Early reports
`
`Randomised comparison of thalidomide versus placebo in toxic
`epidermal necrolysis
`
`EARLY REPORTS
`
`Pierre Wolkenstein, Jacques Latarjet, Jean-Claude Roujeau, Corinne Duguet, Sylvie Boudeau, Loïc Vaillant,
`Michel Maignan, Marie-Hél`ene Schuhmacher, Brigitte Milpied, Alain Pilorget, Hél`ene Bocquet,
`Christian Brun-Buisson, Jean Revuz
`
`Summary
`
`Background Toxic epidermal necrolysis (TEN) is associated
`with a 30% death rate. Tumour necrosis factor ␣ (TNF-␣)
`has been
`implicated
`in
`the pathogenesis of TEN.
`Thalidomide is a potent inhibitor of TNF-␣ action. We did a
`double-blind,
`randomised, placebo-controlled study of
`thalidomide in TEN.
`
`Methods The patients
`received a 5-day course of
`thalidomide 400 mg daily or placebo. The main endpoint
`was the progression of skin detachment after day 7.
`Secondary endpoints were the severity of the disease,
`evaluated with the simplified acute physiology score
`(SAPS), and the mortality. TNF-␣ and interleukin 6 were
`measured.
`
`Findings The study was stopped because there was excess
`mortality in the thalidomide group—ten of 12 patients died
`compared with three of ten in the placebo group (Fisher’s
`exact test with Katz’s approximation, relative risk=2·78,
`p=0·03). After adjustment for SAPS, mortality remained
`significantly higher in the thalidomide group than in the
`placebo group (exact logistic regression mid-p=0·007; 95%
`infinity). Plasma TNF-␣
`CI
`for odds
`ratio 2·7
`to
`concentration was higher in the thalidomide group than the
`placebo group on day 2, though the difference was not
`significant (Wilcoxon rank-sum test p=0·07).
`
`Interpretation Even though few patients were included, our
`data suggest that thalidomide is detrimental in TEN,
`possibly because of a paradoxical enhancement of TNF-␣
`production.
`
`Lancet 1998; 352: 1586–89
`
`Department of Dermatology ( P Wolkenstein MD, J C Roujeau MD,
`H Bocquet MD, J Revuz MD) and Intensive Care Department
`( C Brun-Buisson) Hôpital Henri-Mondor, University Paris XII,
`Créteil; Burn Unit, Hôpital Saint Luc, Lyon (J Latarjet MD); Clinical
`Research Department, Laboratoires Laphal, Paris ( C Duguet MD,
`S Boudeau MD); Department of Dermatology, Hôpital Trousseau,
`Tours (L Vaillant MD); Department of Infectious Diseases, Hôpitaux
`de Brabois, Vandoeuvre ( M Maignan MD, M H Schumacher MD);
`Department of Dermatology, Hôtel-Dieu, Nantes ( B Milpied MD);
`and Burn Unit, Hôtel-Dieu, Nantes, France ( A Pilorget MD)
`Correspondence to: Dr Jean Revuz
`(e-mail: jean.revuz@hmn.ap-hop-paris.fr)
`
`Introduction
`Toxic epidermal necrolysis (TEN) is a rare, acute, and life-
`threatening condition. The incidence is one case per
`million inhabitants per year.1 Apoptosis of cells causes
`erosions of the mucous membranes, extensive detachment
`of the epidermis, and severe constitutional symptoms.
`Cases with the most extensive skin detachment are
`associated with the poorer prognosis, and a 30–40% death
`rate. Milder
`forms are known as Stevens-Johnson
`syndrome (SJS) or SJS/TEN overlap.2 TEN is usually
`drug-related.3
`At present there is no specific treatment for TEN. Some
`retrospective studies claimed a benefit of corticosteroids in
`milder forms,4 whereas several showed no benefit or even
`increased morbidity and mortality.5–7 Plasmapheresis,
`cyclosporin, cyclophosphamide, and N-acetylcysteine have
`been used in isolated cases and short uncontrolled series,8–11
`allowing no conclusion on the efficacy. Nevertheless,
`because the extent of final epidermal detachment is the
`main prognostic factor,12 therapies with the potential to
`stop the process of epithelial necrosis would be valuable
`during the initial phase of the disease.
