`
`
`
`Pharmacyclics Exhibit 2068 - Page 1 of 45Pharmacyclics Exhibit 2068 - Page 1 of 45
`
`Sandoz V. Pharmacyclics - IPR2019-00865Sandoz V. Pharmacyclics - IPR2019-00865
`
`

`

`--.'_.9_.?ao;92.;3-_7_7-_1{.6
`--.'_.9_.?ao;92.;3-_7_7-_1{.6
`
`
`Pharmacyclics Exhibit 2068 - Page 2 of 45Pharmacyclics Exhibit 2068 - Page 2 of 45
`
`Sandoz V. Pharmacyclics - IPR2019-00865Sandoz V. Pharmacyclics - IPR2019-00865
`
`

`

`THE UNIVERSITY o::: ~;1~GHIGAN
`MAY 3 1 2000
`
`TAUBMAN MEDK;AL LIBRARY
`
`.
`
`'
`
`

`

`ADVANCES IN ALLOGENEIC
`HEMATOPOIETIC STEM CELL
`TRANSPLANTATION
`
`edited by
`
`Richard K~ Burt
`and
`Mary M. Brush
`
`Northwestern University Medical Center
`Chicago; USA ·
`
`.
`.. KLUWER ACADEMIC PUBLISHERS
`Boston/ Dordrecht/t..rindtm
`
`\
`
`.
`
`.
`
`.
`
`

`

`Distributors for North, Central and South America:
`Kluwer Academic Publishers
`I 4JL6>
`l O 1 Philip Drive
`Assinippi Park
`r<,C
`Nonvell, Massachusetts 02061 USA
`a. i I
`1 B $ i4
`Distributors for all other countries:
`/!t '3 8
`Kluwer Academic Publishers Group
`Distribution Centre
`l 4 4-ltl
`.Post Office Box 322
`· 1
`3300 AH Dordrecht, THE NETHERLANDS
`
`•
`
`,
`
`Library of Congress Cataloging-in-Publication Data
`
`Copyright © 1999 by Kluwer Academic Publishers
`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, mechanical, photo(cid:173)
`copying, recording, or otherwise, without the prior written pennission of the
`publisher, Kluwer Academic Publishers, 101 Philip Drive, Assinippi Park, Nonvell,
`Massachusetts 02061
`Prjnted on acid-free paper.
`Printed in the United States of America
`
`

`

`TABLE OF CONTENTS
`
`Contributors ........................ · ................................................................................... ,.
`Preface .............................. , .. ,. ...................................................................... , ........ ..
`I.
`ALLOGENElC PERIPHERAL BLOOD STEM CELL TRANSPLANTATION
`FOR HEMATOLOGIC DISEASES
`~v1artin K6rbling ............... ,, ................................................................................... 1
`
`2.
`
`3.
`
`4.
`
`5,
`
`6.
`
`7,.
`
`8.
`
`9.
`
`UNRELATED DONOR MARROW TRANSPLANTATION FOR TREATMENT
`OF CHILDHOOD HEMATOLOGIC MALIGNANCIES-EFFECT OF HLA
`DrSPARITY AND CELL DOSE
`Ann E. Woolfrey. Claudio Anasetti, Effie \V. Petersdorf, Paul J. ivlartin, Jean E.
`Sanders. and John A. Hansen.,., ......................................................................... 25
`
`HAPLOIDENTICAL TRANSPLANTATION
`P. Jean l-lenslee-Downey ..................................................................................... 53.
`
`UMBILICAL CORD BLOOD HEMATOPOIETIC STEM CELL
`TRANSPLANTATION
`E. Gluckman, V. Rocha, C.L. Chastang .............................................................. 79
`
`NON MYELOABLATNE "MINI TRANSPLANTS''
`Sergio Giralt, Issa Khouri and Richard Champlin ............................................... 97
`
`ALLOGENEIC HEMATOPOIETIC STEM CELL TRANSPLANTATlON IN
`RECIPIENTS OF CELLULAR OR SOLID ORGAN ALLOGRAFTS .
`Nonna S. Kenyon, Maria Chatzipetrou, Andreas Tzakis, Joshua Miner, Rodolfo
`Alejandro and Camillo RicordL ................................................. ;.,.:i(:L .............. 109
`
`ALLOGENEIC PERIPHERAL BLOOD PROGENITOR CELL
`TRANSPLANTATION IN SOLID TUMORS
`Naoto T. Ueno, Gabriel N. Hortobagyi and Richard E. Champlin .................... .133
`
`HEMATOPOIETIC STEM CELL TRANSPLANTANTION OF MULTIPLE
`SCLEROSIS, RHEUMATOID ARTHRITIS, ANDSYSTEMIC LUPUS
`ERYTHEMATOSUS
`Richard K. Burt, Ann Traynor and William Burns ............................................. 157
`
`ADVANCES IN TI-IE CONTROL OF CYTOMEGALOVIRUS DISEASE IN
`BONE MARROW TRANSPLANT PATIENTS
`William H. Burns ............................................................................................... 185
`
`

