombin inhibitor; and D-Phe-ProArgCHzC1 (or PPACK), a peptide that directly inactivates thrombin by interacting with the active site of thrombin. Both PPACK and hirudin are able to inactivate fibrin-bound and surface-bound thrombin [66,68] which is a significant advantage. Ximelagatran is also a new oral direct thrombin inhibitor and clinical trials have been very successful: similar or superior efficacy relative to warfarin in some scenarios with reduced bleeding, obviating the need for monitoring. It
has many advantages over warfarin and will likely soon replace it [69]. When studying blood-material interactions in vitro, heparin is usually the anticoagulant of choice as it is the most widely used with cardiovascular devices. However, heparin also possesses some anticomplement activity [70]. PPACK and hirudin, which appear to be more specific than heparin, may then be used especially when the mechanisms of cell activation are studied. However, the associated high cost restricts their use.
3. Complement The complement system plays an important role in the body's defense mechanisms against infection and "non-self' elements. The complement system consists of more than 20 plasma proteins that function either as enzymes or as binding proteins. Complement activation is initiated via the classical or alternative pathways and the terminal pathway is common to both (Fig. 4). Both pathways contain an initial enzyme that catalyses the formation of the C3 convertase, which in turn generates the C5 convertase allowing the assembly of the terminal complement complex (TCC). Various complement products (C3b, C4b and iC3b) bind to particles, surfaces, bacteria and immune complexes in a process called opsonization [71], which facilitates their uptake by inflammatory cells. Activation of complement results in cell lysis when the terminal attack complex is inserted into the cell membrane. Complement activation also releases C3a, C4a and C5a, which are anaphylatoxins. These peptides are humoral messengers that bind to specific receptors on neutrophils, monocytes, macrophages, mast cells and smooth muscle cells. They induce a variety of cellular responses such as chemotaxis, vasodilatation, cell activation and cell adhesion [72]. At high enough concentration, they are responsible for the many systemic effects of complement activation.
3.1. Classicalpathway The classical pathway is normally triggered by antigen-antibody complexes that bind the C1 complex (Clq, Clr, C1 s) through the Clq component. This activates C ls, which is then able to cleave the C4 complement protein into C4a and C4b. C4b attaches to its target surface via its exposed metastable thioester binding site. It is important to note that C4b does not bind efficiently to membrane surfaces and the fluid phase C4b is rapidly inactivated by the loss of its binding site. C2 binds to the attached C4b and is cleaved by C ls, releasing C2a. The classical C3 convertase, C4bC2b, is thus formed and can cleave C3 into C3a (anaphylatoxin) and C3b. The combination of C4bC2b
225
The Biomaterials Silver Jubilee Compendium M.B. Gorbet, M.V. Sefton / Biomaterials 25 (2004) 5681-5703
5687
Fig. 4. Pathways of (a) complement activation [256], copyright JB Lippincott and (b) inhibition [257], copyright Elsevier. In (a), the classical and alternative pathways are shown leading to the production of the C3 convertase and C5 convertase complexes, both of which are part of amplification loops. In (b), various natural and synthetic inhibitors are shown inhibiting downstream portions (C3 and C5) of the complement cascades; only C1 inhibitor is shown as being effective at a level prior to the formation of the C3 or C5 convertases.
and C3b becomes the C5 convertase, which cleaves C5 into C5a (anaphylatoxin) and C5b. C5b is the first component of the terminal complex and has high affinity for C6. C5bC6 then binds C7, C8 and up to 12 molecules of C9 and this forms the TCC C5b-9. If C5b is attached to a biological surface, the TCC (also called the membrane attack complex, mC5b-9) inserts itself into the lipid layers resulting in cell damage and/or lysis. In the absence of a biological membrane, the complex binds to S protein (also known as vitronectin) to create SC5b-9 in the fluid phase.
3.2. Alternative pathway The activation of the alternative pathway does not require antibody or immune complexes and is activated by any foreign surfaces, such as fungal, bacterial polysaccharides, lipopolysaccharides (LPSs), particles and biomaterial surfaces. Complement activation via the alternative pathway occurs spontaneously at a low rate.
Spontaneous hydrolysis of the internal thioester group of C3 occurs in the fluid phase, generating C3. H20. This hydrolyzed C3 can bind and activate Factor B and cleave another C3 molecule into C3a and C3b. The alternative C3 convertase, C3bBb, is formed. In the absence of a surface to support binding of C3b, little complement activation occurs. In the presence of a surface, covalent binding of C3b to hydroxyl or amine groups on the surface may occur via the carbonyl group in the C3b thioester binding site. Attachment of C3b to a surface favors binding of Factor B and Factor D to C3b. Factor D cleaves Factor B into Ba and Bb, and the alternative C3 convertase, C3bBb, is formed again. This attached C3 convertase is able to generate more C3b, resulting in a positive amplification loop. Properdin acts to stabilize the C3 convertase. The clustering of C3b on the surface allows the formation of the alternative C5 convertase, C3bBbC3b, and C5 can be cleaved. The assembly of the TCC follows as for the classical pathway.
226 5688
The Biomaterials Silver Jubilee Compendium M.B. Gorbet, M.V. Sefton / Biomaterials 25 (2004) 5681-5703
3.3. Reyulatory molecules of complement activation The various molecules involved in the regulation of complement activation are illustrated in Fig. 4b. The classical pathway is regulated by two specific mechanisms [29]: C1 inhibitor, a plasma protein, binding Cls or C lq and inhibiting the enzymatic activity of the C1 complex; and the C4b-binding protein inhibiting C4b bound to a membrane or a surface. Factor I is a proteolytic enzyme that binds C4b and C3b, generating iC4b and iC3b, which are further degraded into smaller fragments. C4b-binding protein is a cofactor of Factor I and augments the degradation of C4b by Factor I. Factor H is a cofactor of Factor I for the degradation of C3b. Factor H is also able to displace Bb in the C3 convertase to promote C3b inactivation by Factor I [75]. On cell surfaces, the C3 and C5 convertase activity are regulated by three integral membrane proteins [75]. Decay-accelerating factor (DAF), found on all peripheral blood cells, destabilizes the C3 convertase by promoting the release of factor Bb. The membrane cofactor protein (MCP), expressed on leukocytes and platelets, favors the dissociation of Factor B and promotes C3b association with Factor I. The complement receptor type 1 (CR1), present on all blood cells except platelets, acts like Factor H and displaces Bb from the C3 convertase and facilitates inactivation by Factor I. To prevent lysis of "bystander" blood cells, membrane proteins called homologous restriction factors limit the ability of the terminal complex to properly assemble on autologous cells [75]. Two proteins have been characterized: CD59 and the C8-binding protein (also called MAC inhibiting protein). CD59, found not only on blood cells but on many cells such as hepatocytes and epithelial cells, binds to C8 and C9 and inhibits C9 polymerization. Little is known about the C8-binding protein, which is believed to bind C8 and C9.
3.4. Interactions of complement and coagulation cascade Although the coagulation and complement cascades are discussed as separate entities, the two cascades appear to interact significantly to modulate each other's activity [29,75]. Factor XIIa and kallikrein are known to cleave C ls [29] and thus have the capacity to trigger classical complement activation. Thrombin activates C3, C5, C6 and Factor B; kallikrein cleaves C5 and factor B; and Factor XIIa also cleaves C3. The activity of thrombin on C3 and C5 may actually explain the higher background levels of C3a and C5a in serum versus plasma. Table 1 outlines the various interactions between complement and coagulation factors.
Table 1 Interactions between complement and coagulation systems [29,75] Protein
Type of interaction
Thrombin Factor XIIa Kallikrein Antithrombin III Bb C3bBb C 1 inhibitor S protein (vitronectin) C4b-binding protein
Proteolysis of C3, C5, C6 and factor B Proteolysis of Clr, Cls and C3 Proteolysis of C1, C5 and Factor B Protect RBC from lysis by mC5b-9 Proteolysis of prothrombin Proteolysis of prothrombin Inactivates FXIIa and kallikrein Stabilizes plasminogen activator inhibitor 1 Binds to the vitamin K-dependent protein S
3.5. Complement activation and biomateria& Complement activation is generally treated as if it is part of the inflammatory response induced by biomaterials. For example, complement activation is known to occur during cardiopulmonary bypass and hemodialysis [75-78], and with catheters and prosthetic vascular grafts [79,80]. It is recognized that, both in the short and long term, complement activation plays a role in the leukocyte related clinical sequelae associated with the use of cardiovascular devices such as leukopenia, hypotension and pulmonary injury [81,82]. The thrombotic consequences of leukocyte activation are discussed below. The presence of a biomaterial is believed conventionally to activate complement via the alternative pathway. Biomaterials are usually classified as "activating" or "non-activating" surfaces [74]. On a non-activating surface, negatively charged groups such as carboxyl and sulfate, sialic acid and bound heparin appear to promote high-affinity association between bound C3b and Factor H. On the other hand, an activating surface is usually characterized by the presence of nucleophiles such as hydroxyl and amino groups: these groups will allow covalent binding of C3b and promote formation of the C3 and C5 convertase [2,73]. However, even in the absence of these activating groups on the surface, some biomaterials, such as polyacrylonitrile, are able to activate complement, suggesting that the mechanisms of material-induced complement activation due to nucleophilic groups is not the whole story. A newer hypothesis places emphasis on interaction of Factor H with the surface: an activating material is then defined by its capacity to bind Factor B rather than Factor H [73]. As noted earlier, binding of Factor H would lead to C3b inactivation by Factor I and thus terminate the propagation of the complement cascade. Some activating materials generate high levels of both C3b and C5b-9, while others will generate high C3b level but little C5b-9. Why the efficiency of C5 convertase formation relative to that of C3 formation differs from one activating surface to another is not well understood.
227
The Biomaterials Silver Jubilee C o m p e n d i u m
5689
M.B. Gorbet, M.V. Sefton / Biomaterials 25 (2004) 5681-5703
However, even low amounts of C5b-9 are able to activate leukocytes [74] and thus a low terminal complement activating material may still induce a significant inflammatory response. The question remains as to which are the appropriate levels of complement activation that the host can accept without deleterious effect. We also have to determine if, in the long term, the host will be able to differentiate/discriminate between moderate and high complement activating surfaces. The hypothesis that material-induced complement activation occurs exclusively via the alternative pathway has also been challenged. Reports of complement activation via the classical pathway during cardiopulmonary bypass [18,83] and a delay in complement activation observed with C4-deficient patients undergoing hemodialysis [84] suggest that classical activation plays a role in material-induced activation. The presence of immune complexes may allow for a rapid onset of complement activation, and then subsequently the alternative pathway becomes activated. In vitro studies have also demonstrated activation of the classical pathway by some biomaterials [85-87]. C1 inhibitors such as pentamidine and benzamidine were effective in lowering platelet adhesion and activation by polystyrene beads and polyethylene tubes while sCR1, an inhibitor of both pathways at the level of C3, had no effect [88,89]. Pentamidine was also effective in a canine chronic shunt in eliminating the thrombocytopenia seen with a platelet activating material (a polyvinyl alcohol (PVA) hydrogel)[90]. While some of the consequences of complement activation are well understood, more work is needed to fully understand how a material activates complement. Questions to be resolved include selecting inhibitors that block both pathways and at an early enough stage to inhibit the local (rather than systemic) effects of complement activation. Similarly, controlling the differential adsorption of C3b, Factor D and Factor H may be more important than preparing low adsorption, so-called "non-fouling" surfaces. The latter may lower the adsorption of all proteins but it may be more important to alter the composition of the protein adsorbate.
Platelets respond to minimal stimulation and become activated when they contact any thrombogenic surface such as injured endothelium, subendothelium and artificial surfaces. Platelet activation is initiated by the interaction of an extracellular stimulus with the platelet surface. This interaction involves the coupling of the agonist to specific receptors on the platelet plasma membrane [91]. Plasma proteins such as thrombin and fibrinogen; vascular wall products such as collagen; and molecules derived from inflammatory cells (i.e., leukocytes) or platelets, such as platelet activating factor (PAF) or cathepsin G, are all potent platelet activators. A list of known platelet receptors and their specific agonist/ligand is presented in Table 2. Activation results in at least five physiologic responses [92]. (1) A platelet release reaction in which biologically active compounds stored in intracellular platelet granules are secreted into the microenvironment, such as platelet factor 4, thrombospondin, fl-thromboglobulin, ADP and serotonin. (2) P-selectin (previously referred to as GMP-140 or P A D G E M ) is released and expressed on the platelet membrane after e granule secretion. Table 2 Platelet receptors [91,102] Receptor
(a) Receptors leadin9 to platelet activation
Thrombin receptor Thromboxane A2 receptor V1A receptor PAF receptor 5HT2 receptor ~2 receptor ADP receptor PGE2 receptor C 1q receptor
4.1. Platelet biology
The platelet's main role in hemostasis is to preserve the integrity of the vascular wall through formation of a platelet plug. Platelets are anuclear, disc-shaped cells with a diameter of 3-4gm. They are derived from megakaryocytes in the bone marrow and circulate at an average concentration of 200 x 106 cells/ml, with individual platelet concentrations ranging from 150 to 400 • 106 platelets/ml.