`Apoptosis is the mechanism of keratinocyte death in
`TEN, and tumour necrosis factor ␣ (TNF-␣) is the likely
`cause for this and for constitutional symptoms during
`TEN.10,13 Thus, TNF-␣ production is believed to be an
`early pathogenetic event in TEN.
`Thalidomide is a potent inhibitor of TNF-␣ in vitro and
`in vivo,14,15 and appeared beneficial in several acute
`disorders thought to involve TNF-␣.16,17 We undertook a
`randomised placebo-controlled study of thalidomide in
`patients with TEN with the aim of testing the efficacy and
`safety of thalidomide in stopping the necrolysis process and
`reducing systemic symptoms during the initial phase of
`extension in TEN.
`
`Methods
`Patients
`Patients were enrolled from nine centres representing recruitment
`of about half the cases of TEN in France. Patients over 18 years
`old were eligible if they had detachment of epidermis of more than
`10% of body surface area,2 if the disease was still in the initial
`phase of extension and so had evolved for less than 4 days after the
`first mucocutaneous symptoms, and if they were expected to
`survive longer than 48 h. Diagnosis of TEN had to be confirmed
`with photographs and skin biopsy samples showing full-thickness
`detachment of epidermis, excluding staphylococcal scalded-skin
`syndrome. Patients were not eligible if their skin detachment was
`already above 90% of the body surface area, if skin detachment
`had not progressed during the previous 48 h, and if they had
`received therapies that have been claimed to influence TEN
`evolution (systemic corticosteroids, plasmapheresis, cyclosporin,
`
`THE LANCET • Vol 352 • November 14, 1998
`
`1
`
`IPR2018-00685
`Celgene Ex. 2024, Page 1
`
`

`

`numbers. These numbers were assigned to the capsule boxes.
`Local investigators telephoned a private randomisation service and
`were given an identification number that matched numbers on
`capsule boxes distributed in the centres. This schedule was
`prepared by Laboratoires Laphal; the investigators were unaware
`of the allocation. A set of sealed envelopes containing the code
`were supplied to each centre. In an emergency, the code could be
`broken and the investigator was required to write, sign, and date
`an explanation. During the study, the code was held by the
`Clinical Research Department of Laboratoires Laphal. Patients
`received standard supportive medical therapy (fluids, vasopressors,
`antibiotics, haemodynamic monitoring, as needed), and the
`putative culprit drugs were withdrawn.
`On days 0, 5, and 7 the extent of epidermal detachment
`(erosion, blisters, and areas with positive Nikolsky sign) was
`measured and expressed as the percentage of body surface area
`according to classic burns tables.18
`The main end point of the study was the progression of skin
`detachment after day 7, assessed as the difference of percentage of
`skin detachment between day 7 and day 0. Day 7 was chosen so
`that we could detect a potential rebound after the end of
`treatment. We also examined progression of skin detachment at
`day 5, which corresponded to the end of the treatment and to the
`average delay of skin-detachment progression in TEN. Other
`endpoints were the overall mortality and the severity of disease
`evaluated with the simplified acute physiology score19 (SAPS) at
`days 5 and 7. SAPS is a prognosis score calculated from seven
`clinical variables (age, heart rate, systolic blood pressure, body
`temperature, respiratory rate, urinary output per 24 h, Glasgow
`coma score) and seven biological variables (blood urea, packed-
`cell volume, white bloodcell count, and plasma concentrations of
`glucose potassium, sodium, and bicarbonate). The variations of
`SAPS were expressed as the difference in SAPS at days 5 and 7
`versus day 0.