`

`12.
`
`I J.
`
`I 0.
`
`14.
`
`ADOPTIVE IMMUNOTHERAPY FOR EBY-ASSOCIATED
`MALIGNANCIES
`Kenneth G. Lucas and J. Christian Barrett ......................................................... 203
`11. ADOPTIVE IMMUNOTHERAPY USING DONOR LEUKOCYTE
`INFUSIONS TO TREAT RELAPSED HEMATOLOGIC MALIGNANCIES
`AFTER ALLOGENEIC BONE 1v1ARROW TRANSPLANTATION
`William R. Drobyski ............................................................................................ 233
`CLINlCAL USE OF IRRADIATED DONOR LYMPHOCYTES IN BONE
`MARROW TRANSPLANTATION
`Alan M. Ship and Edumd K. Waller .................................................................. 267
`PENDRITIC CELLS AND THEIR CLINICAL AP PUCA TIONS
`D.N. J. l·lart and G.J. Clark ................................................................................ 283
`ENGINEERING HEMATOPOIET1C GRAFTS USING ELUTRIATION AND
`POSITIVE CELL SELECTION TO REDUCE GVHD
`Stephen]. Noga ................................................................................................. 311
`I 5. MONOCLONAL ANTIBODY AND RECEPTOR ANTAGONIST
`THERAPYFOR GVHD
`James L.M. Ferrara, Ernst Holler and Bmce Blazar ........................................ 331
`16. ADOPTIVE IMMUNOTHERAPY FOR LEUKEMIA: DONOR
`LYMPHOCYTES TRANSDUCED WITH THE HERPES SIMPLEX
`THYivHDINE KINASE GENE
`Charles J. Link, Jr., Ann Traynor, Tatiana Seregina and Richard K. Burt ..... 369
`17. CLINICAL AP.PLICATION OF HEMA TOPOIETIC STEM CELL CUL TIJRE
`AND EXPANSION
`Stephen G. Emerson and Patricia Conrad ......................................................... 377
`18. NEW CYTOKINES AND THEIR CLINICAL APPLICATION
`Ian K. McNiece ................................................................................................. 389
`INDEX .......................................................................................................................... 407
`
`vi
`
`

`

`This material may be protected by Copyright law (Title 17 U.S. Code)
`
`15
`MONOCLONAL ANTIBODY AND
`RECEPTOR ANTAGONIST THERAPY
`FORGVHD
`
`James L.M. Ferrara, M.D.
`University of Michigan Cancer Center, Ann Arbor, MI 48109
`Ernst Holler, M.D.
`Regensberg University Hospital, Regensburg, Germany, D93042
`Bruce Blazar, M.D.
`University of Minnesota, Minneapolis, MN 55455
`
`INFLAMMATORY CYTOKINES IN GVHD: PATHOPHYSIOLOGY
`
`Recent experimental evidence suggests that dysregulation of complex cytokine
`networks occurring in three sequential steps is responsible for many of the
`systemic manifestations of acute graft versus host disease (GVHD) (Figure 1).
`The first step of GVHD pathophysiology begins with the transplant conditioning
`regimen, which in clinical BMT includes total body irradiation (TBI) and/or
`chemotherapy. The conditioning is an important variable in the pathogenesis of
`acute GVHD because it damages and activates host tissues, including intestinal
`mucosa, liver and skin. Activated host cells then secrete inflammatory
`cytokines, e.g. tumor necrosis factor (TNF-a) and IL-I. The presence of
`inflammatory cytokines during this phase may upregulate adhesion molecules
`and MHC antigens by mature donor T cells in the second step of acute GVHD.
`This inflammatory context helps to explain the observation that enhanced risk of
`GVHD after clinical BMT is associated with certain intensive conditioning
`regimens that cause extensive injury to epithelial and endothe1ial surfaces and
`the subsequent release of inflammatory cytokines.
`
`During the second step of acute GVHD, donor T cells proliferate and secrete IL-
`2 and IFN-y (Th 1 cytokines} in response to host alloantigens. During
`autologous BMT, donor and recipient are the same, and thus no allospecific T
`cell activation occurs. Th 1 cytokines play a central role in the expansion of
`donor T cells and the activation of other effector cells such as large granular
`lymphocytes (LGLs) and natural killer (NK) cel1s. LGLs, NK cells and
`
`