Thrombin TxA2, PGH2, PGG2 Vasopressin Platelet activating factor Serotonin or 5-OH tryptamine Epinephrine ADP, ATP PGE2 Clq
(b) Receptors leadin9 to platelet inhibition A2 receptor Adenosine
PGI2 receptor PGD2 receptor
PGI2, PGE~ PGD2
(c) Platelet adhesion receptors (bindin9 may also result in platelet activation)
GPIa/IIa or VLA-2 GPIb/IX or GPIb
4. Platelets
Ligand/agonist
GPIc/IIa or VLA-5 GPIc'/IIa or VLA-6 GPIIb/IIIa GPIV or GPIIIb GPVI Vitronectin receptor PECAM-1 FcT-RII ICAM-2 P-selectin Leukosialin, sialophorin
Collagen Von Willebrand factor (vWF), thrombin Fibronectin Laminin Collagen, fibrinogen, fibronectin, vitronectin, vWF Collagen, thrombospondin, Collagen Vitronectin, thrombospondin Heparin Immune complexes LFA-1 Sialyl-Lex ICAM-1
GP: glycoprotein; VLA: very late antigen.
228 5690
The Biomaterials Silver Jubilee Compendium M.B. Gorbet, M.V. Sefton / Biomaterials 25 (2004) 5681-5703
P-selectin is a cell-surface glycoprotein belonging to the selectin family and plays an important role in mediating adhesion of activated platelets to neutrophils, monocytes and a subset of lymphocytes [93,94]. (3) The platelet eicosanoid pathway is initiated, resulting in the liberation of arachidonic acid from platelet phospholipids and in the synthesis and release of prostaglandins and thromboxane B2. (4) Platelet activation is characterized by a drastic shape change, which promotes platelet-platelet contact and adhesion. The rearrangement of the platelet membrane during activation also promotes the association of the tenase and prothrombinase complexes on its phospholipids. (5) Platelet activation results in the formation of platelet microparticles (PMPs), which are particularly rich in factor Va, platelet factor 3 and phospholipid-like procoagulant activity (phosphatidylserine) [95,96]. PMPs are formed from the surface membrane through exocytotic budding. Their physiologic role remains unclear but in vitro results have shown that they can bind and adhere to fibrinogen and fibrin, and coaggregate with platelets [97,98]. The procoagulant activity of PMPs, generated both in vitro and in vivo, has also been demonstrated [99-101]. Among the different platelet adhesion receptors (Table 2), GPIb and GPIIb/IIIa have the highest density on platelets. GPIb (CD42) is a leucine-rich glycoprotein receptor and approximately 25,000 receptors are present on the platelet surface [92,102]. It is complexed one to one with GPIX but the function of the latter remains unknown. GPIb is a long molecule, making it more susceptible to conformational change upon shear stress. GPIb mediates platelet interaction with von Willebrand factor (vWF). It will not bind plasma vWF unless the antibiotic ristocetin or the snake venom botrocetin is present. On the other hand, GPIb will bind to adsorbed or immobilized vWF on a surface. Shear stress is an important factor in platelet adhesion to vWF as it induces the required conformational change of vWF to bind GPIb. Thrombin also binds to GPIb but the functional significance of this binding has not been elucidated. GPIIb/IIIa (CD41/CD61) is an integrin receptor and is constitutively expressed on platelets. GPIIb/IIIa is the dominant platelet receptor with 40-80,000 receptors present on the surface of a resting platelet. Another 20-40,000 are present inside the platelets, in c~ granule membranes and in the membranes lining the open canalicular system [102]. They are translocated to the platelet membrane during the release reaction. On resting platelets, GPIIb/IIIa is in an inactive form and has a low-affinity binding site for adsorbed fibrinogen [103]. Upon platelet activation, a conformational change occurs leading to the exposure of the high-affinity binding site for soluble fibrinogen. Binding of fibrinogen to GPIIb/IIIa leads to platelet aggregation as well as
formation of platelet-leukocyte aggregates, via crosslinking of GPIIb/IIIa on two different platelets by fibrinogen and crosslinking between GPIIb/IIIa and Mac-1 (on leukocyte) by fibrinogen. Other adhesive glycoproteins containing RGD sequences can also bind to activated GPIIb/IIIa: vWF, thrombospondin, fibronectin and vitronectin. Since GPIIb/IIIa mediates platelet aggregation, its inhibition has generated much interest in the control of prothrombotic states [104,105]. An antibody against GPII/IIIa (7E3 also called Reopro or Abciximab) has been developed and has entered clinical trials with angioplasty [106], myocardial infarction [107] and unstable angina [107,108]. All clinical trials have shown a significant improvement of longterm survival [109,110]. However, higher complication rates such as bleeding and thrombocytopenia have also been observed [111,112]. New clinical trials are underway to determine appropriate regimens [110].
4.2. Platelets and biomaterials Platelet activation (platelet release, PMP formation, P-selectin expression, aggregation) and adhesion is known to occur [1,58] during cardiopulmonary bypass, hemodialysis, as well as with vascular grafts and catheters. The thrombotic complications associated with cardiovascular devices are linked clearly to their ability to activate platelets. Adherent platelets [113] and circulating PMPs generated by material contact [100,101] have been shown to be procoagulant in nature. Association between platelets and leukocytes via Pselectin also occurs in the presence of cardiovascular devices [8] and such associations have become a relatively new parameter to study biocompatibility. However, the implications of this association are mostly unknown: they may directly or indirectly contribute to thrombin generation (via monocyte TF) or participate in the removal of platelets from the circulation since several platelets can be bound to each neutrophil or monocyte. While it is intuitive to suggest that a non-thrombogenic surface should not support platelet adhesion it has not, unfortunately, been that simple. It has been found that platelet contacts with some biomaterial surfaces results in activation leading to high consumption (removal from circulation) characterized by high platelet turnover rather than adherence and by the formation of microemboli rather than occlusive thrombi. Following contact with the layer of adsorbed proteins on the artificial surface, platelets will either adhere or "bounce off" [114], most likely depending on their state of activation and the ligands present at the interface [115]. Platelet adhesion on surfaces is mediated by GPIIb/IIIa and fibrinogen and interaction with GPIb/IIa and vWF can also occur [116-119]. However, the absence of significant platelet adhesion does not preclude platelet
229
The Biomaterials Silver Jubilee Compendium M.B. Gorbet, M.V. Sefton / Biomaterials 25 (2004) 5681-5703
activation as shown by the generation of PMPs with PVA hydrogel both in vitro [120] and ex vivo [90]. Indeed, animal studies have shown that, despite the absence of platelet adhesion, blood contact with some hydrogel surfaces [121,122], like contact with NHLBI reference materials and Silastic | [123] appears to activate platelets, resulting in their removal from the circulation. Furthermore, Hanson et al. [122] demonstrated a direct linear relationship between the water content of hydrogels and the rate of platelet consumption in a baboon AV shunt model. They also noted that platelet consumption was reduced by the antiplatelet agent, dipyridamole [124], similar to its effectiveness in normalizing platelet survival in-patients with artificial heart valves [125]. While there is agreement that the rapidly adsorbed proteins, especially fibrinogen, play a critical role in platelet adhesion, it is not clear, for example, how effective adsorbed fibrinogen is as a platelet agonist in vivo and what other mechanisms are involved in supporting or initiating platelet adhesion. While adsorbed fibrinogen is likely the critical ligand for adhesion, what activates the platelet to adhere in the first place? Further it has not been clear that inhibition of platelet adhesion will inhibit the generation of PMPs or reduce platelet consumption. The strategies available to minimize platelet adhesion (e.g., polyethylene glycol immobilization) have not been sufficient to prevent platelet consumption [126]. The mechanism of materialinduced platelet activation is often presumed to be via the generation of thrombin due to activation of the intrinsic coagulation cascade or the release of ADP from damaged red blood cells or platelets. Even in the presence of heparin, small levels of thrombin generation are generated and may activate platelets. However, the inability of thrombin and kallikrein inhibitors to reduce platelet activation suggests that platelet activation is at least in part mediated by other agonists [127]. For example, a correlation between complement activation and thrombocytopenia has been noted during dialysis [128,129]. Complement activation can lead to platelet activation in many ways. Platelets possess a receptor for C lq that has been shown to induce GPIIb/IIIa activation, P-selectin expression and procoagulant activity [130]. It is currently not known how classical complement activation leads to activated platelets, but in vitro results support a role for C1 [89]. Insertion of C5b-9 in platelets has also been associated with increased platelet procoagulant activity [95]. In vitro studies using human cells have been conducted to probe the effectiveness of various agents to inhibit platelet activation [88] using a microsphere based immunoassay. Platelet adhesion to polystyrene microspheres was found to be unaffected by the complement inhibitor, sCR1 which is otherwise capable of inhibiting material-induced SC5b-9 produc-
5691
tion. In contrast, classical pathway complement inhibitors, pentamidine, benzamidine, pyridoxal-5-phosphate (P5P) and cysteine were able to inhibit platelet adhesion to the polystyrene surface. These agents also inhibited platelet loss and microparticle levels in whole blood after contact within polyethylene tubing. Benzamidine and a derivative, pentamidine, are able to competitively inhibit C ls enzyme (ionic interaction with active site), although pentamidine has an IC50 10 times lower [131-133]. The antiplatelet effects of pentamidine have been documented but the mechanism of action remains unresolved. It has been reported that pentamidine has no disruptive effect on GPIIb/IIIa receptors, intracellular cAMP levels, calcium ion influxes or intracellular pH changes [88,134,135]. P5P, a major coenzyme from Vitamin B6 is also a known inhibitor of C 1 fixation [136,137]. Cysteine is known to inhibit both C ls and the alternative complement pathway [138]. Since sCR1 (in vitro) was not able to block platelet adhesion and activation, it is our hypothesis that C1 is playing a role in material-induced platelet activation and that these agents are effective because of their ability to inhibit C ls. These agents also effectively inhibit other serine proteases such as thrombin, trypsin and plasmin [132] and thus unequivocal delineation of the mechanism is not yet possible. However, during in vitro blood-material contact, inhibiting complement activation has led to conflicting results on the role of the terminal complement pathway in material-induced platelet activation. Monoclonal antibodies to C5 and C8 inhibited platelet activation during simulated extracorporeal circulation (SECC) [139,140]. On the other hand, sCR1 had no effect on platelets in our microsphere assay [88,141] and in a different extracorporeal circulation model [142]. The difference in experimental conditions such as higher flow rate, blood dilution and the presence of mannitol (a hydroxyl scavenger) between the studies may account for the difference in the apparent efficacy of terminal pathway inhibition. There are likely multiple pathways whereby platelets are activated, some more relevant during conditions of high complement and leukocyte activation and some more relevant where these pathways are less well developed. Further research is needed to fully understand which complement protein triggers material-induced platelet activation. In our arterio-venous canine shunt model with a platelet consumptive PVA hydrogel tubing segment, systemic low molecular weight heparin did not reduce material-induced platelet damage: loss of platelets, microparticle generation or reduced lifespan, indicating that thrombin production did not appear to play a role. On the other hand, pentamidine (12 mg/kg, daily, IM) dramatically inhibited thrombocytopenia during the connection of PVA hydrogel test segments (Gemmell, 2001, pers. comm.). This supports our hypothesis that
230 5692
The Biomaterials Silver Jubilee Compendium M.B. Gorbet, M.V. Sefton / Biomaterials 25 (2004) 5681-5703
such inhibitors are useful for blocking biomaterialassociated platelet activation. We have also used ReoPro T M (up to 0.Smg/kg, GPIIb/IIIa antagonist) and not observed any effect on thrombocytopenia caused by PVA. The ReoPro T M dosage was sufficient to inhibit agonist induced platelet aggregation. This preliminary finding could suggest that platelet adhesion (and vesiculation)--both blocked by IIb/IIIa inhibitors--are not responsible for premature platelet consumption and clearance, at least on these smooth surfaces that have few platelet deposits in the absence of any therapeutic agents.
