MICROSCOPY RESEARCH AND TECHNIQUE 50:229–235 (2000)
`
`Anti-TNF Therapies in Rheumatoid Arthritis, Crohn’s
`Disease, Sepsis, and Myelodysplastic Syndromes
`AZRA RAZA*
`Rush Cancer Institute, Rush-Presbyterian-St. Luke’s Medical Center, Chicago, Illinois 60612-3515
`
`KEY WORDS
`
`anti-TNF therapies; rheumatoid arthritis; Crohn’s disease; myelodysplastic syn-
`dromes
`
`An attempt has been made in this article to summarize the state-of-the-art clinical
`ABSTRACT
`experience with the use of anti-TNF therapies in four diseased states with special emphasis on
`myelodysplastic syndromes. Given the central role of TNF-a in initiating and perpetuating the
`chronic damage produced in the diseased organs by controlling a cascade of pro-inflammatory
`cytokines, as well as its acute role in sepsis, theoretically speaking, neutralization of this peptide
`was a natural therapeutic choice. Results of the initial clinical trials appear encouraging and
`sometimes dramatic in their efficacy. The mechanism of response however, is interesting in that
`even when TNF-a is directly targeted by a monoclonal antibody, the resulting benefits can fre-
`quently not be attributed to TNF suppression alone. Rather, it appears that a more general effect
`on the T-lymphocytes is also contributing to the responses being seen. This raises the new
`possibility of combining anti-cytokine and anti-T-cell strategies to treat at least the more chronic
`diseases such as Crohn’s disease and myelodysplastic syndromes. Continued clinical trials testing
`these strategies are clearly warranted. Microsc. Res. Tech. 50:229–235, 2000.
`© 2000 Wiley-Liss, Inc.
`
`INTRODUCTION
`Parsimony in nature is graphically manifested in the
`frequent employment of a single pathway for multiple,
`often opposing functions. One prime example of this
`cross-functionality is provided by the ubiquitous,
`multi-purpose cytokine, tumor necrosis factor alpha
`(TNF-a).While diseases may not share a common eti-
`ology, a plethora of signs and symptoms experienced in
`states as disparate from each other as rheumatoid ar-
`thritis (RA), multiple sclerosis (MS), acquired immune
`deficiency syndrome (AIDS), cancer, septic shock, con-
`gestive heart failure (CHF), and hematopoietic disor-
`ders such as aplastic anemia (AA) and myelodysplastic
`syndrome (MDS) have been directly attributed to the
`excess production of a single cytokine, TNF-a (Can-
`nella and Raine, 1991; Cannon et al., 1990; Dezube et
`al., 1995; Feldmann et al., 1996; Hober et al., 1989;
`Koike et al., 1995; Perrault and Menasche, 1999; Raza
`et al., 1996a–c; Selmaj et al, 1991; Shetty et al., 1996).
`The fact that one protein is capable of mediating the
`danger signs of so many maladies depending upon the
`cytokine cascade it initiates, the organ involved, the
`cells in question, the etiologic agent that is operational,
`and whether the irritation is acute or chronic has only
`recently been appreciated in some totality, and clearly
`represents the tip of the iceberg. Overproduction of
`TNF-a may be acute and intense, or chronic and low-
`grade, each carrying its own set of unique manifesta-
`tions. The grave consequences of extremely high TNF-a
`levels such as those seen acutely in septic shock are
`related to two major effects of this pro-inflammatory
`cytokine. Along with Interleukin-1 (IL-1), it serves as a
`“proximal” master initiator of the cytokine cascade,
`which helps perpetuate septic shock, and, secondly, the
`more “distal” end-organ damage that so often proves to
`be irreversible and fatal (Porter, 1997). However, the
`© 2000 WILEY-LISS, INC.
`
`more subtle consequences of chronic, low-level eleva-
`tions are far more obscure and difficult to decipher,
`regularly defying scrutiny and only recently becoming
`the subject of systematic study. These include the role
`of TNF-a in causing the signs and symptoms of chronic
`diseases such as RA, AIDS, MS, AA, and MDS. At-
`tempts to suppress both the acute and chronic eleva-
`tion of TNF-a are, therefore, only as old as the recog-
`nition that this peptide is the principle mediator of the
`disease manifestations related to a variety of patho-
`logic states. In other words, the field is in its infancy.