`

`Monoclonal Antibody and Receptor Antagonist Therapy for G VHD
`
`monocyte/macrophages appear to be prominent in the effector arm of GVHD
`and may contribute to the pathological damage, i.e. induce the changes of GVH
`disease following the T cell mediated GVH reaction. The initial hypothesis that
`the cytolytic function of lymphocytes directly causes the majority of tissue
`damage and necrosis has been modified and may be more appropriately
`restricted to specific target organs, e.g. the liver.
`
`Figure 1. GVHD Pathophysiology
`
`( l J Recipicnl condlt1onit1g
`
`( 2)
`Dnnpr
`Tedi
`a~tiv,uron
`
`During the step three of acute GVHD physiology, increased secretion of the
`inflammatory cytokines TNF-u and IL-1 mononuclear phagocytes - which had
`been primed with Thl cytokines during step two, occurs after triggering by a
`secondary stimulus which may be provided by lipopolysaccharide (LPS or
`endotoxin). LPS can leak through the intestinal mucosa damaged by the
`conditioning regimen and subsequently stimulate gut-associated lymphocytes
`and macrophages. LPS reaching skin tissues may also stimulate kertinocytes,
`dermal fibroblasts and macrophages to produce similar cytokines in the dermis
`and programmed cell death pathway. TNF-a. mediated apoptosis may be
`particularly important in acute GVHD of the intestine. In addition to these
`preinflammatory cytokines, excess nitric oxide produced by activated
`macrophages may contribute to the deleterious effects on GVHD target tissues,
`particularly immunosuppression. Induction of inflammatory cytokines may thus
`synergize with the lytic component of GVHD provided by CTL and NK cells,
`resulting in the amplification of local tissue injury and further promotion of an
`inflammatory response which ultimately leads to the observed target tissue
`destruction in the BMT host. Blockage of these cytokines, either directly or
`indirectly, thus becomes an attractive therapeutic strategy.
`
`332
`
`