5. Leukocytes
5.1. Leukocyte biology Circulating leukocytes comprise neutrophils, monocytes, lymphocytes, basophils and eosinophils. Only neutrophils and monocytes in blood, but not after they emigrate into tissues, will be addressed in this review as they are the major players in the inflammatory response with cardiovascular devices. Neutrophils are the most abundant white blood cells, representing 40-60% of the leukocyte population (3-5 • 106 neutrophils/ml), while monocytes represent 5% with a concentration of 0.2-1 x 106 monocytes/ml. Under normal circumstances, neutrophils have a very short half-life in blood (8-20 h) but after an inflammatory stimulus, such as LPS or cytokines, their lifespan can increase up to three-fold [143] and their role may be even more active. Monocytes enter the circulation for a short period (36-104h) and they migrate into tissues where maturation and differentiation occur and they become macrophages [144]. They may also be deposited on injured blood vessels and later differentiate into
macrophages. It is important to note that when neutrophils and monocytes are recruited in tissues during inflammation, their half-life increases to several days. Monocytes and neutrophils possess receptors for different complement products and other pro-inflammatory mediators such as PAF and cytokines. Platelet release, such as/~-thromboglobulin and PDGF, has also been reported to activate neutrophils in vitro [145-147]. Other neutrophil and monocyte activating stimuli include bacteria and their products and cell adhesion. A list of the most relevant receptors involved in the inflammatory response is presented in Table 3. As for platelet activation, leukocyte activation results in several physiological responses. (1) Upon activation, changes in expression of membrane receptors occur on neutrophils and monocytes: upregulation of CD1 l b by translocation from intracellular granules [148], shedding of L-selectin by shedding [148], synthesis and expression of TF [149]. TF expression by leukocytes is the subject of another review [63]. (2) Leukocyte activation results in release of inflammatory mediators. Neutrophils contain three types of granules (gelatinase, specific and azurophil granules) and their contents may be released upon activation: among others, elastase, cathepsin G and lactoferrin are important inflammatory mediators. Cytokines such as IL-1, IL-6, IL-8, TNFe, G-CSF and GM-CSF are also released. Arachidonic acid metabolites, such as leukotriene B4 and PAF, are produced and released by activated neutrophils. The released inflammatory mediators have various properties: they may be chemoattractant for leukocytes, promote adherence to endothelial cells, and further activate platelets or leukocytes. (3) Activation may also result in the onset of the oxidative burst whereby neutrophils and monocytes release oxidants, such as O2 and H 2 0 2 . These products damage tissues and activate cells [150].
Table 3 Leukocyte receptors in acute/immediate inflammatory response [254,255] Receptor
(a) Complement receptors Clq R CR1 CR3 or CD 1lb CR4 or CD1 l c C3a R C5a R
(b) Other receptors (R) TNF~, IL-1 IL-8 PAF, LTB4 GM-CSF, G-CSF, IFN7
Ligand
Function
Clq, MBP
Enhance phagocytosis, respiratory burst Immune adherence, phagocytosis
C3b > C4b >iC3b iC3b, fibrinogen, FX, ICAM- 1 iC3b, fibrinogen C3a C5a
Phagocytosis, respiratory burst, adhesion Adhesion, phagocytosis Chemotaxis, degranulation, respiratory burst Chemotaxis, degranulation, respiratory burst
Degranulation, respiratory burst Chemotaxis, degranulation, respiratory burst Strong activation Weak activation, priming
231
The Biomaterials Silver Jubilee Compendium M.B. Gorbet, M.V. Sefton / Biomaterials 25 (2004) 5681-5703
5.2. Leukocyte activation and biomater&ls
Table 4 Leukocyte adhesion receptors Receptor
(a) Selectin family L-selectin
5693
Ligand Mucin-like ligand, lymph node addressin
(b) Integrin family CD 11a/CD 18 (or LFA- 1) CD 11b/CD 18 (CR3 or Mac- 1) CD 11c/CD 18 (CR4 or p 150,95) VLA-1 VLA-2 VLA-3 VLA-4 VLA-5
ICAM-1, ICAM-2, ICAM-3 iC3b, fibrinogen, FX, ICAM-1 iC3b, fibrinogen Laminin, collagen Collagen Fibronectin, laminin, collagen VCAM-1 Fibronectin
(c) Immunoglobulin family PECAM-1 FcT-RI, FcT-RII, FcT-RIII
PECAM- l, heparin Immune complexes
VLA: very late antigen.
(4) Activated neutrophils and monocytes also have an increased adhesive capacity on endothelium and other surfaces [151]. Leukocyte adhesion to the endothelium is an important means by which neutrophils and monocytes participate in the inflammatory response. Adherent leukocytes have been shown to be more activated than their counterpart in the bulk [152,153] but the level of activation of adherent leukocytes depends on the surface [154]. Leukocyte adhesion molecules are divided into three main families [155]: the selectins, the integrins and the immunoglobulin superfamily (Table 4). The mechanism of leukocyte adhesion to endothelial cells, a threestep mechanism, has been well characterized [156]. Step 1: L-selectin is involved in the initial rolling of leukocytes on endothelium. Step 2: The rolling stage enables leukocytes to slow their movement and sample the local environment, and they may become activated due to local stimulation and additional interaction between receptor/ligand. Step 3 : C D l l / C D 1 8 mediates firm adhesion. With activated neutrophils and monocytes, a functional upregulation is observed for CD 11a, while for CD1 l b and CD1 l c both a quantitative and functional upregulation occurs, the upregulation for CD1 l b being more rapid and important than for CDllc. It is important to note that the functional change of CD1 l b upon leukocyte activation and/or adhesion can occur despite no measurable increase in C D l l b surface expression. The functional change is conformational involving receptor phosphorylation and resulting in increase binding affinity for certain ligands (such as fibrinogen and Factor X) [157]. On the other hand, a quantitative increase in CD 11b expression on leukocytes does not imply increased adhesion, unless i t is accompanied by functional change in the receptor [158].
Contact with cardiovascular devices in vivo activates both neutrophils and monocytes. Indicators of leukocyte activation such as L-selectin shedding and CD1 l b upregulation on leukocytes have been widely observed following angioplasty [159-162], hemodialysis [163-165] and cardiopulmonary bypass [20,22,166-169] (for an extensive review of studies on the expression of leukocyte adhesion molecules with in vitro cardiopulmonary bypass, see Asimakopoulos and Taylor [170]). Degranulation with the release of elastase and lactoferrin [171-175] and the presence of cytokines [169,176,177], such as IL-1 and TNF~, have been associated with extracorporeal circulation and further demonstrate leukocyte activation. Activation of the respiratory burst is also a common trait with hemodialysis [164,178]. Material-induced leukocyte activation also results in increased adhesion. As the biomaterial is larger than a micro-organism and cannot then be engulfed by leukocytes, adherent neutrophils and monocytes undergo a frustrated phagocytosis whereby they release their array of potent oxygen metabolites and proteolytic enzymes [144]. Material characteristics and proteins at the interface appear to modulate the level of activation of adherent leukocytes [153,179,180]. In vivo studies have found activated leukocytes adhering to stents [181,182], oxygenators [183] and hemodialysis membranes [184,185]. Material-activated leukocytes also adhere to the endothelium, such as at the anastomoses of a vascular graft or in the lung during extracorporeal circulation. Heparinized human whole blood contact with a PVA hydrogel surface in vitro for 1 h lead to a two-fold upregulation of C D l l b , typical of the degree of leukocyte activation induced by phorbol esters [186]. In contrast, whole blood contact with PE and Silastic T M surfaces resulted in minimal CD1 l b upregulation. We have also reported that many clinical materials can activate isolated neutrophils (without platelets) suspended in plasma and that fibrinogen adsorption (plasma but not serum pretreatment) enhances activation [153]. Longer exposures (2h) to whole blood with polystyrene beads and polystyrene beads grafted with polyethylene glycol (PS-PEG) lead to expression of monocyte TF [187] as well as C D l l b upregulation. Activation was dependent on the surface area to volume ratio, but there was no difference in the extent of activation in comparing PS beads with PS-PEG (or PSPEG-NH2) beads. On the other hand, monocyte TF expression and adherent platelet density were all greater on PS than PS-PEG; there were no differences in leukocyte adhesion densities. The mechanisms of leukocyte adhesion on artificial surfaces are not clear, but it appears to be mediated in
232 5694
The Biomaterials Silver Jubilee Compendium M.B. Gorbet, M.V. Sefton / Biomaterials 25 (2004) 5681-5703
part by the complement product iC3b [72,144]. This is supported by in vitro work showing that inhibition of complement activation in vitro significantly reduced leukocyte adhesion [153,188-190] while fibrinogen also appears to play an important role in leukocyte adhesion to materials [191,192]. In experiments with isolated leukocytes, however, it appears that monocyte TF expression is only partly dependent on complement inhibition but appears to be also dependent on the presence of platelets [64]. The removal of platelets or the blocking of GPIIb/IIIa by monoclonal antibody (7E3) partially lowered the degree of material-induced TF expression while having no effect on CDllb. The presence of platelets on the surface may also mediate leukocyte adhesion via the interaction between P-selectin and PSGL-1 and/or GPIIb/IIIa and CD1 lb [193,194]. Conflicting reports also exist on the requirement for platelets in leukocyte adhesion on artificial surfaces [195,196]. The mechanisms of material-induced leukocyte activation as distinct from adhesion also remain unknown. Whether they are directly activated by contact with a foreign surface, via complement activation, kallikrein or platelet activation has not been fully determined. In vitro and in vivo investigations with protease inhibitors [127,175,197-199], complement inhibitors [21,22,186,200-203] and antiplatelet agents [106] suggest that they all play a role, but no one inhibitor has led to consistent results with a significant reduction of material-induced leukocyte activation. Material-induced leukocyte activation may be mediated by several factors and inhibition of one pathway of activation may not be sufficient to result in a significant impact on leukocyte activation. For example, we have shown that complement inhibition via sCR1 was only partially effective in reducing leukocyte activation [186]. On the other hand, a combination of sCR1 and antiGPIIb/IIIa (which blocks platelet activation) reduced material-induced leukocyte activation to almost background levels. The complexities of leukocyte activation by the PEG modified materials and the inhibitory effects of sCR1 and pentamidine are discussed elsewhere [204]. In summary, the mechanisms that regulate materialinduced leukocyte activation are as yet not well understood, precluding a clear scientific basis for strategies for a significant reduction in the inflammatory response induced by cardiovascular devices.
5.3. Leukocytes, platelets and coagulation Circulating monocytes and neutrophils normally roll on the endothelium. They will however adhere to damaged or stimulated endothelial cells or adherent platelets and further contribute to localized thrombogenesis. We have evidence [64,187] that similar phenom-
ena apply to biomaterials. The different procoagulant activities of leukocytes may be classified as: 9 Membrane-associated procoagulant activity: via TF
expression on the cell membrane (TF-dependent coagulation pathway) or via TF-independent mechanisms through factor X binding to C D l l b receptors leading to factor Xa generation or fibrinogen binding to CD1 l b; or binding of the prothrombinase complex on the membrane. 9 Release of procoagulant mediators: degranulation and oxidative products have the capacity to neutralize various anticoagulant proteins and activate platelets. 9 Association between platelets and neutrophils or monocytes: their interactions may lead to mutual activation and to a microenvironment protected from inhibitors. For example, following blood contact with cardiopulmonary bypass circuits or ventricular assist devices, TF expression on monocytes has been observed in vitro [205-207] and in vivo [60,61,206]. It has also been shown that C D l l b , upregulated on monocytes by cardiopulmonary bypass, was able to directly activate factor X [208] and platelet-leukocyte aggregates have been observed in several scenarios [106,209-211] as noted above. The potential role of leukocytes in thrombogenesis is underscored by the number of studies that have tried to minimize thrombus formation by the administration of drugs specifically targeted at leukocytes. Antibodies to block leukocyte adhesion may prove to be a reasonable therapeutic approach in the prevention of thrombus formation as illustrated in in vivo baboon models [212,213]. Overall, however, there is relatively little known on the potential contribution of expression of leukocyte procoagulant activities to thrombogenesis and thrombotic complications associated with the use of biomaterials and cardiovascular devices.