`The success of anti-TNF or anti-cytokine approaches in
`vitro and in animals have more recently been trans-
`lated into human trials with some equivocal and occa-
`sional dramatic outcomes. Until the precise mecha-
`nisms of response are understood, both the successes
`and the failures must be taken with a grain of salt and
`extreme caution must be exercised in ascertaining that
`the baby is not thrown out with the bathwater. The
`purpose of this article is to briefly review anti-TNF
`therapies in general, with special reference to the pre-
`leukemic disorders grouped under the heading of my-
`elodysplastic syndromes or MDS.
`
`ANTI-TNF THERAPIES AVAILABLE
`There are basically three approaches to suppress the
`activity of TNF-a that have been utilized in clinical
`trials so far. The first is to use antibodies to TNF and
`these include the chimeric monoclonal antibody cA2 or
`Infliximab/ Remicade (Centocor) and the humanized
`antibody of the IgG4 isotype CDP571 (Celltech/ Bayer).
`
`*Correspondence to: Azra Raza, M.D., rofessor, Director Pre-Leukemia and
`Leukemia Program, Rush Cancer Institute, 2242 West Harrison Street, Suite
`108, Chicago, IL 60612-3515. E-mail: araza@rush.edu
`Received 5 January 2000; accepted in revised form 25 February 2000
`
`Dr. Reddy’s Laboratories, Inc. v. Celgene Corp.
`IPR2018-01504
`Exhibit 2014, Page 1
`
`

`

`230
`
`A. RAZA
`
`TABLE 1. Anti-TNF therapeutic agents
`
`Drug
`1. Etanercept
`2. Lenercept
`3. Infliximab
`4. CDP571
`humanized moab
`against TNF
`5. Thalidomide
`6. Pentoxifylline
`7. IL-10
`8. Amiodorone
`
`Company
`Immunex
`Hoffman-LaRoche
`Centacor
`
`Celltech
`Celgene
`Hoescht-Roussel
`Schering
`Wyeth-Ayerst
`
`Status
`FDA approved
`Developmental
`FDA approved
`
`Developmental
`FDA approved
`FDA approved
`Developmental
`FDA approved
`
`An extension of this approach is to block the transcrip-
`tion factor NFkB believed to function upstream of
`TNF-a, critically controlling the onset and perpetua-
`tion of the inflammatory process. A second approach to
`neutralize the activity of TNF-a is based upon provid-
`ing the soluble TNF-a receptors in vivo. Enbrel or p75
`TNF-R:Fc (Immunex) and lenercept, p55 TNF-R:Fc
`(Roche) are examples of this therapy. Finally, agents
`known to interfere with the signaling pathways of a
`variety of cytokines simultaneously though less specif-
`ically, can nonetheless be quite useful, especially in
`combinations. These include drugs such as pentoxifyl-
`line, thalidomide, and dexamethasone. See Table 1 for
`details. The following review will concentrate upon the
`in vivo use of these agents in a variety of human
`diseases.
`
`RHEUMATOID ARTHRITIS
`It has been appreciated for at least two decades that
`cytokines lie at the heart of chronic inflammatory and
`autoimmune diseases. In the case of RA, all cytokines
`were found to be expressed in the joints, suggesting
`that a key aspect of chronic diseases may be the con-
`tinuous production of cytokines, as opposed to the tran-
`sient one seen in response to antigen presentation. It
`was rapidly learnt that most of the destruction in the
`joints of RA patients was mediated via IL-1. Since
`TNF-a regulates IL-1 activity, it was no surprise to
`discover that both TNF and TNF receptors were up-
`regulated in the joints of patients with RA (Buchan et
`al., 1988; Chu et al., 1991; Di Giovine et al., 1988). This
`led to attempts to neutralize TNF-a and thereby ame-
`liorate the ongoing IL-1 mediated joint destruction. At
`least in vitro it was found that a successful suppression
`of TNF-a was simultaneously accompanied by an inhi-
`bition of IL-1, granulocyte macrophage colony stimu-
`lating factor (GM-CSF), IL-6 and IL-8, suggesting that
`the proinflammatory cytokines are not independently
`regulated, but controlled as a module or coordinate
`(Brennan et al., 1989; Haworth et al., 1991). Once
`again, it was no surprise that the critical coordinator of
`this cytokine cascade was none other than TNF-a, and
`IL-1 to a lesser extent (Butler et al., 1995). Prior to the
`discovery that TNF-a often provided the password that
`sets the cytokine cascade into motion, these inflamma-
`tory mediators were considered poor therapeutic tar-
`gets because of their redundant and overlapping roles.