`

`Advances in Allogeneic Hematopoietic Stem Cell Transplantation
`
`The proteins, tumor necrosis factor alpha (TNF-a) and interleukin-I (IL-1), are
`secreted during inflammatory processes of natural or "innate" immunity to
`infectious microbes or other foreign proteins. The cellular sources of TNF-cx ·
`and IL-1 are often, but not always, mononuclear cells, and an abundance of cell
`types are known to produce thent. These proteins are also produced during
`processes of acquired immunity, thus enhancing natural immunity; they often
`have synergistic, pleiotropic, and occasional redundant effects on target cells,
`activating lymphocytes as well as performing certain effector functions. TNF-cx.
`was originally described as a protein produced after activation of macrophages
`following Jipopolysaccharide (LPS) stimulation and which induced hemorrhagic
`necrosis of transplantable murine sarcomas (1). TNF-cx is now known to be a
`critical proinflarnrnatory cytokine capable of mediating the deleterious effects of
`It is· central to
`severe infectious complications such as septic shock (2,3 ).
`proinflarnrnatory responses that result in a wide variety of systemic effects,
`including fever, acute phase respons~s, hematopoietic growth factor production,
`activation of endothelial cells and the destruction of epithelial cells and
`These direct effects are accompanied by increased
`kertinocytes ( 4-6).
`expression of a variety of molecules critical to immune responses such as
`intercellu1ar adhesion molecules (ICAMs) and HLA class I antigens on
`epithelial and endothelial cells (5,6). Furthermore, a costimulatory role for
`TNF-a in .activation of cytotoxic T cells has been reported (7).
`
`IL-1 shares multiple proinflammatory properties with TNF-cx (2), making both
`monokines candidates for involvement in the pathophysiology of BMT related
`complications. IL-1 is a polypeptide with two isoforrns, IL-1 ex and IL-1 f3. IL-1
`was first described as "endogeneous pyrogen" and was later shown to be a
`lymphocyte stimulating factor. Endotoxin is the most important known stimulus
`for IL-I production, and approximately 5 molecules per lymphocyte is sufficient
`to stimulate IL-1 rnRNA transcription. Recent studies suggest that cells
`undergoing apoptosis or programmed cell death produce IL-1 more efficiently
`(8) and IL-1 appears to be important in the regulation of apoptosis. Over(cid:173)
`expression of IL-1 converting enzyme (ICE) induces apoptosis in cell lines (9)
`and ICE-deficient mice are resistant to apoptosis through the Fas pathway
`(10,199). IL-1 and TNF-a appear to be linked in this pathway because the Fas
`antigen (CD95) is a cell-su·rface protein with high homology to the TNF-a
`receptor (12,13).
`
`Endotoxin is a principal inducer of inflammatory cytokines, and stimulation of
`TNF-a and IL-1 by endotoxin is the most likely explanation for the well-lmown
`role of bacterial microflora in development of GVHD. Sensitivity to endotoxin
`and its ability to trigger release of TNF-a are greatly enhanced in mice with
`GVHD, and this priming is predominantly due to the costimulatory role of IFNy
`333
`
`

`

`Monoclonal Antibody and Receptor Antagonist Therapy for GVHD
`
`in the setting of BMT is the result of activation of donor Th I cells, which occurs
`subsequent to the recognition of alloantigens. A second mechanism of IFNy
`induction involves viral infections such as cytomegalovirus (15,16), a pathway
`that is of great potential significance in the context of BMT. Reproducible
`increases of IFNy serum levels have also been noted during pre-transplant
`conditioning by Niederwieser et al (16). There is also increasing evidence for
`induction of inflammatory cytokines by cytotoxic drugs and ionizing irradiation.
`Induction of TNF-a. mRNA had been demonstrated in cultures of human
`sarcoma cells
`(17) and
`in myeloid cell
`lines as well as normal
`monocytes/macrophages (18, 19). In these systems, doses as low as 2 Gy induce
`TNF-a. mRNA after 1 hour and reached a maximum after 3 hours. Clinically,
`increases of serum TNF-cx. and soluble TNF-receptor p55 (sTNF-R p55) have
`been observed following total body irradiation (TBI) with fractions of 4 Gy in
`patients prepared for allogeneic BMT (20). Both TNF-a. and Interleukin 6 (IL-
`6) are released after a single TBI dose of 10 Gy (21). In an experimental SCID
`model, pretransplant conditioning induced systemic TNF-a within four hours of
`TBI. IL-1 a levels increased slightly later, peaking at 72 hours (22). Systemic
`cytokine release was accompanied by increased expression of both TNF-a, and,
`subsequently, IL-6 in colonic tissue. These results have been independently
`confirmed in other experimental models, where IL-1 ex mRNA was induced 100-
`fold in the spleens of mice one week after receiving either syngeneic or
`allogeneic BMT; IL-1 a production then returned to normal in syngeneic (but not
`allogeneic) BMT recipients (23). Although experimental data on cytokine
`induction by cytotoxic drugs that are used for pretransplant conditioning are
`rare, significant induction seems likely. While only minimal amounts of
`systemic and tissue expression of TNF-a and IL-1 a were detected in mice
`conditioned by bulsulphan/cyclophosphamide regimens, large increases in TNF(cid:173)
`a. serum levels have been seen in patients receiving CYB (Cyclophosphamide,
`Etoposide, BCNU) conditioning prior to autologous BMT, suggestive that at
`least this combination chemotherapy can cause TNF-a. release (see below).
`Cyclophosphamide alone is known to trigger TNF-a. release from normal
`mononuclear cells in vitro (24).
`
`Parallel to the heterogeneity of mechanisms resulting in the release of
`inflammatory cytokines, a vast number of cellular sources may be involved in
`this process. Donor monocytes and, to a lesser extent, donor CD4+ T cells can
`produce TNF-a. Host macrophages are also candidate sources of TNF-a. (25).
`Organs such as the lung and the liver (with significant populations of alveolar
`macrophages and Kupffer cells, respectively) are likely to be important
`The concept of macrophage-derived
`producers of systemic cytokines.
`production of cytokines might also help to explain the causal relationship
`between pretransplant risk factors and post-transplant complications, because
`
`334
`
`