6. Other important factors
6.1. Flow Fluid dynamics affects the growth of thrombi and the deposition of fibrin. The composition difference between arterial and venous thrombi is one old example of this, although the underlying mechanisms are still not well understood. Thorough reviews are available [5,65, 214-216]. Flow determines the rates of transport of cells and proteins to the surface; it can also change the level of receptor expression on platelets and leukocytes. As platelets are an important part of the thrombus, the effect of shear on platelets has been studied extensively. Higher shear results in higher platelet deposition and lower fibrin deposition, while at lower shear the inverse
233
The Biomaterials Silver Jubilee Compendium M.B. Gorbet, M.V. Sefton / Biomaterials 25 (2004) 5681-5703
is true [214]. High shear, such as the ones observed at stenotic plaques, is also able to induce platelet aggregation even in the absence of any other exogenous factors [217,218]. Conflicting results have been obtained on the effect of flow on leukocyte adhesion while little is known of its effect on leukocyte activation. High shear has been shown to either reduce, increase or leave unchanged leukocyte adhesion on different substrates [195, 219-225]. These conflicting results may be explained by differences in experimental conditions: the presence of red blood cells [220], platelets [195,223,226] and plasma proteins [195]; the surface studied [223] and the state of leukocyte activation [224]. As for the effects of flow on the coagulation cascade, it has been studied less. Current knowledge is limited to Factor Xa generation initiated by the extrinsic pathway and thrombin generation initiated by the intrinsic pathway (with biomaterials). Factor Xa generation by the complex TF:VIIa increases with shear rate (and shear stress) [227,228]. For thrombin generation by the intrinsic pathway, modeling has identified three types of reactions [65]: at low flow, a significant amount of thrombin is produced after a long lag time (over 10 h); at moderate flow, significant thrombin generation is produced in a short time (within minutes); at high flow, low levels of thrombin are produced within seconds. Turbulent flow can be present at anastomoses, joints, and bifurcations of cardiovascular devices and such turbulence also contributes to the observed thrombosis. It is believed to play a significant role in the failure of mechanical heart valves, for example. Turbulent flow (in distinction to recirculation and stagnation zones, which may or may not have the characteristics of turbulence) results in hemolysis and/or cell activation but the mechanisms leading to thrombus formation are still poorly understood. Platelet deposition remains the focus of most studies [229] but the literature on turbulent flow and thrombus formation is more limited. Nonetheless, much effort is done to design devices so that recirculation or stagnation zones are avoided, since these are known niduses for thrombus growth. While the importance of flow has been recognized, our current understanding of its mechanisms is limited mostly to platelets. Many in vitro and in vivo flow models are available and have been successfully used to assess antithrombotic drugs in whole blood [216]. More fundamental research is required on blood coagulation, leukocytes and flow. Previous research with flow has focused on isolated cells or proteins, which is far from the in vivo situation. The critical role of red blood cells in the in vitro study of mechanisms of leukocyte adhesion was recently demonstrated by Melder et al. [230]. In the absence of erythrocytes, blocking L-selectin had no effect on lymphocyte adhesion to activated endothelial cells, which was in contradiction with their in vivo observation. Upon addition of erythrocytes to
5695
the lymphocytes, L-selectin was then shown to play a significant role in adhesion especially under high shear rate. The use of more physiological experimental conditions (e.g., presence of red blood cells, plasma proteins) should result in significant advances in our knowledge on the effect of mechanical factors on thrombosis and hemostasis.
6.2. Endotoxin Endotoxins, also called LPS, are the component of the outer membrane of gram-negative bacteria and are released into the circulation upon disruption of the intact bacteria (death, cell lysis) [231]. Endotoxin is commonly found everywhere in our environment and it is the most significant pyrogen in parenteral drugs and medical devices. Endotoxins are also present in the digestive system. Their presence in the blood stream may cause septic reactions with a variety of symptoms such as fever, hypotension, nausea, shivering and shock [232]. High concentrations can lead to serious complications such as disseminated intravascular coagulation (DIC), endotoxin shock and adult respiratory distress syndrome (ARDS). Endotoxins are known to activate complement, the kinin system, leukocytes, platelets and endothelial cells [231,232] and are the "enemy" of both in vitro and in vivo study of blood-material interactions. In vivo, they may lead to the complications mentioned above, while in vitro, the presence of this contaminant may affect the results and compromise the conclusions. FDA regulates the acceptable level of endotoxin contamination with medical devices to be 0.5 endotoxin units/ml [233]. There have been few reports of endotoxin contamination with the use of cardiovascular devices. During cardiopulmonary bypass and extracorporeal membrane oxygenation, the presence of endotoxins has been observed in vivo [234,235]. They appear to originate mostly from the gut [236-238] rather than from the materials and are believed to be a reaction to the surgical procedure. During hemodialysis, endotoxin contamination is also an issue and the dialysate is usually the source [232,239]. While endotoxin contamination may be present in vivo in some patients and studies, there has been no investigation showing a significant correlation between the magnitude of endotoxin contamination and postoperative complications [235,240]. On the other hand, endotoxin contamination during in vitro work may be much more common, as sterile conditions are not always available and the laboratoryworking environment contributes to their presence. The most conspicuous source of endotoxin may actually be the water since distillation and deionizing columns do not remove endotoxin. Endotoxin has an effect on platelets only at high concentration (over l~tg/ml
234 5696
The Biomaterials Silver Jubilee Compendium M.B. Gorbet, M.V. Sefton / Biomaterials 25 (2004) 5681-5703
equivalent to 5000 EU/ml) [241], while leukocytes have been reported to be activated by endotoxin concentrations as low as 0.01 ng/ml (equivalent to 0.05 EU/ml) [242,243]. When studying blood-material interactions, endotoxin may be contained in buffers and/or on materials, and its priming and activating effect on leukocytes may affect the observed results. However, it is important to consider that all the studies performed on the effect of endotoxin on leukocyte activation used purified strains of endotoxin while the endotoxin present in laboratory materials and buffers are of an environmental nature. Purified endotoxins are much more potent than environmental endotoxins [244] and even among purified endotoxins, their activity might vary [242,245]. Contrary to a purified strain of endotoxin [239], the presence of relatively high levels of environmental endotoxins (100 EU/ml) was shown to have little impact on the leukocyte response to hemodialysis [246] in vivo. But when tested in vitro, environmental contamination of a material may have a dramatic effect on the results. In the orthopedic area, recent studies [258,259] have focussed on endotoxin contamination of microparticles used to assess the effect of wear debris in vitro and have confirmed the significant effect of environmental endotoxin contamination on cytokines. Many washing procedures are now available to ensure endotoxin removal from materials [247,248] and should ensure that the study of blood-material interactions is not impaired by the presence of endotoxin.
7. Conclusions
cytes; the ability of C D l l b to bind Factor X and fibrinogen; the ability of released inflammatory mediators to activate platelets and block inhibitors of coagulation; and by promoting the association between leukocytes and platelets. In the last 5 years, many leukocyte investigators have discussed the participation of inflammatory cells in coagulation [249-252]. Thrombosis is viewed now more as a multicellular event rather than just a platelet event [253]. In certain situations, blocking leukocyte contributions to thrombin generation may appear to be a reasonable means to reduce the occurrence of thrombotic complications. Such nontraditional approaches to thrombosis control with biomaterials may be a useful opportunity for further study. The mechanism of biomaterial-associated thrombosis is not fully clear. The role of Factor XII is uncertain while that of TF has not been directly assessed. Both the mechanisms of leukocyte and platelet activation by materials remain to be further elucidated. As noted above in the context of coagulation, the timing of the events contributing to thrombin formation is also a complex issue. Both Factor XII activation and platelet activation are able to generate thrombin formation within minutes while thrombin generation via leukocyte TF requires hours since TF has to be synthesized. The contribution of leukocyte proteases will also be affected by time since their effect will be dependent on the presence of inhibitors and other inflammatory mediators that can potentiate their action. As illustrated in Fig. 5, the time course of the underlying steps in biomaterialassociated thrombosis may need more consideration
The complexity of blood-material interactions explains our failure to design a material that is entirely blood-compatible. Our current stage of knowledge is far from providing us with a complete mechanism of material-induced thrombin generation. One issue has been the natural scientific tendency to focus on individual aspects of the whole problem rather than considering the various interactions. For example, the biomaterials community has typically looked at platelet interactions in platelet-rich plasma and so is unable to explore interactions between leukocytes and platelets. Alternatively, an anticoagulant prevents thrombin effects from being considered. Of course, without anticoagulants or with whole blood, the experiments get too complicated or sometimes impossible to perform or analyze. Unfortunately, simplifying the system has not allowed us to make real progress. We have also separated thrombosis from its normal context in inflammation and wound healing. The molecular links between inflammation and thrombosis are undeniable. Inflammation, as characterized by a leukocyte response to a stimulus, may contribute to thrombin generation by the TF expression on mono-
Fig. 5. Time scales of biomaterial-associated thrombogenicity. Each component is associated with a different time scale. Protein adsorption and Factor XII activation (assuming it is relevant) occurs within seconds of blood-material contact, producing low levels of thrombin. Platelet activation occurs within minutes creating the phospholipid surface required for assembly of the platelet bound coagulation enzymes and the production of enough thrombin to cause substantial fibrin formation. Leukocyte activation ( C D l l b upregulation) also occurs within minutes leading to adhesion while TF expression occurs over hours (as transcription and translation must occur first). Complement activation occurs at all these time scales, but whether it is the prelude to leukocyte activation and/or platelet activation is not clear; we suspect that these three are highly interlinked. The interplay between these components will vary depending on the time scale of the situation: a low time constant situation (e.g., high flow, straight tube) would involve different components than a high time constant (e.g., low flow, stagnation zone) situation.
235
The Biomaterials Silver Jubilee Compendium M.B. Gorbet, M.V. Sefton / Biomaterials 25 (2004) 5681-5703
than has hitherto been the case. Experimental studies already have difficulty making the transition from the first few seconds of protein adsorption to the first few minutes of coagulation and cell adhesion. What happens over hours to days as leukocytes synthesise TF and thrombotic deposits become remodeled is almost beyond current experimental capacity. Whether thrombosis leads to passivation or embolization or some other long-term consequence is still largely unknown. The solution to thrombotic complications associated with cardiovascular devices may not be to try to create a new material that will elicit the proper blood response: the inert cardiovascular biomaterial may be impossible. Rather, more success may be achieved by preventing the adverse effects of a biomaterial by actively blocking the pathway responsible for the inherent thrombogenicity of the materials. Rather than minimizing non-specific biomaterial-associated activation, active inhibition may be the only recourse. Despite more than 50 years of research on blood-material interactions, many questions remain unanswered. The intent of this review was to summarize what we know and to highlight what we have yet to learn.
[12]
[13]
[14]
[15]
[16]
[17]
[18]
[19]
References [1] Hanson SR. Device thrombosis and thromboembolism. Cardiovasc Pathol 1993;2:157S-65S. [2] Bittl JA. Coronary stent occlusion: thrombus horribilis. J Am Coll Cardiol 1996;28:368-70. [3] Clagett GP, Eberhart RC. Artificial devices in clinical practice. In: Colman RW, Hirsh J, Marder VJ, Salzman EW, editors. Hemostasis and thrombosis: basic principles and clinical practice. Philadelphia: Lippincott; 1994. p. 1486-505. [4] Bick RL. Hemostasis defects with cardiac surgery, general surgery and prosthetic devices. In: Bick RL, editor. Disorders of hemostasis and thrombosis. Chicago: American Society of Clinical Pathologist Press; 1992. p. 195-222. [5] Eberhart RC, Clagett CP. Catheter coatings, blood flow, and biocompatibility. Semin Hematol 1991;28:42-8. [6] Edmunds LHJ. Is prosthetic valve thrombogenicity related to design or material? Tex Heart Inst J 1996;23:24-7. [7] Geiser T, Sturzenegger M, Genewein U, Haeberli A, Beer JH. Mechanisms of cerebrovascular events as assessed by procoagulant activity, cerebral emboli and platelet microparticles in patients with prosthetic heart valves. Stroke 1998;29: 1770-7. [8] Mickelson JK, Lakkis NM, Villarreal-Levy G, Hughes BJ, Smith CW. Leukocyte activation with platelet adhesion after coronary angioplasty: a mechanism for recurrent disease? J Am Coll Cardiol 1996;28:345-53. [9] Wendel HP, Ziemer G. Coating techniques to improve the hemocompatibility of artificial devices used for extracorporeal circulation. Eur J Cardiothorac Surg 1999; 16:342-50. [10] Defraigne JO, Pincemail J, Larbuisson R, Blaffart F, Limet R. Cytokine release and neutrophil activation are not prevented by heparin-coated circuits and aprotinin administration. Ann Thorac Surg 2000;69:1084-91. [11] Moen O, Hogasen K, Fosse E, et al. Attenuation of changes in leukocyte surface markers and complement activation with
[20]
[21]
[22]
[23]
[24]
[25]
[26]
[27] [28] [29] [30]
5697
heparin-coated cardiopulmonary bypass. Ann Thorac Surg 1997;63:105-11. Baufreton C, Jansen PG, Le BP, et al. Heparin coating with aprotinin reduces blood activation during coronary artery operations. Ann Thorac Surg 1997;63:50-6. Muehrcke DD, McCarthy PM, Kottke-Marchant K, et al. Biocompatibility of heparin-coated extracorporeal bypass circuits: a randomized, masked clinical trial. J Thorac Cardiovasc Surg 1996;112:472-83. Gorman RC, Ziats N, Rao AK, et al. Surface-bound heparin fails to reduce thrombin formation during clinical cardiopulmonary bypass. J Thorac Cardiovasc Surg 1996;111:1-11. Redmond JM, Gillinov AM, Stuart RS, et al. Heparin-coated bypass circuits reduce pulmonary injury. Ann Thorac Surg 1993; 56:474-8. Levy M, Hartman AR. Heparin-coated bypass circuits in cardiopulmonary bypass: improved biocompatibility or not. Int J Cardiol 1996;53:$81-7. Fosse E, Thelin S, Svennevig JL, et al. Duraflo II coating of cardiopulmonary bypass circuits reduces complement activation, but does not affect the release of granulocyte enzymes: a European multicentre study. Eur J Cardiothorac Surg 1997;11: 320-7. Videm V, Mollnes TE, Bergh K. Heparin-coated cardiopulmonary bypass equipment. II. Mechanisms for reduced complement activation in vivo. J Thorac Cardiovasc Surg 1999;117: 803-9. Janvier G, Baquey C, Roth C, Benillan N, Belisle S, Hardy J-F. Extracorporeal circulation, hemocompatibility and biomaterials. Ann Thorac Surg 1996;62:1926-34. Videm V, Mollnes TE, Fosse E. Heparin-coated cardiopulmonary bypass equipment. I. Biocompatibility markers and development of complications in high risk population. J Thorac Cardiovasc Surg 1999;117:794-802. Gillinov AM, DeValeria PA, Winkelstein JA, et al. Complement inhibition with soluble complement receptor type 1 in cardiopulmonary bypass. Ann Thorac Surg 1993;55:619-24. Fitch JC, Rollins S, Matis L, et al. Pharmacology and biological efficacy of a recombinant, humanized, single-chain antibody C5 complement inhibitor in patients undergoing coronary artery bypass graft surgery with cardiopulmonary bypass. Circulation 1999; 100:2499-506. Murkin JM. Cardiopulmonary bypass and the inflammatory response; a role for serine protease inhibitor. J Cardiothorac Vasc Anesth 1997;11:19-23. Sundaram S, Gikakis N, Hack CE, et al. Nafamostat mesilate, a broad spectrum protease inhibitor, modulates platelet, neutrophil and contact activation in simulated extracorporeal circulation. Thromb Haemost 1996;75:76-82. Rao AK, Sun L, Gorman JH, Edmunds Jr LH. GPIIb/IIIa receptor antagonist Tirofiban inhibits thrombin generation during cardiopulmonary bypass in baboons. Thromb Haemost 1999;82:140-4. Vertress RA, Tao W, Kramer GC, et al. Tumor necrosis factor monoclonal antibody prevents alterations in leukocyte populations during cardiopulmonary bypass. ASAIO J 1994;40: M554-9. Schmaier AH. Contact activation: a revision. Thromb Haemost 1997;77:101-7. Furie B, Furie BC. The molecular basis of blood coagulation. Cell 1988;53:505-18. Blajchman MA, Ozge-Anwar AH. The role of the complement system in hemostasis. Prog Hematol 1986;XIV:149-82. Jetsy J, Nemerson Y. The pathways of blood coagulation. In: Beutler E, Lichtman MA, Coller BS, Kipps TJ, editors. Williams hematology. 5th ed. New York: McGraw-Hill; 1995. p. 1227-38.