`Plasma and blister-fluid samples for the measurement of TNF-
`␣ and interleukin 6 by EIA before and during treatment were
`obtained on days 0 and 2 and frozen at ⫺80ºC. Samples for the
`measurement of plasma thalidomide concentration were obtained
`on days 2 and 5. Assays for TNF-␣, interleukin 6, and
`thalidomide were done simultaneously at the end of the study.
`
`Statistics
`A sample size of 50 patients (25 in each group) was chosen on the
`basis of the variability in the maximum percentage of skin
`detachment in previous TEN series from Créteil centre.12,20 This
`sample size allowed us to detect a difference in progression of the
`percentage of skin detachment between placebo and thalidomide
`groups of 10% at day 7, with the assumption of an SD of 14%
`with 80% power at p=0·05 with a one-tailed test.
`t-tests,
`For continuous variables, appropriate two-sample
`Wilcoxon non-parametric rank-sum, or signed-rank tests were
`used. For categorical variables, Fisher’s exact test was used.
`Because of small numbers, mortality was analysed by exact
`statistics methods (StatXact Turbo, LogXact Turbo, Cytel
`Software Corporation, Cambridge, MA, USA). Survival was
`compared by the Kaplan-Meier method and a log-rank test. An
`exact logistic regression21 was done to adjust mortality for potential
`confounding factors. We did all analyses by intention to treat.
`
`Results
`Study population and clinical data
`22 patients were enrolled in the study (figure 1) from May,
`1995, to September, 1996. Of 15 patients at a single centre
`(Créteil), nine (60%) had died. This unusually high
`mortality rate alerted the local investigators. The trial
`coordinator was informed and decided to convene a safety
`board of three experts. The safety board first looked at the
`overall data without breaking the code—13 of 22 patients
`enrolled had died (overall mortality 59%). Because this
`rate was higher than expected from previous series of TEN
`and the baseline severity of patients, the data were analysed
`
`EARLY REPORTS
`
`34 patients approached
`
`12 not included
`10 not eligible
`2 refused
`
`22 randomised
`
`10 assigned and
`received placebo
`
`12 assigned and
`received thalidomide
`
`1 died
`
`2 died
`
`9 analysed for
`main endpoint
`
`8 analysed for
`main endpoint
`
`4 died
`
`6 died
`
`7 completed
`study
`
`2 completed
`study
`
`Figure 1: Trial profile
`
`cyclophosphamide) or other experimental drugs targeted at TNF-
`␣ (such as oxipentifylline or monoclonal antibodies to TNF-␣).
`Women of child-bearing age had to have a negative serum
`pregnancy test before inclusion. HIV-1 infection was not an
`exclusion criterion.
`Eligible patients were included immediately after admission
`(day 0). Each patient or the closest relative gave written informed
`consent for the trial, which had been approved by the ethics
`committee of Henri-Mondor Hospital according to French law.
`
`Methods
`The patients were randomly assigned a 5-day course of
`thalidomide 400 mg daily or placebo (provided by Laboratoires
`Laphal, Paris, France). The thalidomide dose was chosen by
`analogy16 to that used in the treatment of graft-versus-host disease.
`The drug was given twice daily as two 100 mg thalidomide
`capsules or two placebo capsules, orally when possibly or by
`nasogastric
`feeding tube otherwise. The placebo and the
`thalidomide capsules were identical in appearance, and the
`investigators and patients did not know which capsules were given.