`The modular functionality of the cascade capable of
`being switched on and off by a master peptide changed
`all that. Following the in vitro work on cultured RA
`cells, animal studies were undertaken centered on the
`concept of downregulating the inflammatory cytokine
`
`module by neutralizing TNF-a activity in an animal
`model of RA, collagen-induced arthritis. Both joint in-
`flammation and joint destruction were found to be re-
`duced, even if administered after the onset of disease
`(Piguet et al., 1992; Thorbecke et al., 1992; Williams et
`al., 1996) human trials soon followed. A chimeric anti-
`TNF-a antibody called Infliximab (Centocor) produced
`encouraging results (Elliott et al., 1993) and led to
`many confirmatory studies with other antibodies.
`These include CDP571 (Celltech/Bayer), enbrel, p75
`TNF-R Fc (Immunex/AHP), lenercept, p55 TNF-R Fc
`(Roche), and D2E7 (Knoll/CAT) (Feldmann et al., 1998;
`Weinblatt et al., 1999). Overall, these therapies were
`well tolerated in general. Since most trials focused on
`RA patients with severe disease, doing badly on exist-
`ing therapies with reduced life expectancies and other
`morbidity, it is unfair to compare the efficacy of this
`approach with standard therapies especially when
`used in patients with less aggressive disease. That
`said, it is remarkable that the advantages of anti-TNF
`approach have been so widespread. The benefits are
`striking and reproducible, representing a decided ad-
`dition, if not improvement, on currently employed ther-
`apies. It is equally important now to see whether these
`results improve further as the duration of treatment is
`prolonged over the 6-month trials initially employed
`and if this approach serves patients with less severe
`disease better than standard options. Trials with other
`agents such as pentoxifylline have not produced as
`encouraging results as the antibodies or the use of
`soluble receptors (Dubost et al., 1997). A new approach
`being proposed now is to try blocking signals upstream
`of TNF-a such as the transcription factor NFkB that
`may be driving the inflammatory process (Bondeson et
`al., 1999; Wang et al., 1996) and this is being attempted
`via delivery of signals through vectors such as adeno-
`viruses. In other words, with the development of the
`concept, and then the successful demonstration of cy-
`tokine manipulation to control the signs and symptoms
`of a chronic disease like RA, we have officially entered
`the truly exciting, modern era of genetic engineering.
`CROHN’S DISEASE
`Although the incidence of this idiopathic chronic in-
`flammatory bowel disease is increasing in Western Eu-
`rope and the United Stateas, the cause remains un-
`known. Intestinal (bacteria) antigens, T-lymphocytes,
`and cytokines play the key roles in the pathogenesis of
`these disabling, life-long disorders (Fiocchi, 1997; Pow-
`rie, 1995). Increased TNF-a levels lie at the center of
`the cytokine abnormalities (Breese and McDonald,
`1995), making the TNF neutralizing strategy particu-
`larly attractive. Infliximab or the cA2 monoclonal an-
`tibody was the first in this series of agents to produce
`encouraging and often dramatic results (Hanauer,
`1999; Prescent, 1999; Rutgeerts and Paert, 1999; Tar-
`gan et al., 1997; Van Dullermen et al., 1995). Following
`its approval for use in Crohn’s disease by the FDA in
`1998, it has now also received positive advice for the
`Europeans Evaluation Agency in 1999. Infliximab acts
`rapidly, clinical
`improvements becoming apparent
`within 5 days of starting the treatment, and lasting for
`10–12 weeks in most patients. Further investigation
`into the mechanism of Infliximab action yielded new
`insights into the pathogenesis of Crohn’s disease. It
`had been demonstrated that mucosal T-lymphocytes in
`
`Dr. Reddy’s Laboratories, Inc. v. Celgene Corp.