236 5698
The Biomaterials Silver Jubilee Compendium M.B. Gorbet, M.V. Sefton / Biomaterials 25 (2004) 5681-5703
[31] Rapaport SI, Rao VM. Initiation and regulation of the tissue factor-dependent blood coagulation. Arterioscler Thromb 1992; 12:1111-21. [32] Morrissey JH, Macik BG, Neuenschwander PF, Comp PC. Quantitation of activated factor VII levels in plasma using a tissue factor mutant selectively deficient in promoting factor VII activation. Blood 1993;81:734-44. [33] Osterud B, Rapaport SI. Activation of Factor IX by the reaction product of tissue factor and Factor VIIa; additional pathway for initiating blood coagulation. Proc Natl Acad Sci USA 1977;74: 5260-4. [34] Lawson JH, Kalafatis M, Stram S, Mann KG. A model for the tissue factor pathway to thrombin. An empirical study. J Biol Chem 1994;269:23357-66. [35] Rand MD, Lock JB, van't Veer C, Gaffney DP, Mann KG. Blood clotting in minimally altered whole blood. Blood 1996;88: 3432-45. [36] Bauer KA, Rosenberg RD. Control of coagulation factors. In: Beutler E, Lichtman MA, Coller BS, Kipps TJ, editors. Williams hematology. 5th ed. New York: McGraw-Hill; 1995. p. 1239-50. [37] Hirsh J. Oral anticoagulant drugs. N Engl J Med 1991;324: 1865-75. [38] Novotny WF, Girard TJ, Miletich JP, Broze Jr GB. Platelets secrete a coagulation inhibitor functionally and antigenetically similar to the lipoprotein associated coagulation inhibitor. Blood 1988;72:2020-5. [39] McGee MP, Foster S, Wang X. Simultaneous expression of tissue factor and tissue factor pathway inhibitor by human monocytes. A potential mechanism for localized control blood coagulation. J Exp Med 1994;179:1847-54. [40] Cooke ED, Bowcock SA, Lloyd M J, Pilcher MJ. Intravenous lignocaine in prevention of deep venous thrombosis after elective surgery. The Lancet 1977;ii:797-9. [41] Higuchi DA, Wun TC, Likert KM, Broze GJ. The effect of leukocyte elastase on tissue factor pathway inhibitor. Blood 1992;79:1712-9. [42] Esmon CT. Regulatory mechanisms in hemostasis: natural anticoagulants. In: Colman RWet al,, editor. Hemostasis and thrombosis. Philadelphia: Lippincott; 1997. p. 1597-9. [43] Horbett TA. Principles underlying the role of adsorbed plasma proteins in blood interactions with foreign materials. Cardiovasc Pathol 1993;2:137S-48S. [44] Ziats N, Pankowsky DA, Tierney BP, Ratnoff OD, Anderson JM. Adsorption of Hageman Factor (Factor XII) and other human plasma proteins to biomedical polymers. J Lab Clin Med 1990;116:687-96. [45] Mulzer SR, Brash JL. Identification of plasma proteins adsorbed to hemodialyzers during clinical use. J Biomed Mater Res 1989; 23:1483-504. [46] Cornelius RM, Brash JL. Identification of proteins adsorbed to hemodialyser membranes from heparinized plasma. J Biomater Sci Polym Ed 1993;4:291-304. [47] Matata BM, Courtney JM, Sundaram S, et al. Determination of contact phase activation by the measurement of the activity of supernatant and membrane surface-adsorbed factor XII: its relevance as a useful parameter for the in vitro assessment of haemodialysis membranes. J Biomed Mater Res 1996;31:63-70. [48] Elam J-H, Nygren H. Adsorption of coagulation proteins from whole blood on to polymer materials: relation to platelet activation. Biomaterials 1992;13:3-8. [49] Van der Kamp KWHJ, van Oweren W. Factor XII fragment and kallikrein generation in plasma during incubation with biomaterials. J Biomed Mater Res 1994;28:349-52. [50] Van der Kamp KWHJ, Kauch KD, Feijen J, Horbett TA. Contact activation during incubation of five different polyurethanes or glass in plasma. J Biomed Mater Res 1995;29:1303-6.
[51] Hong J, Nilsson Ekdahl K, Reynolds H, Larsson R, Nilsson B. A new in vitro model to study interaction between whole blood and biomaterials. Studies of platelet and coagulation activation and the effect of aspirin. Biomaterials 1999;20:603-11. [52] Blezer R, Willems GM, Cahalan PT, Lindhout T. Initiation and propagation of blood coagulation at artificial surfaces studied in a capillary flow reactor. Thromb Haemost 1998;79:296-301. [53] Irvine L, Sundaram S, Courtney JM, Taggart DP, Wheatley DJ, Lowe GDO. Monitoring of Factor XII activity and granulocyte elastase release during cardiopulmonary bypass. ASAIO Trans 1991 ;XXXVII: 569-71. [54] Boisclair MD, Lane DA, Philippou H. Mechanisms of thrombin generation during surgery and cardiopulmonary bypass. Blood 1993;82:3350-7. [55] Boisclair MD, Philippou H, Lane DA. Thrombogenic mechanisms in the human: fresh insights obtained by immunodiagnostic studies of coagulation markers. Blood Coagul Fibrinolysis 1993;4:1007-21. [56] Burman JF, Chung HI, Lane DA, et al. Role of Factor XII in thrombin generation and fibrinolysis during cardiopulmonary bypass. Lancet 1994;344:1192-3. [57] Colman RW. Mechanisms of thrombus formation. Cardiovasc Pathol 1993;2:23S-31S. [58] Ratner BD. The blood compatibility catastrophe. J Biomed Mater Res 1993;27:283-7. [59] Edmunds Jr LH. Is prosthetic valve thrombogenicity related to design or material? Tex Heart Inst J 1996;23:24-7. [60] Chung JH, Gikakis N, Rao K, Drake TA, Colman RW, Edmunds LHJ. Pericardial blood activates the extrinsic coagulation pathway during clinical cardiopulmonary bypass. Circulation 1996;93:2014-8. [61] Wilhelm CR, Ristich J, Kormos RL, Wagner WR. Monocyte tissue factor expression and ongoing complement generation in ventricular assist devices patients. Ann Thorac Surg 1998;65: 1071-6. [62] Ernofsson M, Thelin S, Siegbahn A. Monocyte tissue factor expression, cell activation and thrombin formation during cardiopulmonary bypass: a clinical study. J Thorac Cardiovasc Surg 1997;113:576-84. [63] Gorbet MB, Sefton MV. Leukocyte activation and procoagulant activities, in preparation. [64] Gorbet MB, Sefton MV. Material-induced tissue factor expression but not CD1 l b upregulation depends on the presence of platelets. J Biomed Mater Res 2003;67A:792-800. [65] Basmadjian D, Sefton MV, Baldwin SA. Coagulation on biomaterials in flowing blood: some theoretical considerations. Biomaterials 1997; 18:1511-22. [66] Hirsh J, Weitz JI. Antithrombin therapy. In: Califf RM, editor. Acute myocardial infarction and other acute ischemic syndromes. St Louis: Mosby; 1996. [67] Hirsh J, Raschke R, Warkentin TE, Dalen JE, Deykin D, Poller L. Heparin: mechanism of action, pharmacokinetics, dosing considerations, monitoring, efficacy and safety. Chest 1995;108: 258S-75S. [68] Rubens FD, Weitz JI, Brash JL, Kinlough-Rathbone RL. The effect of antithrombin III-independent thrombin inhibitors and heparin on fibrin accretion onto fibrin-coated polyethylene. Thromb Haemost 1993;69:130-4. [69] Gustafsson D, Elg M. The pharmacodynamics and pharmacokinetcis of the oral direct thrombin inhibitor ximelagatran and its active metabolite melagatran: a mini-review. Thromb Res 2003;109:$9-15. [70] Edens RE, Lindhart RJ, Weiler JM. Heparin is not just an anticoagulant anymore: six and one-half decades of studies on the ability of heparin to regulate complement activity. Complement Profiles 1993;1:96-120.
The Biomaterials Silver Jubilee Compendium
237
M.B. Gorbet, M.V. Sefton / Biomaterials 25 (2004) 5681-5703
[71] Misoph M, Schwender S, Babin-Ebell J. Response of the cellular immune system to cardiopulmonary bypass is independent of the applied pump type and of the use of heparin-coated surfaces. Thorac Cardiovasc Surg 1998;46:222-7. [72] Remes R, Williams DF. Immune response in biocompatibility. Biomaterials 1992;13:731-43. [73] Hakim RM. Complement activation by biomaterials. Cardiovasc Pathol 1993;2:187S-97S. [74] Kazatchkine MD, Carreno M-P. Activation of the complement system at the interface between blood and artificial surfaces. Biomaterials 1988;9:30-5. [75] Johnson RJ. Complement activation during extracorporeal therapy: biochemistry, cell biology and clinical relevance. Nephrol Dial Transplant 1994;9:36-45. [76] Chenoweth DE. Complement activation during hemodialysis: clinical observations, proposed mechanisms, and theoretical implications. Artif Organs 1984;8:281-90. [77] Agostini A, Gardinali M. Complement activation during hemodialysis. J Biomater Appl 1989;4:102-22. [78] Videm V, Fosse E, Mollnes TE, Garred P, Svennevig JL. Time for new concepts about measurement of complement activation by cardiopulmonary bypass? Ann Thorac Surg 1992;54:725-31. [79] Kottke-Marchant K, Anderson JM, Miller KM, Marchant RE, Lazarus H. Vascular graft-associated complement activation and leukocyte adhesion in an artificial circulation. J Biomed Mater Res 1987;21:379-97. [80] Shepard AD, Gelfand JA, Callow AD, O'Donnell Jr TF. Complement activation by synthetic vascular grafts. J Vasc Surg 1984;1:829-38. [81] Johnson RJ. Immunology and the complement system. In: Ratner BD, Hoffman AS, Schoen FJ, Lemons JE, editors. Biomaterials science. San Diego: Academic Press; 1996. p. 173-88. [82] Craddock PR, Fehr J, Dalmasso AP, Brighan KL, Jacob HS. Hemodialysis leukopenia. Pulmonary vascular leukostasis resulting from complement activation by dialyzer cellophane membranes. J Clin Invest 1977;59:879-88. [83] Tulunay M, Demiralp S, Tatsan S, et al. Complement (C3, C4) and C-reactive protein responses to cardiopulmonary bypass and protamine administration. Anaesth Intens Care 1993;21:50-5. [84] Lhotta K, Wurzner R, Kronenberg F, Opperman M, Konig P. Rapid activation of the complement system by cuprophane depends on complement component C4. Kidney Int 1998;53: 1044-51. [85] Thylen P, Fernvik E, Lundhal J, Hed J, Jacobson SH. Modulation of CD1 lb/CD18 on monocytes and neutrophils following hemodialysis membrane interaction in vitro. Int J Artif Organs 1996; 19:156-63. [86] Nilsson UR, Larm O, Nilsson B, Storm KE, Elwing H, Nilsson Ekdahl K. Modification of the complement binding properties of polystyrene; effects of end-point heparin attachment. Scand J Immunol 1993;37:349-54. [87] Sefton MV, Gemmell CH, Gorbet MB. What really is blood compatibility. J Biomater Sci Polym Ed 2000;11:1165-82. [88] Gemmell CH. Platelet adhesion onto artificial surfaces: inhibition by benzamidine, pentamidine, and pyridoxal-5-phosphate as demonstrated by flow cytometric quantification of platelet adhesion to microspheres. J Lab Clin Med 1998;131:84-92. [89] Gemmell CH. Activation of platelets by in vitro whole blood contact with materials: increases in microparticle, procoagulant activity, and soluble P-selectin blood levels. J Biomater Sci Polym Ed 2001;12(8):933-43. [90] Gemmell CH, Yeo EL, Sefton MV. Flow cytometric analysis of material-induced platelet activation in a canine model: elevated microparticle levels and reduced platelet life span. J Biomed Mater Res 1997;37:176-81.