`The randomisation was done in blocks of six patients stratified
`according to two categories of study centres—dermatological
`centres or burns and intensive-care units. Two lists (one for each
`category of centres) were generated from tables of random
`
`Thalidomide (n=12)
`Placebo (n=10)
`
`100
`
`90
`
`80
`
`70
`
`60
`
`50
`
`40
`
`30
`
`20
`
`10
`
`0
`
`Cumulative survival (%)
`
`0
`
`5
`
`20
`15
`10
`Time from inclusion (days)
`
`25
`
`30
`
`Figure 2: Overall survival of thalidomide-treated patients
`compared with those receiving placebo
`
`2
`
`THE LANCET • Vol 352 • November 14, 1998
`
`IPR2018-00685
`Celgene Ex. 2024, Page 2
`
`

`

`EARLY REPORTS
`
`Median (range) in group
`
`Placebo group*
`
`Thalidomide group†
`
`p
`
`34 (16–392)
`36 (2–432)
`
`58 (16–775)
`93 (38–636)
`
`183 (16–1288)
`104 (0–314)
`
`323 (95–1285)
`103 (45–1279)
`
`0·62
`0·07
`
`0·51
`0·56
`
`206 (30–32 398)
`278 (70–20 087)
`
`146 (23–24 109)
`883 (57–11 424)
`
`0·57
`0·76
`
`Plasma TNF-␣ (ng/L)
`Day 0
`Day 2
`
`Blister fluid TNF-␣ (ng/L)
`Day 0
`Day 0
`
`Plasma interleukin 6 (ng/L)
`Day 0
`Day 2
`
`0·84
`0·60
`
`Blister-fluid interleukin 6 (ng/L)
`1412 (65–19 468)
`1732 (79–2843)
`Day 0
`7147 (88–37 802)
`722 (142–43 971)
`Day 2
`*n=8 on day 0; n=7 on day 2 (except for blister-fluid TNF-␣).
`†n=12 for day 0 plasma values; n=11 for day 0 blister-fluid values; n=9 for day 2
`plasma values; n=10 for day 0 blister-fluid values.
`Table 3: Plasma and blister-fluid TNF-␣ and interleukin 6
`concentrations
`day 2 (table 3). At day 2, the plasma TNF-␣ concentration
`was higher in the thalidomide group than the placebo
`group, though the difference was not significant (p=0·07).
`
`Discussion
`This study is the first double-blind, randomised, placebo-
`controlled trial of any therapy in TEN. Thalidomide was
`not effective in halting the necrolysis process during the
`initial phase of extension. On the contrary, thalidomide
`treatment was associated with increased mortality. This
`difference was not the result of an unexpectedly low rate of
`death in the placebo group, because the mortality rate in
`that group was of the same order in previous series.12,20 The
`mortality remained significantly higher in thalidomide
`receipients after adjustment for SAPS. The causes of death
`(multiple organ failure, septic shock, and acute respiratory
`distress) are the usual causes of death in most series of
`TEN.12,18
`Thalidomide has been used in several disorders in which
`TNF-␣ is thought to play a part14 and has shown evidence
`of efficacy in several.16,17,22 Nevertheless, thalidomide was
`not successful as prophylaxis for chronic graft-versus-host
`disease—patients receiving thalidomide had a higher rate of
`this complication, which resulted in a higher mortality
`rate.23
`The excess mortality in our study could be explained by
`three main causes. First, thalidomide therapy might have
`resulted in increased mortality through some of its known
`side-effects—central depression of ventilation by sedative
`effect or increased bacterial translocation by decrease in
`gastrointestinal motility.24 We observed no evidence in
`favour of these mechanisms. Second, anti-TNF-␣ might
`have a protective effect during TEN, as has been suggested
`in septic shock, in which anti-TNF-␣ agents may result in
`increased mortality.25 However, our data suggest that
`thalidomide did not inhibit TNF-␣ production. Third,
`thalidomide might have
`resulted
`in paradoxical
`overproduction of TNF-␣, explaining in part the excess
`mortality. In our study, plasma concentrations of TNF-␣
`tended to increase after treatment with thalidomide in
`comparison with the placebo group. Jacobson and
`colleagues22 observed a similar unexpected increase in the
`plasma concentrations of TNF-␣ and soluble TNF-␣
`receptors with thalidomide treatment for oral aphthous
`ulcers in HIV-1-infected patients. In-vitro findings26,27
`suggest that thalidomide, at concentrations achieved in
`vivo, could either enhance or suppress the synthesis of
`TNF-␣ depending on the type of cells stimulated. Thus,
`
`Age (years)
`M/F
`Weight (kg)
`Skin detachment (% BSA)*
`SAPS
`
`Group
`
`Placebo (n=10)
`
`Thalidomide (n=12)
`
`50·5 (23–58)
`4/6
`72 (46–105)
`30·5 (10–85)
`10·5 (6–17)
`
`53 (23–81)
`6/6
`56·5 (45–104)
`43·5 (26–90)
`11·5 (6–19)
`
`Data are median (range).
