Thalidomide in Ankylosing Spondylitis F

Thalidomide in ankylosing spondylitis F. Huang1, J.C.C. Wei2, M. Breban3

1Chinese PLA General Hospital, Beijing, ABSTRACT born to mothers exposed to thalidomide China; 2Family Medicine Department, Despite potential side effects dominat - during their first trimester of pregnancy Chia Yi Veterans Hospital, Taiwan; ed by terat ogenicity and peri p h e ra l which lead to withdrawal of the drug 3 INSERM U567 and Service de Rhumatolo- n e u ro p at hy, thalidomide has re c e n t ly f rom the market. Unfo rt u n at e ly, t h i s gie B, Hôpital Cochin, Paris, France. been used to tre at seve re anky l o s i n g did not occur before 10,000 to 12,000 Feng Huang, MD, James C.C. Wei, MD, spondylitis (AS). Over 50 patients have ch i l d ren with birth defects had been Maxime Breban, MD, PhD. been treated across several 6-12 month born from mothers exposed to the drug Please address correspondance to: open studies. A l t ogether 68% of the (6). Maxime Breban, MD, PhD, INSERM U567 and Service de Rhumatologie B, Hôpital p atients improved and the dro p - o u t Since that time however, thalidomide Cochin, 27 rue du Faubourg Saint-Jacques, rate was 19%. Inhibition of NF- B has continuously been used under re- 75014, Paris, France. and/or TNF is a putative mechanism stricted conditions, because of its re- E-mail: [email protected] for thalidomide efficacy in AS. ported efficacy on rare refractory mani- paris.fr fe s t ations involving immu n o l ogi c a l Clin Exp Rheumatol 2002; 20 (Suppl. 28): History mechanism, such as erythema nodosum S158-S161. Thalidomide ( -N-phthalimidoglutari- leprosum (ENL), chronic cutaneous lu- © Copyright CLINICAL AND EXPERIMEN- mide) is a synthetic derivative of glu- pus, Behçet’s disease, or chronic graft- TAL RHEUMATOLOGY 2002. tamic acid, which was originally devel- versus-host disease (6). More recently, oped by a German company and put on Food and Drug Administration has ap- Key words: Ankylosing spondylitis, the market as a sedative, in Canada, proved thalidomide for the treatment of spondylarthropathy, thalidomide, South America, throughout Europe and ENL in US after strict rules and regula- TNF . some countries in Asia. The drug was tions we re established between the known for rapid onset of action, lack of agency and Celgene, the US manufac- hangover, and absence of respiratory or turer. s keletal mu s cle coord i n ation impair- ment. The body’s handling of this lipo- Possible mechanism of action philic molecule is still incompletely The mechanism of action of thalido- k n own. Pharmacokinetic studies in mide is progressively broken down (6). healthy human male volunteers suggest Va rious anti-infl a m m at o ry, i m mu n o- a one compartment, first order absorp- modulatory and anti-angiogenic activi- tion and elimination model, with peak ties have been described (Table I). plasma levels reached within 4-5 hr and Most remarkably, thalidomide is a rela- a half-life of 8.7 ± 4.1 hr.Thalidomide tively weak inhibitor of TNF synthe- is metab o l i zed mainly through non- sis in vitro. At therapeutic concentra- enzymatic hydrolytic cleavage generat- tions, it reduces TNF production by ing a multitude of active intermediates approximately 40%. This TNF inhibi- with largely unknown functions (Chen tory capacity occurs at the transcrip- et al., 1989). tional level through a reduction of the In 1960, two types of major side-effects TNF mRNA half-life and is selective, we re rep o rt e d, i . e. peri p h e ral neuro- with no significant effect on cytokines pathies in patients with long-term ex- s u ch as interleukin (IL)-l, I L - 6 , a n d posure, and congenital abnormalities of granu l o cy t e - m a c ro p h a ge colony- s t i m u- the limbs (phocomelia) in ch i l d re n lating fac t o r , by human monocytes sti-

Table I. Immunomodulatory and anti-inflammatory properties of thalidomide

Inhibits leukocyte chemotaxis into site of inflammation Alters density of TNFa -induced adhesion molecules on leukocytes Reduces phagocytosis by polymorphonuclear leukocytes Enhances mononuclear cell production of interleukins-4 and -5; inhibits interferon-g production Inhibits production of interleukin-12 Inhibits production of TNF by monocytes and macrophages by reducing half-life of mRNA.