5699
[91] Blockmans D, Deckmyn H, Vermylen J. Platelet activation. Blood Rev 1995;9:143-56. [92] Marcus AJ. Platelet activation. In: Fuster V, Ross R, Topol EJ, editors. Atherosclerosis and coronary artery disease. Philadelphia: Lippincott-Raven Publishers; 1996. p. 607-37. [93] Rinder HM, Tracey JL, Rinder CS, Leitenberg D, Smith BR. Neutrophil but not monocyte activation inhibits P-selectinmediated platelet adhesion. Thromb Haemost 1994;72: 750-6. [94] Rinder HM, Bonan JL, Rinder CS, Ault KA, Smith BR. Dynamics of leukocyte-platelet adhesion in whole blood. Blood 1991;78:1730-7. [95] Sims PJ, Wiedmer T. Induction of cellular procoagulant activity by the membrane attack complex of complement. Semin Cell Biol 1995;6:275-82. [96] Jy W, Mao W-W, Horstman LL, Tao J, Ahn YS. Platelet microparticles bind, activate and aggregate neutrophils in vitro. Blood Cells Mol Dis 1995;21:217-31. [97] Siljander P, Carpen O, Lassila R. Platelet derived microparticles associate with fibrin during thrombosis. Blood 1996;87:4651-63. [98] Holme PA, Solum NO, Brosstad F, Pedersen T, Kveine M. Microvesicles bind soluble fibrinogen, adhere to immobilized fibrinogen and coaggregates with platelets. Thromb Haemost 1998;79:389-94. [99] Tans G, Rosing J, ChristeUa M, et al. Comparison of anticoagulant and procoagulant activities of stimulated platelets and platelet-derived microparticles. Blood 1991;77:2641-8. [100] Nieuwland R, Bercmans RJ, Rotteveel-Eijkman RC, et al. Cellderived microparticles generated in patients during cardiopulmonary bypass are highly procoagulant. Circulation 1997;96: 3534-41. [101] Gemmell CH. Assessment of material-induced procoagulant activity by a modified Russell viper venom coagulation time test. J Biomed Mater Res 1998;42:611-6. [102] Ware JA, Coller BS. Platelet morphology, biochemistry and function. In: Beutler E, Lichtman MA, Coller BS, Kipps TJ, editors. Williams hematology. New York: McGraw-Hill; 1995. p. 1161-225. [103] Calvete JJ. Clues for understanding the structure and function of a prototypic human integrin: the platelet glycoprotein IIb/IIIa complex. Thromb Haemost 1994;72:1-15. [104] van der Wieken LR. Stents and IIb/IIIa receptor blockers combined: usefulness in various types of coronary artery disease. Semin Interv Cardiol 1999;4:77-83. [105] Ellis SG, Bates ER, Schaible T, Weisman HF, Pitt B, Topol EJ. Prospects for the use of antagonists to the platelet glycoprotein IIb/IIIa receptor to prevent postangioplasty restenosis and thrombosis. J Am Coll Cardiol 1991;17:89B-95B. [106] Mickelson JK, Ali MN, Kleiman NS, et al. Chimeric 7E3 Fab (Reopro) decreases detectable C D l l b on neutrophils from patients undergoing coronory angioplasty. J Am Coll Cardiol 1999;33:97-106. [107] Coller BS, Anderson KM, Weisman HF. The anti-GPIIb/IIIa agents: fundamental and clinical aspects. Haemostasis 1996;26: 285-93. [108] The EPIC Investigators. Use of a monoclonal antibody directed against the platelet glycoprotein IIb/IIIa receptor in high risk coronary angioplasty. N Engl J Med 1994;330:956-61. [109] Topol EJ, Lincoff AM, Kereiakes DJ, et al. Multi-year follow-up of Abciximab therapy in three randomized, placebo controlled trials of percutaneous coronary revascularization. Am J Med 2002;113:1-6. [110] Spinler SA, Hilleman DE, Cheng JWM, et al. New recommendations from the 1999 American College of CArdiology/ American Heart Association acute myocardial infarction guidelines. Ann Pharmacotherapy 2002;35:589-617.
238 5700
The Biomaterials Silver Jubilee Compendium M.B. Gorbet, M.V. Sefton / Biomaterials 25 (2004) 5681-5703
[111] Cote AV, Berger PB, Holmes Jr DR, Scott CG, Bell MR. Hemorrhagic complications after percutaneous coronary interventions with adjunctive Abciximab. Mayo Clin Proc 2002;76: 890-6. [112] Makoni SN. Acute profound thrombocytopenia following angioplasty: the dilemma in the management and a review of the literature. Heart 2002;86:e18-9. [113] Grunkemeier JM, Tsai WB, Horbett TA. Hemocompatibility of treated polystyrene substrates: contact activation, platelet adhesion, and procoagulant activity of adherent platelets. J Biomed Mater Res 1998;41:657-70. [114] Godo MN, Sefton MV. Characterization of transient platelet contacts on a polyvinyl alcohol hydrogel by video microscopy. Biomaterials 1999;20:1117-26. [115] Sheppard JI, McClung WG, Feuerstein IA. Adherent platelet morphology on adsorbed fibrinogen: effects of protein incubation time and albumin addition. J Biomed Mater Res 1994;28:1175-86. [116] Chinn JA, Ratner BD, Horbett TA. Adsorption of baboon fibrinogen and the adhesion of platelets to a thin film polymer deposited by radio-frequency glow discharge of allylamine. Biomaterials 1992;13:322-32. [117] Hanson SR, Harker LA. Blood coagulation and blood-materials interactions. In: Ratner BD, Hoffman AS, Schoen FJ, Lemons JE, editors. Biomaterials science. San Diego: Academic Press; 1996. p. 193-9. [118] Tsai WB, Grunkemeier JM, Horbett TA. Human plasma fibrinogen adsorption and platelet adhesion to polystyrene. J Biomed Mater Res 1999;44:130-9. [119] Bailly AL, Laurent A, Lu H, et al. Fibrinogen binding and platelet retention: relationship with the thrombogenicity of catheters. J Biomed Mater Res 1996;30:101-8. [120] Gemmell CH, Ramirez SM, Yeo EL, Sefton MV. Platelet activation in whole blood by artificial surfaces: identification of platelet-derived microparticles and activated platelet binding to leukocytes as material-induced activation events. J Lab Clin Med 1995;125:276-87. [121] Cholakis CH, Zingg W, Sefton MV. Effect of heparin-PVA hydrogel on platelets in a chronic canine AV shunt. J Biomed Mater Res 1989;23:417-41. [122] Hanson SR, Harker LA, Ratner BD, Hoffman AS. In vivo evaluation of artificial surfaces with a nonhuman primate model of arterial thrombosis. J Lab Clin Med 1980;95: 289-304. [123] Ip WF, Sefton MV. Platelet consumption by NHLBI reference materials and silastic. J Biomed Mater Res 1991;25:1321-4. [124] Hanson SR, Harker LA. Effects of platelet-modifying drugs on arterial thromboembolism in baboons. Aspirin potentiates the antithrombotic actions of dipyridamole and sulfinpyrazone by mechanisms independent of cyclooxygenase inhibition. J Clin Invest 1999;75:1591-9. [125] Weily HS, Steele PP, Davies H, Pappas G, Genton E. Platelet survival inpatients with substitute heart valves. N Eng J Med 1974;290:534-6. [126] Llanos GR, Sefton MV. Immobilization of poly(ethylene glycol) onto poly(vinyl alcohol) hydrogel: 2. Evaluation of thrombogenicity. J Biomed Mater Res 1993;27:1383-91. [127] Wachtfogel YT, Bischoff R, Bauer R, et al. Alpha 1-antitrypsin Pittsburgh (Met358~Arg) inhibits the contact pathway of intrinsic coagulation and alters the release of human neutrophil elastase during simulated extracorporeal circulation. Thromb Haemost 1994;72:843-7. [128] Simon P, Ang KS, Cam G. Enhanced platelet aggregation and membrane biocompatibility: possible influence on thrombosis and embolism in haemodialysis patients. Nephron 1987;45: 172-3.
[129] Hakim RM, Schafer A1. Hemodialysis-associated platelet activation and thrombocytopenia. Am J Med 1985;78:575-80. [130] Peerschke EIB, Ghebrehiwet B. Platelet receptors for the complement component Clq: implication for hemostasis and thrombosis. Immunobiology 1998;199:239-49. [131] Bing DH. Inhibition of guinea pig complement by aromatic amidine and guanidine compounds. J Immunol 1970;105: 1289-91. [132] Andrew JM, Roman Jr DP, Bing DH. Inhibition of four human serine proteases by substituted benzamidines. J Medicinal Chem 1978;21:1202-7. [133] Hauptmann J, Markwardt F. Inhibition of the haemolytic complement activity by derivatives of benzamidine. Biochem Pharmacol 1977;26:325-9. [134] Cox D, Motoyama Y, Seki J, Aoki T, Dohi M, Yoshida K. Pentamidine: a non-peptide GPIIb/IIIa antagonist--in vitro studies on platelets from humans and other species. Thromb Haemost 1992;68:731-6. [135] Cox D, Aoki T, Seki J, Motoyama Y, Yoshida K. Pentamidine is a specific, non-peptide, GPIIb/IIIa antagonist. Thromb Haemost 1996;75:503-9. [136] Allan R, Rodrick M, Knobel HR, Isliker H. Inhibition of the interaction between the complement component C lq and immune complexes. Int Arch Allergy Appl Immunol 1979;58: 140-8.
[137] Subbarao K, Kuchubhotla J, Kakkar VV. Pyridoxal 5'phosphate--a new physiological inhibitor of blood coagulation and platelet function. Biochem Pharmacol 1979;28:531-4. [138] Takada Y, Arimoto Y, Mineda H, Takada A. Inhibition of the classical and alternative pathways by amino acids and their derivatives. Immunology 1978;34:509-15. [139] Rinder CS, Rinder HM, Smith M J, et al. Selective blockade of membrane attack complex formation during simulated extracorporeal circulation inhibits platelet but not leukocyte activation. J Thorac Cardiovasc Surg 1999;118:460-6. [140] Rinder CS, Rinder HM, Smith BR, et al. Blockade of C5a and C5b-9 generation inhibits leukocyte and platelet activation during extracorporeal circulation. J Clin Invest 1995;96: 1564-72. [141] Gorbet MB, Sefton MV. Role of complement and platelets in leukocyte activation induced by polystyrene and PEG-immobilized polystyrene beads in whole blood. J Biomed Mater Res, in preparation. [142] Larsson R, Elgue G, Larsson A, Ekdahl KN, Nilsson UR, Nilsson B. Inhibition of complement activation by soluble recombinant CR1 under conditions resembling those in a cardiopulmonary circuit: reduced up-regulation of CD1 l b and complete abrogation of binding of PMNs to the biomaterial surface. Immunopharmacology 1997;38:119-27. [143] Cassatella MA. The production of cytokines by PMN. Immunol Today 1995;16:21-6. [144] Anderson JM. Mechanisms of inflammation and infection with implanted devices. Cardiovasc Pathol 1993;2:33S-41S. [145] Del Maschio A, Corvazier E, Maillet F, Kazatchkine MD, Maclouf J. Platelet dependent induction and amplification of polymorphonuclear leukocyte lysosomal enzyme release. Br J Haematol 1989;72:329-35. [146] Bazzoni G, Dejana E, Del Maschio A. Platelet-dependent modulation of neutrophil function. Pharmacol Res 1992;26: 269-72. [147] Del Maschio A, Dejana E, Bazzoni G. Bidirectional modulation of platelet and polymorphonuclear leukocyte activities. Ann Hematol 1993;67:23-31. [148] Rebuck N, Finn A. Polymorphonuclear granulocyte expression of CDlla/CD18, CDllb/CD18 and L-selectin in normal individuals. FEMS Immunol Med Microbiol 1994;8:189-96.