`*% of body surface area.
`Table 1: Demographic and severity features at entry
`
`by group; there were three deaths among ten patients in
`one treatment group compared with ten deaths among 12
`patients in the other. The decision to break the code was
`then taken, and the high death rate was found to be
`associated with thalidomide treatment. Mortality was
`confirmed to be signifcantly higher in the thalidomide
`group than in the placebo group (Fisher’s exact test with
`Katz’s approximation, relative risk 2·78 [95% CI
`1·04–7·40]; p=0·03). The safety board thus advised that
`the trial be stopped.
`Survival curves are plotted in figure 2. Median survival
`time was 10 days and more than 30 days, respectively, in
`the thalidomide and placebo groups. Demographic and
`severity features in the placebo and thalidomide groups are
`shown in table 1. Among the known prognosis indicators
`in TEN—percentage of skin detachment and SAPS at
`entry—only SAPS appeared to be a predictive factor for
`mortality. After adjustment for SAPS by exact logistic
`regression, mortality remained significantly higher in the
`thalidomide group than in the placebo group (exact logistic
`regression mid-p=0·007; 95% CI for odds ratios 2·7 to
`infinity).
`The progression of skin detachment and SAPS did not
`differ at day 5 or day 7 between the groups (table 2).
`
`Causes of death
`According to the investigators, death was attributed to
`several causes: multiple organ failure (two in placebo
`group, six in thalidomide group) septic shock (three in
`placebo group, five in thalidomide group) and acute-
`respiratory distress syndrome (three in thalidomide group).
`
`Biological data
`Serum concentrations of thalidomide in the treated group
`were 0·619 mg/L (SD 0·322) at day 2 (data on eight
`patients) and 0·454 mg/L (0·230) at day 5 (data on seven
`patients).
`The plasma and blister-fluid concentrations of TNF-␣
`and interleukin 6 were high in both groups at day 0 and
`
`p(
`
`Wilcoxon test)
`
`Median (range) change in group*
`
`Placebo
`
`Thalidomide
`
`Skin detachment
`Day 5 minus day 0† ⫺5% (⫺20% to 23%)
`Day 7 minus day 0‡ ⫺5·7% (⫺45% to 20%)
`
`⫺4% (⫺65% to 74%)
`1·00
`4·9% (⫺81% to 74%) 0·78
`
`SAPS
`Day 5 minus day 0† ⫺1 (⫺6 to 4)
`Day 7 minus day 0‡ ⫺0 (⫺9 to 4)
`
`⫺4 (⫺3 to 7)
`⫺2 (⫺4 to 9)
`
`0·14
`0·89
`
`Mortality
`0·03
`10/12 (83%)
`3/10 (30%)
`*Positive values indicate that skin detachment or SAPS increased to day shown;
`negative values indicate a decrease.
`†Data based on nine patients in each group (one patient in placebo group and three in
`thalidomide group died before day 5).
`‡Data based on nine patients in placebo group and eight patients in thalidomide group
`(one and four patients, respectively, died before day 7).
