Thalidomide: A Novel Template for Anticancer Drugs

David Stirling

Thalidomide is an immunomodulatory agent that is approved for use in patients with cutaneous manifesta-

tions of erythema nodosum leprosum (ENL). It is cur-

rently under clinical investigation in a wide range of

malignancies and immunologic-based disorders. Struc- tural analogs of thalidomide have been synthesized in

order to explore potential molecular targets of thalid-

omide, as well as to identify new therapeutics. Mem-

bers of one class of analogs, termed immunomodula-

tory drugs (IMiD), are IO0 to 1,000 times more potent than thalidomide in regulating cytokine production by

peripheral blood mononuclear cells and providing a

costimulatory signal for T-cell proliferation. Notably,

the nature of the stimulus and the cell type are impor-

tant determinants as to whether the IMiDs or thalido- mide cause an inhibitory or stimulator-y effect. The

IMiD CDC-501 has been selected for clinical develop-

ment; it had been safely administered to volunteers in

single doses of 50 to 400 mg and in multiple doses of

100 mg for 7 days. Phase I/II studies of CDC-501 in multiple myeloma are in progress.

Semin Oncol 28:602-606. Copyright 0 2001 by W.6.

Saunders Company.

T HALIDOMIDE WAS originally developed as a sedative/hypnotic agent, but was later

withdrawn from the market due to teratogenic and neurotoxic effects.1 Interest in thalidomide was renewed when it was recognized to effectively sup- press the clinical manifestations of erythema no- dosum leprosum (ENL), an inflammatory condi- tion associated with leprosy.2 These findings led to the approval of thalidomide (Thalomid, Celgene Corp, Warren, NJ) by the Food and Drug Admine istration for acute and maintenance treatment of the cutaneous manifestations of moderate to se- vere ENL.3 On the basis of its immunomodulatory and antiangiogenic properties, thalidomide is currently under clinical evaluation in various ma- lignancies, including multiple myeloma, myelo- dysplastic syndromes, renal cell carcinoma, glio- blastoma multiforme, and other solid tumors; immunological and autoimmune disorders, rang- ing from human immunodeficiency virus (HIV)

From Celgene Carp, Warren, NJ. Address reprint requests to David Stirling, PhD, Celgene COQO-

ration, 7 Powder Horn Dr, Warren, NJ 07059. Copyright 0 2001 by W.B. Saunders Company 0093-7754/01/2806-0002$35.00/O doi:10.1053/sonc.2001.28595

602

infection to rheumatoid arthritis; and complica- tions associated with these disorders, such as ca- chexia (Table 1).4-T In multiple myeloma, for ex- ample, approximately 25% to 35% of patients with refractory disease responded to thalidomide in clinical trials.@ These findings are especially en- couraging, because many such patients had previ- ously failed to respond to other active regimens, including bone marrow transplantation. Presently, more than 90% of current thalidomide usage is in oncology, particularly in patients with multiple myeloma.

The exact molecular mechanism of action of thalidomide in these clinical disorders remains un- known. On the basis of in vitro studies, thalido- mide has been shown to produce a number of desirable immunomodulatory effects, including tu- mor necrosis factor-alpha (TNFcx) inhibition, an- tiangiogenic activity, cell surface alterations, and immune system stimulation. However, thalido- mide also produces sedation and peripheral neu- ropathy that can negatively impact patient com- pliance, and as stated previously, it is recognized for its teratogenic potential. Many of these actions may be interrelated, and consequently, efforts to dissect how the drug is working have been prob- lematic. Moreover, thalidomide is a rather small and simple molecule, which can be hydrolyzed and degraded to metabolites that could contribute to or account for some of the observed biological actions. Accordingly, a large research effort has been undertaken to rationalize clinical observa- tions made with thalidomide to the drug’s actions on a biological level, as well as at a molecular target level. It is expected that these studies will also provide insight for identifying viable molecu- lar targets for discovery of new therapeutic agents.

IN VITRO EFFECTS OF THALIDOMIDE AND STRUCTURAL ANALOGS

Thalidomide inhibits lipopolysaccharide (LPS)- induced production of TNFa by peripheral blood mononuclear cells (PBMCs) in a concentration- dependent manner over the range of 0.2 to 500 pmol/L. However, the 50% inhibitory concentra- tion (I&) for TNFa inhibition is approximately 200 pmol/L, which is one order of magnitude higher than serum drug levels achieved in humans.