S-158 Thalidomide in ankylosing spondylitis / F. Huang et al. mul ated with lipopoly s a c ch a ri d e wish to conceive, do not practice relia- Thalidomide use in ankylosing (LPS). ble contra c ep t i o n , or are judged by spondylitis Thalidomide’s inhibition of TNF pro- their health care provider to be inca- Up to recently, there was no effective duction is not always achieved in vivo. pable of practicing reliable contracep- therapy for patients with severe AS, re- Decreased TNFa serum concentration tion must be excluded from using tha- f ra c t o ry to non-steroidal anti-infl a m- have been described in patients suffer- lidomide. Furthermore, patient educa- matory drugs (NSAID), especially for ing from ENL or tuberculosis, treated tion about the proper and safe use of those with pre d o m i n a n t ly axial in- with thalidomide (8). However, TNFa thalidomide and serial pregnancy tests volvement. Given its pharmacological levels in blood conversely increased in are mandatory (6). properties, thalidomide has been con- t wo ra n d o m i zed controlled trials de- sidered as a putative anti-TNF agent. m o n s t rating thalidomide’s effi c a cy in Peripheral neuropathy The presence of TNF was demonstra- healing oral aphtae in the setting of Th a l i d o m i d e - a s s o c i ated peri p h e ra l ted in inflamed sacro-iliac joints of pa- HIV infection as well as reducing HIV- neuropathy has been described in detail tients affected with AS, both at messen- associated wasting. A similar observa- among the results of several large retro- ger RNA and protein levels (3). This tion was made in a study of patients spective studies published throughout observation prompted one of us to exa- with toxic ep i d e rmal necro ly s i s , i n the medical literature (6). Frequently, mine the efficacy of thalidomide in se- which outcome was adversely affected patients complain of painful paresthe- vere refractory AS. A regimen of 100- by thalidomide (18). Thus, it seems that sias and tingling sensations in the lower 300 mg/day was used.The first two thalidomide’s ability to inhibit TNFa ex t remities. Motor disturbances are patients who received the drug had a may depend on the stimulus or cellular generally mild and reversible on dis- clear benefit from thalidomide (5). Ef- source of cytokine production (6). continuation of the drug, whereas sen- ficacy of the drug concerned both axial Besides pro - i n fl a m m at o ry cy t o k i n e s sory symptoms may persist or incom- and peripheral clinical manifestations, inhibition, thalidomide exerts other in pletely resolve. The issue of irreversi- and also biological markers of inflam- vi t r o im mu n o m o d u l a ting activi t i e s , su c h bility is clouded by the retrospective mation (ESR and CRP). One of the 2 as T-cell costimu l at i o n , t h e reby en- nature of these studies and the collec- patients had prompt relapse after dis- hancing T cell response.This costimu- tion of data from patients whose condi- continuation of the drug, whereas the l at o ry effect ap p e a rs gre ater on the tions may themselves produce neuro- second was lost of follow-up. Efficacy CD8+ T cells than on the CD4+ T cells pathy. The reported incidence of tha- of