`Table 2: Responses according to skin detachment, SAPS, and
`mortality
`
`THE LANCET • Vol 352 • November 14, 1998
`
`3
`
`IPR2018-00685
`Celgene Ex. 2024, Page 3
`
`

`

`EARLY REPORTS
`
`thalidomide could enhance, in certain circumstances, the
`production of TNF-␣.28 Although the trial was stopped
`after inclusion of only a few patients, our findings suggest
`that thalidomide is detrimental in TEN, possible because
`of a paradoxical enhancement of TNF-␣ production.
`
`Contributors
`Jean Revuz, Pierre Wolkenstein, Jean-Claude Roujeau, and
`Jacques Latarjet designed the study. Corinne Duguet and Sylvie Boudeau
`collected and checked the data. Loïc Vaillant, Michel Maignan,
`Marie-Hél`ene Schuhmacher, Brigitte Milpied, Alain Pilorget, and
`Hél`ene Bocquet did the investigations. Christian Brun-Buisson
`participated on the safety board and analysed the data with
`Pierre Wolkenstein. Pierre Wolkenstein, Jean-Claude Roujeau,
`Christian Brun-Buisson, and Jean Revuz wrote the paper.
`Acknowledgments
`We thank Eric Roupie for comments on the paper, and Ariane Auquier,
`Françoise Doyon, Sylvie Bastuji-Garin, and Emmanuelle Girou for
`statistical analyses. The study was supported by a grant from Laboratoires
`Laphal, Allauch, France.
`
`References
`1 Roujeau J-C, Stern RS. Severe cutaneous adverse reactions to drugs.
`N Engl J Med 1994; 331: 1272–85.
`2 Bastuji-Garin S, Rzany B, Stern RS, Shear NH, Naldi L, Roujeau J-C.
`A clinical classification of cases of toxic epidermal necrolysis,
`Stevens-Johnson syndrome and erythema multiforme. Arch Dermatol
`1993; 129: 92–96.
`3 Roujeau J-C, Kelly JP, Naldi L, et al. Medication use and the risk of
`Stevens-Johnson syndrome or toxic epidermal necrolysis. N Engl J Med
`1995; 333: 1600–07.
`4 Sherertz EF, Jegasothy BV, Lazarus GS. Phenytoin hypersensitivity
`reaction presenting with toxic epidermal necrolysis and severe hepatitis.
`Report of a patient treated with corticosteroid “pulse therapy”. J Am
`Acad Dermatol 1985; 12: 178–81.
`5 Halebian PH, Corder VJ, Madden MR, Finklestein JL, Shires GT.
`Improved burn center survival of patients with toxic epidermal
`necrolysis managed without corticosteroids. Ann Surg 1986; 204:
`503–12.
`6 Kim PS, Goldfarb IW, Gaisford JC, et al. Stevens-Johnson syndrome
`and toxic epidermal necrolysis: a pathophysiologic review with
`recommendations for a treatment protocol. J Burn Care Rehabil 1983; 4:
`91–100.
`7 Kelemen JJ, Cioffi WG, McManus WF, Mason AD, Pruitt BA. Burn
`center care for patients with toxic epidermal necrolysis. J Am Coll Surg
`1995; 180: 273–78.
`8 Kamanabroo D, Schmitz-Landgraf W, Czarnetski BM. Plasmapheresis
`in severe drug-induced toxic epidermal necrolysis. Arch Dermatol 1985;
`121: 1548–49.
`9 Hewitt J, Ormerod AD. Toxic epidermal necrolysis treated with
`cyclosporin. Clin Exp Dermatol 1992; 17: 264–65.
`10 Heng MC, Allen SG. Efficacy of cyclophosphamide in toxic epidermal
`
`necrolysis: clinical and pathophysiologic aspects. J Am Acad Dermatol
`1991; 25: 778–86.
`11 Redondo P, Defelipe I, Delapena A, Aramendia JM, Vanaclocha V.