`

`Advances in Allogeneic Hematopoietic Stem Cell Transplantation
`
`macrophages have a half-life of about 2 to 3 months and are only slowly
`In patients with chronic rnyeloid
`replaced by donor monocytes (26,27).
`leukemia (CML), tissue macrophages may be a source of cytokine dysregulation
`throughout the period of acute GVHD. Other cells capable of TNF-cx.
`production are NK cells and kcratinocytes (28). Mast cells have also recently
`been recognized as potent reservoirs of performed TNF-a, which can be
`released rapidly during the process of degranulation; these cells are thus ideal
`candidates for initiation of an inflammatory cascade without prior de novo
`synthesis of cytokines (29,30).
`
`Mixed lymphocyte cultures are thought to represent in vitro models of host-vs. -
`graft and graft-vs.-host reactions, but MLC reactivity does not generally
`correlate with the incidence and severity of GVHD (J. Hansen, personal
`communication). Nevertheless, in many centers MLCs are used in the selection
`of suitable donors in allogeneic BMT. In a study of HLA-disparate MLCs,
`Dickinson used MLC supematants to induce GVHD histology in skin explant
`In these experiments, histological severity of GVHD was
`assays (31 ).
`significantly associated with the amount of TNF-a and IFNy (31). Thus, in a
`simplified model of GVHD in vitro, TNF-a is not merely an index of cellular
`activation but is also a direct mediator of target organ pathology.
`
`Convincing data regarding a direct pathophysiological role of TNF-a were first
`presented by Piguet in 1987 (33). In a P ➔ Fl model of MHC-incompatible
`BMT, lethally irradiated mice were injected with bone marrow and lymph node
`T cells. The observed skin and gut lesions of acute GVHD as well as weight
`loss (a systemic indicator of GVHD) were significantly prevented by weekly
`administration {day +7 to day +35) of a polyclonal antibody that neutralized
`TNF-a (33,34). Since that time, several investigators have confirmed _Piguefs
`observations (30,35-37). In different models of acute GVIID, administration of
`TNF-a antibodies starting either after BMT or from day +7 have reduced the
`mortality from GVHD and its associated signs, including immunosuppression
`and mast cell degranulation. In most of these studies, antibodies neutralizing
`TNF-a were less effective than the use of a T cell-depleted bone marrow in
`preventing GVHD (30, 33, 35-37). In one study> the protective effects of anti
`TNF-antibodies seemed to be restricted to acute GVHD (30). It should be noted
`that in some studies, expression of TNF-a, was less pronounced than the
`expression of IL-1 or IFNy, demonstrating that TNF-a is but one important
`cytokine among several that are dysregulated (38, 39).
`
`Convincing data regarding a direct role for IL-1 in GVHD were published by
`Abyankhar et al (23). Using a well-described mouse model of GVHD to minor
`histocompatibility antigens in a BMT model, mRNA for IL-1, IL-2, and TNF-a
`were evaluated after transplant by a semi-quantitative RT-PCR technique. In the
`335
`
`