`IPR2018-01504
`Exhibit 2014, Page 2
`
`

`

`ANTI-TNF THERAPIES
`
`231
`
`Crohn’s disease are apoptosis-resistant (Biorivant et
`al., 1999). Infliximab was shown to increase the ratio of
`Bax (pro-apoptotic) and Bcl-2 (anti-apoptotic), result-
`ing in an increased rate of programmed cell death in
`previously activated Jurkat T-lymphocytes in vitro
`(Ina et al., 1999). These data suggest that in addition to
`its direct anti-TNF activity, Infliximab may also have
`important anti-inflammatory effects by targeting acti-
`vated T-lymphocytes. This is not the first time an agent
`has produced the desired effect in an unpredictable
`manner. A precise understanding of the mechanism of
`action of anti-TNF agents will doubtless lead to the
`development of more effective therapies. The positive
`clinical experience with Infliximab has opened the door
`for trials with other anti-TNF agents in Crohn’s dis-
`ease. The humanized CDP571 antibody has shown ef-
`ficacy in this disease and other TNF-neutralizing strat-
`egies are being tested in clinical trials now (Stack et al.,
`1997; Van Deventer, 1997).
`SEPSIS
`Endotoxins in septic states lead to an outpouring of
`TNF-a from monocytes and macrophages resulting in
`the clinical manifestations of septic shock (Beutler et
`al., 1986). Suppression of TNF-a in this setting should
`theoretically lead to a dual benefit (Hinshaw et al.,
`1990). On the one hand, it should interrupt the cyto-
`kine cascade being proximally initiated by its activity,
`and, on the other hand, prevent the end-organ damage
`being mediated distally (Hinshaw et al., 1990). In an-
`imal studies, blocking TNF-a effects after experimen-
`tally producing septic shock has been highly effective
`in reducing both morbidity and mortality (Tracey and
`Cerami, 1993). It must be remembered however, that
`in animals, septic shock is produced acutely in other-
`wise healthy animals with TNF-a, while actual septic
`shock in humans usually is the end-stage manifesta-
`tion of an enormously immunocompromised and, usu-
`ally, chronically ill patient with previous multi-organ
`disease already present (Blackwell and Christman,
`1996). To begin with, therefore, a comparison of anti-
`TNF therapies in these two situations is plagued by
`widely different starting points. It is no surprise then
`that at least two clinical trials using anti-TNF mono-
`clonal antibodies have failed to show a significant ben-
`efit in 28-day all-cause mortality in patients with sep-
`sis (Abraham, 1998; Abraham and Wunderink, 1995;
`Cohen and Carlet, 1996; Porter, 1997). A trend in im-
`proved survival was observed in the subgroup of pa-
`tients with severe sepsis, including those with dysfunc-
`tion of two or more organ systems or with septic shock
`associated with the dysfunction of at least one organ
`system (Abraham, 1998), and, therefore, the final word
`on the efficacy of suppressing TNF activity in septic
`shock must await the results of larger multi-center
`trials presently underway. Other approaches such as
`IL-1 receptor antagonist administration to septic pa-
`tients is also being tested (Fisher et al., 1994).