thalidomide resumed after reintro- (16). There is no evidence that thalido- lidomide neuropathy varies from 0.5% ducing the treatment in the first patient. mide possesses immu n o s u p p re s s ive to 50% between series. Gardner-Med- Since this initial description, 10 addi- properties, and the drug has not been win and Powell have published guide- tional SpA patients have been treated a s s o c i ated with opportunistic infe c- lines for the use of electrophysiologic by the same author (4, 10). The dura- tions. A n t i a n gi ogenic pro p e rties of studies to monitor for neuro t ox i c i t y. tion of tre atment was 6 months fo l- thalidomide have also been found (9). They suggest baseline electromyogram lowed by a carry-over observation peri- Most recently, blocking of NF-kB acti- and nerve conduction velocity study re- od of 4 months. Of 12 patients enrolled vation through a mechanism that in- p e ated at 10 g or 6-month interva l s in this small-size open-labeled study, 5 volves the inhibition of IkB kinase was (which ever comes first) during treat- stopped thalidomide before 6 months shown with thalidomide, and was pro- ment. A decline in the sensory nerve because of side-effects. Evidence of posed as an upstream event responsible action potential of 50% or more, sub- clinical effic a c y was found in 7 of 12 pa- for the diffe rent actions of the dru g je c t i ve neurol o gic complaints at month- tients, among whom 2 achieved a re- (14). ly follow-up, or physical examination duction of Bath AS Disease Activity In- abnormalities necessitate a dose reduc- dex (BASDAI) greater than 50%. The Toxicity tion or discontinuation of thalidomide. most consistent efficacy however was Teratogenecity Whether electrophysiologic monitoring on biological param e t e r s of infla m ma- The most serious toxicity of thalido- p rovides tangi ble benefits avo i d i n g tion (ESR and/or CRP) mide is its teratogenicity dominated by neurotoxicity above and beyond detail- Lee at al. subsequently reported effica- phocomelia. Animal models sugge s t ed serial clinical assessments remains cy of thalidomide in one female patient that these teratogenic effects may occur to be proved (6). a ffected with seve re re f ra c t o ry SpA through the drug’s anti-angiogenic ef- (15). fects on the developing fetal tissues (6). Other adverse effects Huang et al. have completed a one-year The joint effort between FDA and tha- Taking thalidomide at bedtime can help open-labeled study in 30 male patients lidomide manufacturer resulted in the m i n i m i ze somnolence, d i z z i n e s s , a n d affected with severely active refractory establishment of the STEPS, (System morning hangovers. These effects tend AS, in Beijing (13). In this study, pa- for Thalidomide Education and Pre- to resolve after 2 to 4 weeks of continu- tients were required to meet six entry s c ribing Safety) progra m , a strin ge n t ous therapy. Constipat i o n , h e a d a ch e, criteria on two occasions in the month protocol aimed at avoiding the terato- nausea, weight gain, edema, and transi- prior to study entry: BASDAI > 4, total genic effects of the drug.Women who ent skin eruptions may also occur. body or spinal pain values >2 when