`

`Monoclonal Antibody and Receptor Antagonist Therapy for GVHD
`
`spleen, lL-.1 ex levels were increased almost two orders of magnitude the first
`week after transplant in both the spleen and the skin. One month after
`transplant, when GVHD was clinically apparent, IL-let mRNA levels were 600-
`fold higher. Administration of IL-1 receptor antagonist from day + 1 O to day
`+20 after transplant significantly increased survival from approximately 20% to
`80%. Thus, IL-I ex appears to be a critical effector molecule in this experimental
`model of acute GVHD.
`
`The observation of systemic inflammatory cytokine release during pretransplant
`conditioning and its correlation with poor outcome following BMT in clinical
`studies (see below) suggest that release of inflammatory cytokines before the
`transplant might increase activation of donor T cells, perhaps by upregulating
`HLA and adhesion molecules on epithelial and endothelial targets (5, 6). To
`explore the pathophysiological relevance of this pathway in an experimental
`model, a P ➔ Fl mouse system was investigated giving only two injections of a
`polyclonal neutralizing. TNF-a antibody prior to irradiation BMT. Using a low
`dose of donor T cells for induction of GVHD, this approach was as effective in
`prevention of mortality and weight loss. The role of conditioning-related release
`ofTNF-a in the induction of GVHD was elegantly confirmed in a study by Xun
`and his colleagues (22). In this study, lethal, acute GVHD was prevented by
`prolonging the interval between TBI and BMT for at least four days, i.e., to a
`time when conditioning-related TNF~a levels had declined.
`Injection of the
`soluble rhuTNFR:Fc, an engineered TNF-ex-antagonist made by linking two
`soluble TNF-receptors (p80) with the Fe portion of human IgG C resulted in a
`comparable protection from GVHD without any interval between TBI and
`BMT. It should be noted, however, that neither of these approaches completely
`eliminated acute GVHD, nor did the use of a less aggressive conditioning
`regimen with busulfan and cytoxan.
`
`The importance of timing between the conditioning regimen and the injection of
`donor cells has been confirmed by several independent laboratories, including
`an analysis of a murine model of GVHD using cyclophosphamide alone ( 40).
`The role of pretransplant conditioning, especially ionizing· irradiation, in the
`induction of GVHD is also supported by the clinical observation of intensified
`GVHD lesions in irradiated areas ( 41 ). This mechanism has been confirmed by
`experimental studies where GVHD histology occurred primarily in murine skin
`that had been irradiated and transplanted prior to the induction of GVHD in the·
`recipients (42).
`
`Chronic GVHD has been observed in experimental models where acute GVHD
`is prevented with anti-TNF-a, suggesting· that early release of TNF-a and
`conditioning-related tissue damage accelerate and amplify acute GVHD but are
`not relevant to the induction of chronic GVHD. This dichotomy implies that
`336
`
`

`

`Advances in Allogeneic Hematopoietic Stem Cell Transplantation
`
`prophylactic neutralization of inflammatory cytokines will not be able to induce
`tolerance to host antigens, a fact only one study has documented TNF-a
`expression in a murine model of chronic GVHD (38). In this report, TNF-a
`mRNA expression was not increased, IFNy was suppressed, and IL-4 was over(cid:173)
`expressed in splenocytes, suggesting a shift towards a Th2 profile in chronic
`GVHD. These results would suggest that the proinflammatory cascade (IFNy ->
`1NF-a) does not operate during the pathophysiology of chronic GVHD, but
`further studies are needed.
`
`A review of the published data on systemic TNF-o: levels during the first three
`months following BMT clearly demonstrates that TNF-o: serum levels are
`increased in patients not only developing acute GVHD, but those experiencing
`acute endothelial complications such as VOD or capillary leakage syndrome
`( 43-45). This association of elevated TNF-a levels with a variety of
`pathological conditions., as well as its wide range of normal values, has so far
`precluded its use as an independent diagnostic parameter for acute GVHD. We
`have shown that a close correlation exists between acute GVHD and elevated
`sTNFR p55 levels (46) (Table 1 ) . The shedding of soluble receptors subsequent
`to induction of TNF-a has been reported for a variety of clinical conditions; in
`the future, sTNFR levels might prove to be a more suitable and sensitive
`approach for clinical monitoring, because soluble receptors are more stable than
`native cytokines and circulate in normal individuals with significant (nanogram)
`levels, which may facilitate the detection of minor shifts in their concentrations.
`
`A second and more precise approach to predictive assays may be the analysis of
`(PBMC).
`cytokine production by peripheral blood mononuclear cells
`Quantitative data can be obtained by evaluation of cytokine expression using
`either immunodetection of intracytoplasmatic cytokine protein ( 4 7) or PCR
`analysis of cytokine mRNA (23, 45). Such techniques have shown increased
`expression of both TNF-a and IL-1 in patients with GVHD. It should be noted
`that PBMC analysis preswnes that peripheral circulating cells are relevant to the
`pathophysiology of acute GVHD; other critical cell populations, such as tissue
`macrophages or endothelial cells, may be major producers of relevant cytokines
`but may be inaccessible to analysis. Detailed correlations of serum or PBMC
`cytokines levels with tissue cytokine expression are therefore needed.
`
`One of the most important findings in studies regarding the systemic release of
`TNF-a. is the strong correlation between the time of first elevated levels ofTNF(cid:173)
`a. and the occurrence of transplant-related complications (TRC) ( 48). In an
`update of studies in Munich, 222 patients receiving either allogeneic related
`donor (n= 161), unrelated donor (n= 34) or autologous BMT (n= 35) have been
`analyzed for daily serum TNF-a levels starting from admission until day +10
`337
`
`