`MYELODYSPLASTIC SYNDROMES or MDS
`A particularly interesting paradox has been de-
`scribed in the baffling; heterogeneous hematopoietic
`disorders grouped under the umbrella of MDS (Heaney
`and Golde, 1999). Besides the dysplastic morphology,
`the hallmark of this stem cell disease, which generally
`affects the elderly, is a variable peripheral cytopenia in
`
`the face of cellular bone marrows (BM). Given the
`monoclonal nature of the disease (Janssen et al., 1989),
`it is clear that the initial abnormality strikes a pluri-
`potential stem cell, an event that confers a proliferative
`advantage on the transformed cell leading to an un-
`checked clonal expansion (Raza et al., 1996a–c). The
`robustness of this clone is such that eventually a mono-
`clonal hematopoiesis is established with all blood cells
`including the B-lymphocytes and monocytes sharing a
`single parent ancestral cell (Prehal et al., 1978). The
`unexpected finding in view of the obvious intramedul-
`lary fitness of the rapidly dividing clone is the inability
`of the maturing daughter cells to survive and make it
`to the periphery. A series of studies designed to inves-
`tigate the cellular dynamics of the bone marrows in
`MDS patients have been conducted in the last decade
`and have led to some rather interesting insights into
`the pathogenesis of this disease. Briefly, using sophis-
`ticated in vivo labeling of S-phase cells by the thymi-
`dine analogs iododeoxyuridine and/or bromodeoxyuri-
`dine (IudR and BrdU, respectively), we were able to
`show that there is increased proliferative activity in
`the marrows of MDS patients with large numbers of
`cells actively synthesizing DNA (Raza et al., 1997).
`Peripheral cytopenia in the face of this actively prolif-
`erating BM was shown to be the result of excessive
`cytokine-mediated intramedullary apoptotic death of
`maturing hematopoietic cells (Raza et al., 1995). The
`two major pro-apoptotic cytokines were subsequently
`identified as being IL-1b (Mundle et al., 1996) and
`TNF-a (Shetty et al., 1996). In other words, the very
`cytokines driving the rapid proliferation of the expand-
`ing clone were also responsible for causing the innocent
`bystander death in their daughters. A new paradigm
`for MDS was proposed thereafter, which can be sum-
`marized as follows (Raza et al., 1996). An as yet poorly
`understood event or series of events transforms an
`early hematopoietic progenitor, leading to clonal ex-
`pansion and eventual monoclonal hematopoiesis. In
`this setting, pro-inflammatory cytokines such as
`TNF-a serve the dual function of stimulating the pro-
`liferation of the dividing cells and inducing apoptosis in
`their maturing progeny. We further proposed that if
`this model is correct, then suppression of the responsi-
`ble cytokine would also lead to a dual benefit, on the
`one hand an improvement in the cytopenias should be
`apparent as the maturing cells stop apoptosing, and, on
`the other hand, removal of the cytokine providing the
`stimulus for the proliferation of the MDS clone should
`lead to its regression with eventual restoration of poly-
`clonal hematopoiesis. Rarely have insights into the
`biology of a disease become translated into improved
`therapies for the patients with such alacrity as was
`accomplished in the case of translational research con-
`ducted in MDS. In just the last 4 years, a series of
`clinical trials have been completed to test the efficacy of
`anti-cytokine therapies in improving the cytopenias
`experienced by MDS patients. These will be summa-
`rized briefly here.
`Pentoxifylline (PTX), Ciprofloxacin (Cipro), and
`Dexamethasone or PCD Therapy With or
`Without Amifostine
`Initial attempts to suppress TNF-a were undertaken
`using the rather non-specific drug PTX, which is known
`to interfere with the second messenger pathways of a
`
`Dr. Reddy’s Laboratories, Inc. v. Celgene Corp.