239
The Biomaterials Silver Jubilee Compendium M.B. Gorbet, M.V. Sefton / Biomaterials 25 (2004) 5681-5703
[149] Camerer E, Kolsto A-B, Pridz H. Cell biology of tissue factor, the principal initiator of coagulation. Thromb Res 1996;81:1-41. [150] Smith JA. Neutrophils, host defense, and inflammation; a double-edged sword. J Leukocyte Biol 1994;56:686. [151] Kuijipers TW, Hakkert BC, Van Mourik JA, Roos D. Distinct adhesive properties of granulocytes and monocytes to endothelial cells under static and stirred conditions. J Immunol 1990; 145:2588-94. [152] van Ginkel CJW, van Haken WG, Oh JIH, Vreeken J. Stimulation of monocyte procoagulant activity by adherence to different surfaces. Br J Haematol 1977;37:35-45. [153] Gorbet MB, Yeo EL, Sefton MV. Flow cytometric study of in vitro neutrophil activation by biomaterials. J Biomed Mater Res 1999;44:289-97. [154] Ginis I, Tauber AI. Activation mechanisms of adherent human neutrophils. Blood 1990;76:1233-9. [155] Jones DA, Smith CW, McIntire LV. Leukocyte adhesion under flow conditions: principles important in tissue engineering. Biomaterials 1996;17:337-47. [156] Butcher EC. Leukocyte-endothelial cell recognition: three or (more) steps to specificity and diversity. Cell 1991;67:1033-6. [157] Arnaout MA. Structure and function of the leukocyte adhesion molecules CD11/CD18. Blood 1990;75:1037-50. [158] Philips MR, Buyon JP, Winchester R, Weissmann G, Abramson SB. Upregulation of the iC3b receptor is neither necessary nor sufficient to promote neutrophil aggregation. J Clin Invest 1988;82:495-501. [159] Takala AJ, Jousela IT, Takkumen OS, et al. Time course of/~2 integrin CD1 lb/CD18 upregulation on neutrophils and monocytes after coronary artery bypass grafting. Scand J Thorac Cardiovasc Surg 1996;30:141-8. [160] Serrano CV, Ramires JA, Venturinelli M, et al. Coronary angioplasty results in leukocyte and platelet activation with adhesion molecules expression. J Am Coll Cardiol 1997;29: 1276-83. [161] Inque T, Sakai Y, Morooka S, Hayashi T, Takayanagi K, Takabatake Y. Expression of polymorphonuclear adhesion molecules and its clinical significance in patients treated with percutaneous transluminal coronary angioplasty. J Am Coll Cardiol 1996;28:1127-33. [162] Neumann FJ, Ott I, Gawaz M, Puchner G, Sch6mig A. Neutrophil and platelet activation at balloon-injured coronary artery plaque in patients undergoing angioplasty. J Am Coll Cardiol 1996;27:819-24. [163] Rousseau Y, Carreno M-P, Poignet J-L, Kaztchkine MD, Haeffner-Cavaillon N. Dissociation between complement activation, integrin expression and neutropenia during hemodialysis. Biomaterials 1999;20:1959-67. [164] Cristol JP, Canaud B, Rabesandratana H, Gaillard I, Serre A, Mion C. Enhancement of reactive oxygen species production and cell surface markers expression due to hemodialysis. Nephrol Dial Transplant 1994;9:389-94. [165] Von Appen K, Goolsby C, Mehl P, Goewert R, Ivanovich P. Leukocyte adhesion molecule as biocompatibility markers for hemodialysis membranes. ASAIO J 1994;40:M609-15. [166] Rinder C, Fitch J. Amplification of the inflammatory response: adhesion molecules associated with platelet/white cell responses. J Cardiovasc Pharmacol 1996;27(Suppl. 1):$6-12. [167] E1 Habbal MH, Carter H Smith L, Elliot MJ, Strobel S. Neutrophil activation in paediatric extracorporeal circuits: effect of circulation and temperature variation. Cardiovasc Res 1995;29:102-7. [168] Gillinov AM, Bator JM, Zehr KJ, et al. Neutrophil adhesion molecule expression during cardiopulmonary bypass with bubble and membrane oxygenators. Ann Thorac Surg 1993;56: 847-53.
5701
[169] Cameron DE. Initiation of white cell activation during cardiopulmonary bypass: cytokines and receptors. J Cardiovasc Pharmacol 1996;27:S1-5. [170] Asimakopoulos G, Taylor KM. Effects of cardiopulmonary bypass on leukocyte and endothelial adhesion molecules. Ann Thorac Surg 1998;66:2135-44. [171] Horl WH, Riegel W, Steinhauer HB, et al. Granulocyte activation during hemodialysis. Clin Nephrol 1986;26: $30-4. [172] Grooteman MP, van TA, van HA, et al. Hemodialysis-induced degranulation of polymorphonuclear cells: no correlation between membrane markers and degranulation products. Nephron 2000;85:267-74. [173] Haag-Weber M, Schollmeyer P, Horl WH. Beta-2-microglobulin and main granulocyte components in hemodialysis patients. Artif Organs 1989; 13:92-6. [174] Stegmayr BG, Esbensen K, Gutierrez A, et al. Granulocyte elastase, /~ thromboglobulin, and C3d during acetate or bicarbonate hemodialysis with hemophan compared to a cellulose acetate membrane. Int J Artif Organs 1992;15:10-8. [175] Wachtfogel YT, Kucich U, Greenplate J, et al. Human neutrophil degranulation during extracorporeal circulation. Blood 1987;69:324-30. [176] Peek GJ, Firmin RK. The inflammatory and coagulative response to prolonged extracorporeal membrane oxygenation. ASAIO J 1999;45:250-63. [177] Frering B, Philip I, Dehoux M, Rolland C, Langlois JM, Desmont JM. Circulating cytokines in patients undergoing normothermic cardiopulmonary bypass. J Thorac Cardiovasc Surg 1994;108:636-41. [178] Himmelfarb J, Lazarus JM, Hakim R. Reactive oxygen species production by monocytes and polymorphonuclear leukocytes during dialysis. Am J Kidney Dis 1991;XVII:271-6. [179] Kaplan SS, Basford RE. Mechanisms of biomaterial induced superoxide release by neutrophils. J Biomed Mater Res 2000;28: 377-86. [180] Kuwahara T, Markert M, Wauters JP. Proteins adsorbed on hemodialysis membranes modulate neutrophil activation. Artif Organs 1989;13:427-31. [181] Serruys PW, Strauss BH, van Beusekom HM, van der Giessen WJ. Stenting of coronary arteries: has a modern Pandora's box been opened. J Am Coll Cardiol 1991;17:143B-54B. [182] van Beusekom HMM, van der Giessen WJ, Van Suylen RJ, Bos E, Bosman FT, Serruys PW. Histology after stenting of human saphenous vein bypass grafts: observations from surgically excised grafts 3-320 days after stent implantation. J Am Coll Cardiol 1993;21:45-54. [183] Jobes DR. Safety issues in heparin and protamine administration for extracorporeal circulation. J Cardiothorac Vasc Anesth 1998;12:17-20. [184] Grooteman MPC, Bos JC, Van Houte AJ, van Limbeek J, Schoorl M, Nube MJ. Mechanisms of intra-dialyser granulocyte activation: a sequential dialyser elution study. Nephrol Dial Transplant 1997;12:492-9. [185] Castiglione A, Pagliaro P, Romagnoni M, Beccari M, Cavaliere G, Veneroni G. Flow cytometric analysis of leukocytes eluted from haemodialysers. Nephrol Dial Transplant 1991;(Suppl. 2):31-35. [186] Gemmell CH, Black JP, Yeo EL, Sefton MV. Material-induced up-regulation of leukocyte CD1 l b during whole blood contact: material differences and a role for complement. J Biomed Mater Res 1996;32:29-35. [187] Gorbet MB, Sefton MV. Expression of procoagulant activities on Leukocytes following contact with polystyrene and PEG grafted polystyrene beads. J Lab Clin Med 2001;137: 345-55.
240 5702
The Biomaterials Silver Jubilee Compendium M.B. Gorbet, M.V. Sefton / Biomaterials 25 (2004) 5681-5703
[188] Kao WJ, Sapatnekar S, Hiltner A, Anderson JM. Complementmediated leukocyte adhesion on poly(etherurethane ureas) under shear stress in vitro. J Biomed Mater Res 1996;32:99-109. [189] McNally AK, Anderson JM. Complement C3 participation in monocyte adhesion to different surfaces. Proc Natl Acad Sci USA 1994;91:10119-23. [190] Cheung AK, Hohnholt M, Gilson J. Adherence of neutrophils to haemodialysis membranes: roles of complement receptors. Kidney Int 1991;40:1123-33. [191] Tang L, Lucas AH, Eaton JW. Inflammatory responses to implanted polymeric biomaterials: role of surface adsorbed immunoglobulin G. J Lab Clin Med 1993;122:292-300. [192] Tang L, Eaton JW. Fibrinogen mediates acute inflammatory responses to biomaterials. J Exp Med 1993;178:2147-56. [193] Eriksson O, Nygren H. Polymorphonuclear leukocytes in coagulating whole blood recognize hydrophilic and hydrophobic titanium surfaces by different adhesion receptors and show different patterns of receptor expression. J Lab Clin Med 2001; 137:296-302. [194] Gemmell CH. Flow cytometric evaluation of material-induced platelet and complement activation. J Biomater Sci Polym Ed 2000;11(11):1197-210. [195] Morley DJ, Feuerstein IA. Adhesion of polymorphonuclear leukocytes to protein-coated and platelet adherent surfaces. Thromb Haemost 1989;62:1023-8. [196] Herzlinger GA, Cumming RD. Role of complement activation in cell adhesion to polymer blood contact surfaces. ASAIO Trans 1980;XXXVI: 165-71. [197] Wachtfogel YT, Kettner C, Hack CE, et al. Thrombin and human plasma kallikrein inhibition during simulated extracorporeal circulation block platelet and neutrophil activation. Thromb Haemost 1998;80:686-91. [198] Wachtfogel YT, Kucich U, Hack CE, et al. Aprotinin inhibits the contact, neutrophil, and platelet activation systems during simulated extracorporeal perfusion. J Thorac Cardiovasc Surg 1993;106:1-9. [199] Himmelfarb J, Holbrook D, McMonagle E. Effects of aprotinin on complement and granulocyte activation during ex vivo hemodialysis. Am J Kidney Dis 1994;24:901-6. [200] Himmelfarb J, McMonagle E, Holbrook D, Toth C. Soluble complement receptor 1 inhibits both complement and granulocyte activation during ex vivo hemodialysis. J Lab Clin Med 1995;126:392-400. [201] Gillinov AM, Redmond JM, Winkelstein JA, et al. Complement and neutrophil activation during cardiopulmonary bypass: a study in the complement-deficient dog. Ann Thorac Surg 1994;57:345-52. [202] Finn A, Morgan BP, Rebuck N, et al. Effects of inhibition of complement activation using recombinant soluble receptor 1 on neutrophil CD 11b/CD 18 and L-selectin expression and release of interleukin-8 and elastase in simulated cardiopulmonary bypass. J Thorac Cardiovasc Surg 1996;111:451-9. [203] Rinder CS, Rinder HM, Johnson K, et al. Role of C3 cleavage in monocyte activation during extracorporeal circulation. Circulation 1999;100:553-8. [204] Gorbet MB, Sefton MV. The importance of platelets and complement in material-induced Leukocyte activation in vitro. PhD Thesis, University of Toronton, 2001. [205] Kappelmayer J, Bernabei A, Edmunds Jr LH, Edgington TS, Colman RW. Tissue factor is expressed on monocytes during simulated extracorporeal circulation. Circ Res 1993;72: 1075-81. [206] Barstad RM, Hamers MJAG, Moller A-S, Sakariassen KS. Monocyte procoagulant activity induced by adherence to an artificial surface is reduced by end-point immobilized heparin coating of the surface. Thromb Haemost 1998;79:302-5.