`

`}.tfonoclonal Antibody and Receptor Antagonist Therapy for GVHD
`
`after BMT. Patients receiving HLA-identical sibling donor BMT are divided
`into 4 subgroups: I) patients with increased · TNF-a. serum levels prior to
`treatment and without any clinical symptoms of "chronic'' cytokine release (as
`previously described) (49); 2) patients with low levels prior to chemotherapy
`and an acute increase during pretransplant conditioning; 3) patients with acute
`release observed between the day + 1 and day + 10 following BMT; and 4)
`patients without any pathological TNF-a serum levels prior to day + 10 after
`BMT. Patients with de novo release of TNF-a (group 2) have an extremely
`poor prognosis due to occurrence of severe treatment related complications
`(TRC) including acute GVHD. Their outcome is significantly worse when
`release of TNF-a. occurs during pretransplant conditioning. indicating that host:(cid:173)
`derived cytokine release is important to the induction of GVHD.
`
`Table 1. Maximal release of TNF-a. soluble TNF-receptor p55 (sTNFR p55) in
`sequential serum samples obtained between day 0 and day + 100 following BMT
`(Mean± SEM, pg/ml). TNFa was determined by ELISA. Patients were
`grouped according to the type of BMT and occurrence or absence of
`complications. Maximal levels between groups were compared by Wilcoxon
`tests; both TNFa and sTNFR were significantly elevated in patients with GVHD
`grade II (p<0.01). TNFa levels in frozen control samples (n=20) were 28 ± 36
`pg/ml, sTNFR p55 levels 3000 ± 300pg/ml. TNFa levels in fresh control
`samples (n=30) were 48 ± 20pg/ml.
`
`BMT
`
`Type
`
`Autologous Allogeneic
`
`complication --
`
`a GVHD
`
`aGVHD
`
`aGVHD
`
`VOD/
`
`grade 0-1
`
`grade II
`
`grade III/IV Capillary
`
`TNFn
`
`(cryo)
`
`TNFet
`
`(fresh)
`
`sTNFR
`
`pSS
`
`30
`
`(46)
`
`n=4
`
`111
`
`(28)
`
`n=14
`
`5,500
`
`50
`
`(11)
`
`""'21
`
`181
`
`(48)
`
`n"'13
`
`4,100
`
`(1,800)
`
`(1,200}
`
`n=5
`
`n=9
`
`97
`
`(36)
`
`n=8
`
`365
`
`(80)
`
`n=l1
`
`8,500
`
`(2,400)
`
`n=l2
`
`338
`
`326
`
`(152)
`
`n"'17
`
`477
`
`(103)
`
`n=l4
`
`14,300
`
`(5,700)
`
`n=l l
`
`Leak
`
`726
`
`(360)
`
`n=6
`
`554
`
`(161)
`
`n=l3
`
`26,400
`
`(10,800)
`
`n=6
`
`