S-159 Thalidomide in ankylosing spondylitis / F. Huang et al.

Table II. Thalidomide in severe refractory SpA*.

Duration No. of premature No. with Study Patients treated: HLA- Dosage of treatment discontinua- clinical improve- diagnosis (n) B27 (mg/day) (mos.) tions (%) improvement (%) Center

Breban et al. AS (9), uSpA (2), PsA (1) 10/12 100-300 6 5 (42) 7/12 (58) Paris Lee et al. uSpA (1) 1/1 100 6 0 1/1 (100) Sydney Huang et al. AS (30) 30/30 200 12 4 (13) 21/30 (70) Beijing Wei et al. AS(10) NA 200 3 1 (10) 7/10 (70) Chia Yi

* SpA: spondylarthropathy; AS: ankylosing spondylitis; uSpA: undifferentiated spondylarthropathy; PsA: psoriatic arthrittis; NA: not available. evaluated on a four point Likert scale, Mechanism of action e ffects). Only one patient re q u i red a and early morning stiffness >30 min- As indicated ab ove, thalidomide wa s higher dose of 300 mg/day. In Taiwan, u t e s , ESR as well as CRP measure- initially tried in AS because it was one the initial dose was 100 mg/day for 1 ments higher than normal. Using a 20% of the few conventional drugs thought week followed by 200 mg/day for 11 improvement of 4/7 indices as primary to target TNF . In agreement with this weeks. Only one patient droped out af- response definition, 80% of the 26 pa- hypothesis, biological agents blocking ter 2 weeks for skin eruption. In the tients who completed the one year thal- T N F p roved re m a rk ably effi c a c i o u s French experience, a loading dose of idomide study demonstrated a positive in refractory AS (2, 17). Parallel effica- 300 mg/day was rather used and taper- response. About 50% of patients show- cy between thalidomide and anti-TNF ed there a f t e r. Frequent side-effe c t s ed a >50% improvement in 4/7 indices. biological therapy was also observed in were drowsiness, constipation, and diz- H oweve r, complete remission with refractory Crohn’s disease (11), as op- ziness. No major side-effect of thalido- improvement >75% in 7/7 indices was posed to rheumatoid art h ritis wh i ch mide, such as birth defects or neuro- not attainable.This positive observa- se e m ed to respond rather poorly to sim- pathy, has been reported so far in AS. tion of prim a r y outcome measure im- ilar doses of thalidomide (6). pr ovement was supported by improve- H oweve r, as alre a dy discussed, t h e Future directions ment of five parameters of secondary mechanism of action of thalidomide is Despite its tragic past, thalidomide has outcome measures. Most notable were more complex than just direct inhibi- recently shown promise in treating sev- significant decreases in both ESR and tion of TNF . Interestingly, Huang et eral disorders, including AS.Thalido- CRP. ESR and CRP were high in all al. have demonstrated inhibition of sev- mide is much less expensive than anti- patients prior to therapy. They returned e ral pro - i n fl a m m at o ry genes ex p re s- TN F bi o l o gical agents. However , co n - to normal in 4 and 14 patients respec- sion, in peripheral blood mononuclear trolled studies are still needed to assess tively after 6 and12 months of thalido- cells after 3 months of thalidomide the benefits and risks from short-term mide. therapy. Gene expression profiles were and long-term use. In recent years, sev- In Taiwan, Wei et al. have now com- tested by a 588-gene microarray. Seven e ral stru c t u ral analogues of thalido- pleted a 3 months open-labeled study thalidomide suppressed genes we re mide have been demonstrated to have of thalidomide in 10 male patients af- i d e n t i fi e d : T N F , I L - 1b, M 1 P - 1 , increased potency at inhibiting TNF fected with severely active AS, refrac- M1P-2 , c-jun, OX40 ligand and the T production as well as enhanced solubil- tory to NSAID. According to AS as- lymphocyte maturation-associated pro- ity and stability. Two different groups sessment study criteria (1) which were tein MAL. In contrast, such inhibition of molecules were obtained. used to evaluate efficacy of treatment, 7 was not observed during treatment with The immu n o m o d u l at o ry imide dru g s of the 9 completers (78%) were respon- infliximab (12, 13). (IMiDs) inhibit the production of ders (4 and 3 patients achieved 50% TNF , IL-1b and IL-12, and enhance and 20% improvement levels respec- Tolerance the production of IL-10, in monocyte ti vel y). Stati s t i c a l l y significant improve- O ve rall tolerance was we a ker in the s t i mu l ated by LPS. In add i t i o n , a n d ment was achieved for BASDAI, pa- Fre n ch patients tre ated with thalido- s i m i l a rly to thalidomide, these com- tients global assessment, ESR and mide than in those from China or Tai- pounds exert a costimulatory effect on CRP. wan. This could be due to genetic dif- T cells. In contrast, the selective cyto- Results from these different reports are ferences in the enzymatic capacity to kine inhibitory drugs (SelCiDs), which summarized in Table II. In both French detoxify the drug, but also to differ- a re potent phosphodiesterase Type 4 and Chinese studies, relapse seemed to ences in the pres c r iption schema. Hence inhibitors, are more selective inhibitors occur after discontinuation of the drug, in Beijing, thalidomide was started at of TNF . It is expected that some of meaning that long-term trea tment wou l d 50 mg/day and doubled every ten days these agents with increased safety pro- pre s u m a bly be necessary to maint a i n to a steady dose of 200 mg/day, which file, as compared to thalidomide, will efficacy in SpA patients responding to was well tolerated (none of the 30 pa- rep resent useful anti-TNF d rugs in thalidomide. tients dropped out because of side- the future (8).

S-160 Thalidomide in ankylosing spondylitis / F. Huang et al.