[207] Stahl RF, Fisher CA, Kucich U, et al. Effects of simulated extracorporeal circulation on human leukocyte elastase release, superoxide generation and procoagulant activity. J Thorac Cardiovasc Surg 1991;101:230-9. [208] Parratt R, Hunt BJ. Direct activation of factor X by monocytes occurs during cardiopulmonary bypass. Br J Haematol 1998; 101:40-6. [209] Rinder CS, Bonan JL, Rinder HM, Mathew J, Hines R, Smith BR. Cardiopulmonary bypass induces leukocyte-platelet adhesion. Blood 1992;79:1201-5. [210] Gawaz M, Bogner C. Changes in platelet membrane glycoproteins and platelet leukocyte interactions during hemodialysis. J Clin Invest 1994;72:424-9. [211] May AE, Neumann F-J, Gavaz M, Ott I, Walter H, Sch6mig A. Reduction of monocyte-platelet interaction and monocyte activation in patients receiving antiplatelet therapy after coronary stent implantation. Eur Heart J 1997;18:1913-20. [212] Palabrica T, Lobb R, Furie BC, et al. Leukocyte accumulation promoting fibrin deposition is mediated in vivo by P-selectin on adherent platelets. Nature 1992;359:848-51. [213] Toombs CF, De Graaf GL, Martin JP, Geng JG, Anderson DC, Shebuski RJ. Pretreatment with a blocking antibody to P-sel accelerates pharmacological thrombolysis in a primate model of arterial thrombosis. J Pharmacol Exp Ther 1995;275:941. [214] Turitto VT, Weiss HJ, Baumgartner HR. The effect of shear rate on platelet interaction with subendothelium exposed to citrated human blood. Microvasc Res 1980;19:352-65. [215] Turitto VT, Hall CL. Mechanical factors affecting hemostasis and thrombosis. Thromb Res 1998;92:$25-31. [216] Hanson SR, Sakariassen KS. Blood flow and antithrombotic drug effects. Am Heart J 1998;135:S132-45. [217] O'Brien JR. Shear induced platelet aggregation. Lancet 1990;335:711-3. [218] Oda A, Yokoyama K, Murata M, et al. Protein tyrosine phosphorylation in human platelets during shear-stress induced aggregation (SIPA) is regulated by GPIb/IX as well as GPIIb/ IIIa and requires intact cytoskeleton and endogenous ADP. Thromb Haemost 1995;74:736-42. [219] Kujiper PHM, Gallardo Torres HI, van der Linden JAM, et al. Platelet-dependent primary hemostasis promotes selectin- and integrin-mediated neutrophil adhesion to damaged endothelium under flow conditions. Blood 1996;87:3271-81. [220] Yeo EL, Sheppard J-OI, Feuerstein IA. Role of P-selectin and leukocyte activation in polymorphonuclear cell adhesion to surface adherent activated platelets under physiological shear conditions (an injury vessel wall model). Blood 1994;83: 2498-507. [221] Weber C, Springer TA. Neutrophil accumulation on activated, surface-adherent platelets in flow is mediated by interaction of MAc-1 with fibrinogen bound to ~IIb/~3 and stimulated by platelet activating factor. J Clin Invest 1997;100: 2085-93. [222] Kujiper PHM, Gallardo Torres HI, Lammers J-WJ, Sixma JJ, Koenderman L, Zwaginga JJ. Platelet and fibrin deposition at the damaged vessel wall: cooperative substrates for neutrophil adhesion under flow conditions. Blood 1997;89:166-75. [223] Bruil A, Sheppard JI, Feijen J, Feuerstein IA. In vitro leukocyte adhesion to modified polyurethane surfaces: III Effects of flow, fluid medium and platelets on PMN adhesion. In: Cooper SL, Bamford CH, Tsuruta T, editors. Polymer biomaterials in solution, as interfaces and as solid. Zeist, The Netherlands: VSP; 1995. p. 357-71. [224] Kujiper PHM, Gallardo Torres HI, van der Linden JAM, et al. Neutrophil adhesion to fibrinogen and fibrin under flow conditions is diminished by activation and L-selectin shedding. Blood 1997;89:2131-8.
241
The Biomaterials Silver Jubilee Compendium M.B. Gorbet, M.V. Sefton / Biomaterials 25 (2004) 5681-5703 [225] Hernandez MR, Escolar G, Bozzo J, Galan AM, Ordinas A. Inhibition of fibrin deposition on the subendothelium by a monoclonal antibody to polymorphonuclear leukocyte integrin CDllb. Studies in a flow system. Haematologica 1997;82: 566-71. [226] Bahra P, Nash GB. Sparsely adherent platelets support capture and immobilization of flowing neutrophils. J Lab Clin Med 1998;132:223-8. [227] Gemmell CH, Nemerson Y, Turitto VT. The effects of shear rate on the enzymatic activity of the tissue factor-factor VIIa complex. Microvasc Res 1990;40:327-40. [228] Nemerson Y, Contino PB. Tissue factor, flow and the initiation of coagulation. In: Seghatchian MJ, Samana MM, Hecker SP, editors. Hypercoagulable states. Boca Raton, FL: CRC Press; 1996. p. 21-8. [229] Bluestein D, Rambod E, Gharib M. Vortex shedding as a mechanism for free emboli formation in mechanical heart valves. J Biomech Eng 2000; 122:125-34. [230] Melder RJ, Munn LL, Yamada S, Ohkubo C, Jain RK. Selectin and integrin-mediated T lymphocyte rolling and arrest on TNFc~activated endothelium: augmentation by erythrocytes. Biophys J 1995;69:2131-8. [231] Lynn WA, Golenbock DT. Lipopolysaccharide antagonists. Immunol Today 1992;13:271-6. [232] Lemke HD. Methods for the detection of endotoxins present during extracorporeal circulation. Nephrol Dial Transplant 1994;9:90-5. [233] US Department of Health and Human Services/Public Health Services/Food and Drug Administration. Guideline on validation of the limulus amebocyte lysate test as an end-product endotoxin test for human and animal parental drugs, biological products and medical devices. 1987. p. 1-30. Available at http:// www.fda.gov/cber/gdlns/lal.pdf [234] Hirthler M, Simoni J, Dickson M. Elevated levels of endotoxin, oxygen-derived free radicals and cytokines during extracorporeal membrane oxygenation. J Pediatr Surg 1992;27:1199-202. [235] Nilsson L, Kulander L, Nystr6m S-O, Eriksson O. Endotoxin in cardiopulmonary bypass. J Thorac Cardiovasc Surg 1990; 100:777-80. [236] Butler J, Rocker GM, Westaby S. Inflammatory response to cardiopulmonary bypass. Ann Thorac Surg 1993;55:552-9. [237] Khabar KS, elBarbary MA, Khouqeer F, Devol E, al-Gain S, alHalees Z. Circulating endotoxin and cytokines after cardiopulmonary bypass: differential correlation with duration of bypass and systemic inflammatory response/multiple organ dysfunction syndrome. Immunol Immunopathol 1997;85:97-103. [238] Rocke DA, Gaffin SL, Wells MT, Koen Y, Brock Utine JG. Endotoxemia associated with cardiopulmonary bypass. J Thorac Cardiovasc Surg 1987;93:832-7. [239] Laude-Sharp M, Caroff M, Simard L, Pusineri C, Kazatchkine MD, Haeffner-Cavaillon N. Induction of IL-1 during hemodialysis: transmembrane passage of intact endotoxin (LPS). Kidney Int 1990;38:1089-94. [240] Khazazmi A, Andersen LW, Baek L, Valerius NH, Laub M, Rasmussen JP. Endotoxemia and enhanced generation of oxygen radicals by neutrophils from patients undergoing cardiopulmonary bypass. J Thorac Cardiovasc Surg 1989;98:381-5.
5703
[241] Casko G, Suba EA, Elin RJ. Endotoxin-induced platelet activation in human whole blood in vitro. Thromb Haemost 1988;59:375-82. [242] Weingarten R, Sklar LA, Mathison JC, et al. Interactions of lipopolysaccharide with neutrophils in blood via CD14. J Leukocyte Biol 1993;53:518-24. [243] Wright SD, Ramos RA, Hermanowski-Vosatka A, Rockwell P, Detmers PA. Activation of the adhesive capacity of CR3 on neutrophils by endotoxin: dependence on lipopolysaccharide binding protein and CD14. J Exp Med 1991;173:1281-6. [244] Pearson FC. A comparison of the pyrogenicity of environmental endotoxins and lipopolysaccharides. In: Ten Cate JW, Bfiller HR, Sturk A, Levin J, editors. Bacterial endotoxins: structure, biomedical significance, and detection with the Limulus Amebocyte Lysate test. New York: Alan Riss, Inc; 1985. p. 251-63. [245] Enders G, Brooks W, Von Jan N, Lehn N, Bayerd6rfer E, Hatz R. Expression of adhesion molecules on human granulocytes after stimulation with Helicobacter pylori membrane proteins: comparison with membrane proteins from other bacteria. Infect Immun 1995;63:2473-7. [246] Tielemans C, Husson C, Schurmans T, et al. Effects of ultrapure and non-sterile dialysate on the inflammatory response during in vitro hemodialysis. Kidney Int 1996;49:236-43. [247] Ragab AA, Van de Motter R, Lavish SA, et al. Measurement and removal of adherent endotoxin from Titanium particles and implant surfaces. J Orthop Res 1999;17:803-9. [248] Hitchins VM, Merritt K. Decontaminating particles exposed to bacterial endotoxin (LPS). J Biomed Mater Res 1999;46:434-7. [249] Gillis S, Furie BC, Furie B. Interactions of neutrophils and coagulation proteins. Semin Hematol 1997;34:336-42. [250] Altieri DC. Inflammatory cell participation in coagulation. Semin Cell Biol 1995;6:269-74. [251] Esmon CT. Introduction: cellular regulation of blood coagulation. Semin Cell Biol 1995;6:257-8. [252] May AE, Neumann FJ, Preissner KT. The relevance of bloodcell wall adhesive interactions for vascular thrombotic disease. Thromb Haemost 1999;82:962-70. [253] Marcus AJ. Thrombosis and inflammation as multicellular processes: significance of cell-cell interactions. Semin Hematol 2000;31:261-9. [254] Sengelov H. Complement receptors in neutrophils. Crit Rev Immunol 1995;15:107-31. [255] Pabst MJ. Priming of neutrophils. In: Hellewell PG, Williams TJ, editors. Immunopharmacology of the neutrophil. London: Academic Press; 1994. p. 195-222. [256] Sims PJ. Interaction of platelets with complement system. In: Kunicki TJ, Georges JN, editors. Platelet immunobiology: molecular and clinical aspects. Philadelphia: Lippincott; 1989. p. 354-83. [257] Kirschfink M. Controlling the complement system in inflammation. Immunopharmacology 1997;38:51-62. [258] Hitchins VM, Merritt K. Decontaminating particles exposed to bacterial endotoxin (LPS). J Biomed Mater Res 1999;46:434-7. [259] Ragab AA, Van de Motter R, Lavish SA, Goldberg VM, Ninomiya JT, Carlin CR, Greenfield EM. Measurement and removal of adherent endotoxin from titanium particles and implant surfaces. J Orthop Res 1999;17:803-9.
242
This Page Intentionally Left Blank
The Biomaterials Silver Jubilee Compendium
243
The Biomaterials Silver Jubilee Compendium
Author Index Author Adell, R. Aebischer, P. Albrektsson, T. Anderson, J.M. Bergsma, J.E. Black, J. Blanchard, C.R. Boering, G. Bos, R.R. Boyan, B.D. Branemark, P.I. Brannon-Peppas, L. Breuer, C. Bruijn de, W.C. Carreno, M.P. Cochran, D.L. Davies, J.E. Gopferich, A. Gorbet, M.B. Heller, J. Hem, D. Hollinger, J.O. Hubbell, J.A. Hunt, J.A. Hutmacher, D.W. Ingber, D.E. Ishihara, K. Kane, R.S. Kazatchkine, M.D. Kikuchi, A. Kojima, M. Langer, R. Leite, S.M. Lekholm, U. Leong, K. Lincks, J. Liu, Y. Lohmann, C.H. Lundkvist, S. Matsuda, T. Mazzoni, C.L. McNamara, A. McNamara, K. Mikos, A.G. Miller, K.M. Mooney, D.J.
Page No. 17 61 17 21 101 27 147 101 101 147 17 51 129 101 45 147 35 117 219 1 129 139 203 73 175 161 69 161 45 211 69 93 93 17 139 147 147 147 17 35 129 9 129 93 21 129
244
The Biomaterials Silver Jubilee Compendium
Nakabayashi, N. Okano, T. Okuhara, M. Ostuni, E. Park, Y.D. Peppas, N.A. Remes, A. Rockler, B. Rozema, F.R. Sakai, H. Sakurai, Y. Saltzman, W.M. Sarakinos, G. Sefton,M.V. Shimizu, T. Takayama, S. Tan, J. Tirelli, N. Tresco, P.A. Vacanti, J.P. Vince, D.G. Wahlberg, L. Watanabe, A. Williams, D. F. Winn, S.R Whitesides, G.M. Yamada, N. Yamato, M.
69 109, 211 109 161 203 51 79 17 101 109 109 191 93 219 211 161 191 203 61 93, 129 73 61 69 v, 9, 73, 79 61 161 109 211
The Biomaterials Silver Jubilee Compendium
This Page Intentionally Left Blank
245
246
This Page Intentionally Left Blank
The Biomaterials Silver Jubilee Compendium