`IPR2018-01504
`Exhibit 2014, Page 3
`
`

`

`232
`
`A. RAZA
`
`variety of cytokines including TNF-a, IL-1b, and trans-
`forming growth factor beta (TGF-b) as well as its ac-
`tivity in downregulating the mRNA for TNF-a. Cipro
`was simultaneously used to block the hepatic degrada-
`tion of PTX by interfering with the P450 reductase
`system and dexamethasone was added with the intent
`of potentiating TNF-a suppression since it is known to
`inhibit the translation of mRNAs into proteins. The
`first clinical trial yielded a response rate of approxi-
`mately 40%, however no complete responses were ob-
`served (Raza et al., 1998). The patients with partial
`responses generally showed improvements in the cyto-
`penias that have been reported in detail elsewhere. An
`unexpected finding was the observation of occasional
`disappearance of karyotypically abnormal cells. Since
`cytogenetically marked clones could only be eliminated
`with cytotoxic therapies thus far, their disappearance
`in response to anti-cytokine therapy made this ap-
`proach even more attractive. Furthermore, regression
`of these clones in response to anti-cytokine therapy
`also provided direct and incontrovertible proof that
`these clones were dependent for their proliferative ad-
`vantage on the cytokines being suppressed in vivo by
`therapy. Occasionally these abnormal metaphases dis-
`appear despite continued presence of MDS, suggesting
`that within the marrow of an individual MDS patient,
`there is continuous evolution of competing clones, some
`of which are more fit to survive than others. These
`evolving clones are obviously driven by the cytokine
`whose suppression led to their disappearance, whereas
`the parent clone may be dependent for its proliferation
`on cytokines other than those being neutralized or be-
`ing cytokine independent completely and autono-
`mously driven. Thus, this first study confirmed many of
`the hypotheses proposed to explain the complex, para-
`doxical pathology of MDS in a direct, in vivo setting. In
`the next study, PCD was combined with the cytopro-
`tective agent amifostine to attempt further improve-
`ment in the clinical outcome of patients (List et al.,
`1997). This combination resulted in a response rate of
`approximately 60%, which was better than that seen
`with either therapy alone (Raza et al., 1999a,b). Unfor-
`tunately, in both clinical trials, it was impossible to
`precisely define the mechanism of improvement in the
`hematopoietic indices since multiple agents were em-
`ployed in each study and none of the drugs were spe-
`cifically targeting any particular cytokine. Despite this
`limitation, we were able to show that the mechanism of
`action is cytokine related since responders showed the
`most sustained reductions in TNF-a levels (Reza et al.,
`1998). Obviously, the real test of the efficacy of anti-
`TNF therapy in ameliorating the cytopenias of MDS
`patients can only come when this cytokine is directly
`suppressed by using antibodies or soluble receptors
`described for RA and Crohn’s disease earlier. These
`studies are described below. Nonetheless, the early,
`albeit modest, success of these clinical trials estab-
`lished the feasibility of this approach and paved the
`way for the development of an entire new area of ther-
`apeutic research based upon the principle of cytokine
`modulation to improve the cytopenias in MDS.
`Enbrel
`In an attempt to directly test the proposition that the
`pathognomonic cytopenias in MDS patients are cyto-
`kine, and more specifically TNF-a mediated and that
`
`neutralizing TNF-a levels in these patients will result
`in improvements in the blood counts, TNF-a receptor
`Enbrel was administered in vivo to these individuals
`(Raza et al., 1999). A dimer of two monomers of the
`extracellular p75 TNF receptor fused to the Fc portion
`of a type I human immunoglobulin (TNFR:Fc) has been
`marketed by Immunex under the brand name, Enbrel.
`We administered Enbrel at a dose of 25 mg/dose twice
`weekly to 20 MDS patients for a period of 12 weeks. Of
`the 20 MDS patients, 9 were in the refractory anemia
`(RA) category, 5 had RA with ringed sideroblasts
`(RARS), and 6 had RA with excess blasts (RAEB) ac-
`cording to the French-American-British (FAB) classifi-
`cation. The median age was 67 years, there were 9
`females and 11 males, and all patients had primary de
`novo MDS without a prior history of toxic or chemical
`exposure. Enbrel was remarkably well tolerated in this
`group of patients with no serious adverse events. Two
`patients experienced a rash at the injection site. A total
`of 18 patients were able to complete the twelve weeks
`of therapy prescribed and were available for response
`evaluation. There were no complete responders, which
`is defined as the return of all peripheral blood indices