`

`Advances in Allogeneic Hematopoietic Stem Cell Transplantation
`
`A still puzzling observation is the extremely low incidence of TRC in patients
`Increased serum levels in these
`with "chronic" release of TNF-a (group 1).
`patients correlate with increased spontaneous production of TNF-cx by PBMC in
`culture. Unique features of this subgroup are its low incidence of GVHD and
`to stage, age or previous infectious
`the absence of disease correlatio11,s
`complications, implying existence of some intrinsic or even genetically fixed
`mechanisms responsible for the increased production of TNF-cx. Analyses of
`polymorphisms within the TNF-cx gene are presently being performed to
`investigate this hypothesis. In thalassemia patients, a subgroup with high TNF(cid:173)
`a levels prior to BMT has also been observed; although there was some
`correlation of TNF-cx level with liver fibrosis, there is a stronger association
`with HLA-DQwl, suggesting ~ mechanism of genetically fixed cytokine
`dysregulation (51). In spite of the small numbers of patients with spontaneous
`TNF-a. release, further evaluation of this group may lead to important insights
`regarding mechanisms of cytokine mediated damage. For example, chronic
`TNF-a. secretion may lead to desensitization to TNF-a., or other cytokines may
`be simultaneously induced in this network that have protective effects.
`Recently, Tan and colleagues have identified a novel factor produced by Thl T
`cell clones that protects NOD mice from the development of diabetes (52). In
`addition, polymorphisms of the TNF-a. gene and promoter region have been
`found to correlate with increased 1NF-a production in heart transplant
`recipients (53). Further insights into regulation of cytokine networks and their
`impact on disease could lead to therapeutic interventions and provide an
`alternative approach to the prevention ofTRC in the future.
`
`Correlation of conditioning-related release of TNF-a and BMT outcome
`has now been observed in a Swedish study (54).
`Experimental findings from several laboratories confirm murine studies reported
`by Xun (22) and .Lehnert (40). Although the murine models have shown an
`almost exclusive role for TBI in the induction of excess 1NF-a, comparison of
`conditioning regimens in patients in Munich have not yet shown a significantly
`higher incidence of TNF-a release in those receiving TBI-CY regimens (23%
`TNF-a release during conditioning, 24% between day +l and day +10)
`compared to BUCY regimens (16% TNF-a release during conditioning. 15%
`between day + 1 and + 10). These data, already suggest occurrence of some
`cytokine release during non-TEI-containing regimens. but a more detailed
`analysis allowing identification of responsible cytotoxic agents is not yet
`available. Recently, this issue was clarified by analyzing circadin kinetcs of
`receiving monoclonal anti-TNF-a
`in patients
`TNF-anti-TNF-cornplexes
`throughout conditioning (55). In these patients, binding ofTNF-a. to circulating
`339
`
`

`

`Monoclonal Antibody and Receptor Antagonist Therapy for GVHD
`
`antibody facilitated detection of TNF-a production during both TBI/CY and
`BU/CY conditioning regimens. In TBI/CY-treated patients, every single dose of
`TBI (as well as every dose of CY) was followed by an increase of TNF-anti(cid:173)
`TNF-complexes; in BU/CY-treated patients, only CY induced significant peaks.
`In contrast, oral application of BU had no effect on levels of these complexes.
`These data clearly indicated TNF-a is induced by both TBl and CY in vivo,
`confirming that cytokine release is not restricted to TBI in humans. As already
`discussed, the concept of chemotherapy induced cytokine release is further
`supported by analysis of patients receiving autologous BMT following CVB(cid:173)
`conditioning, where 8/35 patients showed significant release of TNF-a during
`the course of conditioning. Even in autologous BMT, transplant related
`mortality was increased (37.5% versus7.4% in patients with or without TNF-a
`re1ease, respective]y) which was mainly due to pulmonary complications.
`
`Some clinical risk factors predisposing to acute release of TNF-a. during the
`Increases of TNF-a levels
`course of conditioning have been identified (36).
`were significantly (p<0.001) correlated with a diagnosis of CML or
`myelodysplastic syndrome. Fever or skin exanthems during conditioning were
`strongly associated (p<0.001) with TNF-a. release; in addition, failure of
`gastrointestinal decontamination, as indicated by the presence of pathological
`stool specimens on the day of BMT was also associated with acute release of
`TNF-a (p<0.05). Interestingly, these risk factors mirror the concept of three of
`the major mechanisms involved in host-associated cytokine dysregulation:
`pretransplant conditioning itself, underlying disease, and translocation of
`bacteria from a damaged intestinal mucosa.
`
`CLINICAL USE OF ANTAGONISTS OF
`CYTOKINES
`
`INFLAMMATORY
`
`Corticosteroid
`When discussing the use of specific cytokine antagonists for the treatment and
`prevention of acute GVHD, it should be remembered that corticosteroids, which
`are still the treatment of choice in first line treatment of acute GVHD exhibit
`potent and broad cytokine antagonism. There is increasing evidence that
`corticosteroids act via suppression of cytokine gene activation rather than by
`