`Drug-induced hypersensitivity syndrome and toxic epidermal
`necrolysis—treatment with N-acetylcysteine. Br J Dermatol 1997; 136:
`645–46.
`12 Revuz J, Penso D, Roujeau J-C, et al. Toxic epidermal necrolysis:
`clinical findings and prognosis factors in 87 patients. Arch Dermatol
`1987; 123: 1160–65.
`13 Paul C, Wolkenstein P, Adle H, Wechsler J, Revuz J, Roujeau J-C.
`Apoptosis as a mechanism of keratinocyte death in toxic epidermal
`necrolysis and Stevens-Johnson syndrome. Br J Dermatol 1996; 134:
`710–14.
`14 Klausner JD, Freedman VH, Kaplan G. Thalidomide as an anti-TNF-␣
`inhibitor: implications for clinical use. Clin Immunol Immunopathol 1996;
`81: 219–23.
`15 Moreira AL, Sampaio EP, Zmuidzinas A, Frindt P, Smith KA,
`Kaplan G. Thalidomide exerts its inhibitory action on tumor necrosis
`factor ␣ by enhancing mRNA degradation. J Exp Med 1993; 177:
`1675–80.
`16 Vogelsang GB, Farmer ER, Hess AD, et al. Thalidomide for the
`treatment of chronic graft-versus-host disease. N Engl J Med 1992; 326:
`1055–58.
`17 Klausner JD, Makonkawkeyoon S, Akarasewi S, et al. The effect of
`thalidomide on the pathogenesis of HIV-1 and M tuberculosis infection.
`J AIDS Hum Retroviruses 1996; 11: 247–57.
`18 Lund CC, Browder NC. The estimation of areas burns. Surg Gynecol
`Obstet 1994; 79: 352–59.
`19 Le Gall JR, Loirat P, Alperovitch A, et al. A simplified acute physiology
`score for ICU patients. Crit Care Med 1984; 12: 975–77.
`20 Correia O, Chosidow O, Saiag Ph, et al. Evolving pattern of drug-
`induced toxic epidermal necrolysis. Dermatology 1993; 186: 32–37.
`21 Mehta CV, Patel NR. Exact logistic regression: theory and examples.
`Stat Med 1995; 14: 2143–60.
`22 Jacobson JM, Greenspan JS, Spritzler J, et al. Thalidomide for the
`treatment of oral aphthous ulcers in patients with human
`immunodeficiency virus infection. N Engl J Med 1997; 336:
`1487–93.
`23 Chao NJ, Parkers PM, Niland JC, et al. Paradoxical effect of
`thalidomide prophylaxis on chronic graft-vs-host disease. Biol Blood
`Marrow Transplant 1996; 2: 86–92.
`24 Günsler V. Thalidomide in human immunodeficiency virus (HIV)
`patients. Drug Saf 1992; 7: 116–34.
`25 Fisher CJ, Agosti JM, Opal SM, et al. Treatment of septic shock with
`the tumor necrosis factor receptor: Fc fusion protein. N Engl J Med
`1996; 334: 1698–702.
`26 Shannon EJ, Sandoval F. Thalidomide can be either agonistic or
`antagonistic to LPS evoked synthesis of TNF-␣ by mononuclear cells.
`Immunopharmacol Immunotoxicol 1996; 18: 59–72.
`27 Nishimura K, Hashimoto Y, Iwasaki S. Enhancement of phorbol ester-
`induced production of tumor necrosis factor ␣ by thalidomide. Biochem
`Biophys Res Commun 1994; 199: 455–60.
`28 Zwingenberger K, Wnendt S. Immunomodulation by thalidomide:
`systematic review of the literature and of unpublished observations.
`J Inflamm 1996; 46: 177–211.
`
`4
`
`THE LANCET • Vol 352 • November 14, 1998
`
`IPR2018-00685
`Celgene Ex. 2024, Page 4
`
`