`to normal as well as normalization of the blast count in
`the BM. Partial responses (PR) were defined as one of
`the following: (1) Reduction in red cell transfusions by
`at least 50% or an increase in hemoglobin by at least
`2 Gm/dL; (2) Increase in absolute neutrophil count
`(ANC) by 500/ml over the baseline; and (3) Increase in
`platelet count by 30,000/mL over the baseline. Amongst
`the 18 evaluable patients, 10/18 showed a partial re-
`sponse. Two patients had a trilineage response while
`the rest either had a bilineage response or responded in
`a single lineage. Of the 10 responders, 2 showed a
`.50% reduction in transfusions, 8 had a greater than
`30,000/mL increase in the platelets, while 6 had a
`.500/mL increase in ANC. The increase in ANC was
`statistically significant when responders were com-
`pared to non-responders (P 5 0.051). Patients who had
`a higher pre-therapy biopsy cellularity tended to re-
`spond better than those who had a hypocellular BM
`(median cellularity 80% vs. 40%, respectively, P 5
`0.033). Thirteen patients presented with abnormal cy-
`togenetics while 7 had a normal karyotype prior to
`starting therapy. The most frequent cytogenetic abnor-
`mality was that affecting chromosomes 5 and/or 7 (5
`patients), while one had trisomy 8, three had deletions
`of 20q, and four had other non-specific karyotypes. It
`was interesting to note that 5/7 patients with normal
`cytogenetics showed a PR whereas 7 of the non-re-
`sponders had an abnormal karyotype. Responding pa-
`tients belonged to all different FAB categories. In sum-
`mary, therefore, it appears that suppression of TNF-a
`using the soluble receptor can indeed have profound
`biological effects in MDS patients and that some of
`these effects can clearly be translated into clinical re-
`sponses as well. The patients most likely to respond are
`those having normal cytogenetics and a hypercellular
`BM, indicating that the pathogenesis in MDS patients
`with abnormal karyotypes and/or hypocellular mar-
`rows may be more complicated, with the cytokine-
`driven component playing a less critical role in these
`patients than in the case of MDS with normal cytoge-
`netics and a hypercellular marrow. Another possibility
`is that the predominant abnormality in the non-re-
`sponders may reside in the autonomously proliferating
`
`Dr. Reddy’s Laboratories, Inc. v. Celgene Corp.
`IPR2018-01504
`Exhibit 2014, Page 4
`
`

`

`ANTI-TNF THERAPIES
`
`233
`
`stem cells rather than being the result of a major
`contribution from any single cytokine. Given the het-
`erogeneous nature of MDS, it is not too far-fetched to
`assign a variety of etiologies to the variable cytopenias
`encountered. Finally, the types of responses seen in
`this pilot study also warrant further comment. Anemia
`is the hallmark of disease in MDS and gross morpho-
`logical abnormalities of the erythroid series in the bone
`marrow are the rule. Yet, the majority of patients who
`responded to Enbrel, albeit partially, showed an im-
`provement in platelets (8/10) or in ANC (6/10), while
`only 2/10 responders had a reduction in transfusion
`requirements but without any substantial increase in
`their hemoglobin levels. The precise reason for this is
`unclear at the moment, but it is very possible that this
`only indicates an inadequate duration of therapy, with
`partial improvement in the less serious cytopenias be-
`coming apparent earlier than the more profound ones.
`Thalidomide
`In addition to a cytokine-mediated excessive in-
`tramedullary apoptosis of hematopoietic cells, the bone
`marrows of MDS patients have also been found to show
`increased angiogenesis, and higher than normal levels
`of vascular endothelial growth factor (VEGF) (Pruneri
`et al., 1999). Thalidomide would be a natural choice for
`MDS patients since this drug both suppresses TNF-a
`and inhibits neo-angiogenesis (Klausner et al., 1996). A
`pilot study has been conducted in MDS patients using
`thalidomide as a single agent at a starting dose of
`100 mg po hs and increasing as tolerated to 400 mg po
`hs (Raza et al., 1999). Therapy was continued for at
`least 12 weeks and further in responding patients or
`continued at the discretion of the investigator. Thirty-
`three patients were accrued on the study at the time of
`writing and 20 patients are evaluable for a response.
`Ten patients have demonstrated a partial response
`according to the criteria described above, while no pa-
`tient had a complete response. The median age for this
`group was 69 years. As per FAB classification, 20 pa-
`tients had RA1RARS, 12 had RAEB, and 2 had chronic
`myelomonocytic leukemia (CMMoL). Of the 10 re-
`sponding patients, 8 demonstrated an erythroid re-
`sponse, 4/10 showed a platelet response, and one had
`improved ANC. The therapy was quite well tolerated.