References 6. CA L A B R E S E L , FLEISCHER A B : Th a l i d o- 12. HUANG F, GU J , BRAUN J et al.: Identifying 1. ANDERSON JJ,BARON G, VAN DER HEIJDE D, mide: Current and potential clinical applica- the targets of thalidomide and anti-TNF anti- FELSON DT, D O U G A D O S M: A n ky l o s i n g tions. Am J Med 2000; 108: 487-95. b o dy tre atments in ankylosing spondy l i t i s s p o n dylitis assessment group pre l i m i n a ry 7. CH E N T, VOGELSANG G, PETTY B et al. : (AS). Arthritis Rheum 2000; 43: S396. definition of short-term improvement in an- Plasma pharmacokinetics and urinary excre- 13. HUANG F, GU J , ZHAOW et al.: A one year kylosing spondylitis. Arthritis Rheum 2001; tion of thalidomide after oral dosing in heal- open label trial of thalidomide in ankylosing 44: 1876-86. thy male volunteers. Drug Metab Dispos Biol spondylitis. Arthritis Rheum (Arthritis Care 2. BRANDT J, HAIBEL H, C O R N E LYD et al. : 1989; 17: 402-5. and Research) 2002; 47: 249-54. Successful tre atment of active anky l o s i n g 8. CORRAL LG, HASLETT PA, MULLER GW et 14. KEIFER JA,GUTTRIDGE DC,ASHBURNER BP, spondylitis with the anti-tumor necrosis fac- al.: Differential cytokine modulation and T BALDWIN AS,JR: Inhibition of NF- activity tor monoclonal antibody infliximab. Arthritis cell activation by two distinct classes of tha- by thalidomide through suppression of kinase Rheum 2000; 43: 1346-52. lidomide analogues that are potent inhibitors activity. J Biol Chem 2001; 276: 22382-7. 3. BRAUN J, BOLLOW M, NEURE L et al.: (Use of TNF-alpha. J Immunol 1999; 163: 380-6. 15. LEE L, LAWFORD R, MCNEIL P:The efficacy of immunohistologic and in situ hybridization 9. D ’ AM ATO RJ, LOUGHMAN MS, F LYNN E, of thalidomide in sever e seron e gati ve spondyl o - techniques in the examination of sacroiliac FOLKMAN J:Thalidomide is an inhibitor of art h ro p at h y. Art h r itis Rheum 2 0 0 1 ;4 4 :2 4 5 6 - 7 . joint biopsy specimens from patients with an- angiogenesis. Proc Nat Acad Sci USA 1994; 16. MARRIOTT JB, MULLER G,DALGLEISH AG: kylosing spondylitis. Arthritis Rheum 1995; 91: 4082-5. Thalidomide as an emerging immun o t h e r ape u - 38: 499-505. 10. El HASSANI S, DOUGADOS M, GOMBERTB tic agent. Immunol Today 1999; 20: 538-40. 4. BREBAN M, DOUGADOS M:The efficacy of et al.: Treatment of severe refractory spondy- 17. MARZO-ORTEGA H, MCGONAGLE D, O’CON- thalidomide in severe refractory seronegative larthropathy (SPA) with thalidomide: Results NOR P, EMERY P: Efficacy of etanercept in the spondyloarthropathy: Comment on the letter of an open study. Arthritis Rheum 1999; 42: treatment of the entheseal pathology in resis- by Breban et al. Reply. Arthritis Rheum 2001; S373. tant spondy l a rt h ro p at hy. A rt h ritis Rheum 44: 2457-8. 11. GI N S BURG PM, DASSOPOULOS T, E H R E N- 2001; 44: 2112-7. 5. BREBAN M, GOMBERT B, AMOR B, DOUGA- PREIS ED:Thalidomide treatment for refrac- 18. WOLKENSTEIN P, LATARJET J,ROUJEAU JC DOS M: Efficacy of thalidomide in the treat- tory Crohn’s disease:A review of the history, et al. : Randomised comparison of thalido- ment of re f ra c t o ry ankylosing spondy l i t i s . pharmacological mechanism and clinical lit- mide versus placebo in toxic epidermal ne- Arthritis Rheum 1999; 42: 580-581. terature. Annals Med 2001; 33: 516-25. crolysis. Lancet 1998; 352: 1586-9.

S-161