`The most important thing about the responders was
`the fact that three long-term transfusion dependent
`individuals (1, 3, and 3 years) all of whom had refrac-
`tory anemia became transfusion-independent and their
`hemoglobin levels showed an increase as well. Com-
`panion biologic studies conducted to examine the mech-
`anism of response have failed to identify either a de-
`finitive suppression of TNF-a levels or a decrease in
`angiogenesis as being consistently associated with a
`response. In other words, although there is a general
`tendency for both the TNF-a levels as well as micoves-
`sel density to decrease in the post-therapy marrows of
`the majority of treated patients, these biological effects
`do not necessarily translate into a clinical response. It
`must be noted that the protocol is still accruing pa-
`tients and that these conclusions are by no means final.
`A curious observation, however, is the matter of occa-
`sional “delayed” responses being seen with this ap-
`proach, which suggest that thalidomide may be exert-
`ing some of its actions via immune modulation. In one
`case, hemoglobin continued to improve 100 days after
`
`therapy began, and in another, following little improve-
`ment during a 12-week course of thalidomide and 10
`weeks after termination of therapy during which the
`patient only received supportive care, a trilineage re-
`sponse gradually set in. These effects are reminiscent
`of anti-thymocyt globulin (ATG) induced responses in
`aplastic anemia where it can take as long as 6 months
`after the administration of ATG for a response to be-
`come apparent (Teramura and Mizoguchi, 1996). In
`fact, ATG has now shown activity in MDS as well
`(Molldrem et al., 1997) and it is very likely that at least
`in some patients, thalidomide and MDS are producing
`a response via the same mechanism of T-cell suppres-
`sion. In summary, therefore, thalidomide appears to be
`an exciting new addition in the therapeutic armament
`of anti-TNF therapies, and specifically brings some
`long-awaited good news to the MDS cases for most of
`which supportive care has remained the “standard”
`treatment of choice to date.
`SUMMARY OF THE RATIONALE AND
`RESULTS OF ANTI-TNF THERAPY IN MDS
`Myelodysplastic syndromes are baffling hematopoi-
`etic disorders whose precise etiology continues to evade
`researchers. The disease is more common in the elderly
`and in those exposed to chemicals (such as benzene),
`and toxic agents such as chemotherapies for a prior
`malignancy. While abnormalities of chromosomes 5
`and/or 7 are common, especially in those with second-
`ary MDS, yet there is no proof to date that these ab-
`normalities are etiologically related. Rather, they ap-
`pear to represent derivative populations and evolving
`clones since they usually constitute a sub-population of
`cells in an otherwise monoclonal BM. There is increas-
`ing suspicion that these states may represent a chronic
`inflammatory response rather than a true malignant
`disease (Raza, 1998a,b). An infectious etiology has also
`been proposed with some intriguing initial observa-
`tions implicating the herpesviruses as possible culprits
`in at least some of the MDS patients. However, disease
`association awaits further proof and research (Mundle
`et al., 1999). In this context, a cytokine-based reper-
`toire of signs and symptoms assumes yet another sig-
`nificant dimension. As noted in the section on rheuma-
`toid arthritis, TNF-a can play the role of the master
`switch that turns a cascade of cytokines on and off,
`causing much chronic pathology and end-organ dam-
`age. A similar scenario in MDS can be visualized where
`a perpetual ongoing inflammatory process in the bone
`marrow could result in the selection of a clone capable
`of cytokine-driven rapid proliferation, whose eventual
`expansion would lead to monoclonal hematopoiesis,
`while cytokine-induced innocent bystander death of
`maturing hematopoietic cells would result in a variable
`cytopenia. Beautiful hypotheses have a nasty habit of
`being unraveled by ugly facts. In this case, however,
`the very first pilot study using rather non-specific
`drugs (PCD) to suppress a variety of pro-inflammatory
`cytokines produced exciting responses giving credence
`to the proposed new model of MDS that the major signs
`and symptoms of the disease were cytokine-mediated.
`Thus, with the etiology of MDS still shrouded in mys-
`tery, it became possible to provide substantial pallia-
`tio