Seminars in Oncology, ‘401 28, No 6 (December), 2001: pp 602-606

THALIDOMIDE, A NOVEL DRUG TEMPLATE 603

Table I. Disease Applications for Thalidomide Under Clinical Investigation

Discipline Disorder

OtlCOlOgy

HIV

Infectious disease

Gartroenterology

Ophthalmology

Cardiology

Neurology

Dermatology

Rheumatology

Multiple myeloma; myelodysplastic

syndromes; renal cell; glioblastoma

multiforme; colorectal, prostate, liver,

head/neck, and non-small cell lung

cancer; melanoma: leukemias; cachexia

Aphthous ulcers; cachexia; protozoa1

diarrhea; Kaposi’s sarcoma

Tubercular meningitis; erythema nodosum

leprosumlleprosy

Crohn’s disease; scleroderma; ulcerative

colitis

Macular degeneration; diabetic retinopathy

Congestive heart failure

Dementia: adrenoleukodystrophy; multiple

sclerosis

Behget’s disease; discoid lupus

Rheumatoid arthritis; sarcoidosis;

ankylosing spondylitis

At first glance, the discrepancy between thalido- mide’s potency in vitro and its serum drug levels in vivo suggests that TNFol inhibition may not be a clinically meaningful mechanism of action. How- ever, it is important to recognize that the experi- mental conditions of the PBMC assay may not adequately reflect the optimal stimulus or cell types needed for thalidomide to express its maxi- mal inhibitory activity. Moreover, the microenvi- ronment in which thalidomide exerts its actions may differ under in vitro as compared with in vivo conditions. For example, in an experimental adju- vant-induced arthritis model in rats, thalidomide was essentially inactive, whereas structural analogs that inhibit TNFa and phosphodiesterase (PDE) 4, such as CC-1069, were highly active.9 Finally, the effects of thalidomide appear to be species- dependent, inasmuch as it did not inhibit TNFol production by PBMC from nonhuman primates. Accordingly, the effect of thalidomide in human PBMCs cannot be extrapolated to various disease models in animals.

Two groups of thalidomide-based structural an- alogs have been identified, which inhibit TNFa production by PBMCs. lo The first group, termed selective cytokine inhibitory drugs (SelCID), inhibit PDE 4 to elevate cyclic adenosine mono- phosphate (CAMP) levels.llJz In contrast, the other

group, termed immunomodulatory drugs (IMiD), do not inhibit the PDE 4 enzyme.13 Notably, the IMiD class of agents appears to possess a similar, if not the same, molecular target as the parent molecule, thalidomide. These IMiDs were tested initially in PBMCs for their ability to inhibit LPS-induced TNFa! production. Notably, some members of the IMiD class were greater than 1,000 times more potent than thalidomide in blocking TNFar pro- duction. For example, IMiDs designated as CI-A, CI-B, and CI-C had IC,, values of 0.01 to 0.1 pmol/L.rO These agents tended to produce a rela- tively shallow concentration-response relation ship, which is also typically seen with thalidomide. The higher potency of these agents suggests that they may be helpful as research tools for delineat- ing the molecular mechanism of action of thalid- omide.

Although initial studies of thalidomide on human PBMCs suggested that it was acting as a selective TNFa inhibitor,i+ recent evidence with the IMiDs as well as thalidomide itself shows that production of other cytokines is also regulated.rQl5J6 In one study, the IMiDs were evaluated in LPS-stimulated PBMCs at three dif- ferent concentrations, ie, their approximate IC,, for TNFa! inhibition and at three times and 10 times higher concentrations. At their ICs, for TNFa inhibition, the IMiDs reduced interleukin (IL)-lp and IL-12 production by slightly more than 50%, whereas at higher drug concentrations, production of these cytokines was inhibited more completely. IL-8 production was not affected by the IMiDs, whereas IL-6 production was reduced only at the higher drug concentrations. In addi- tion, the IMiDs stimulated IL-10 production by threefold at their respective TNFa IC,, concen- trations and by fourfold at a lo-fold higher con- centration.1° Similar results were obtained with thalidomide, but very high drug concentrations were required. The inhibitory effect of IMiDs and thalidomide on TNFa was not due to the stimu- lation of IL-10 production, because it was still evident in the presence of anti-IL-10 antibodies.

Thalidomide provides a costimulatory signal for T-cell proliferation induced by activation of the T-cell receptor by anti-CD3 (Fig l).iy In pu- rified T cells, the IMiDs were also potent costimu- lators of T-cell proliferation; they increased cell proliferation in a concentration-dependent man- ner, with maximal responses achieved at concen-

604 DAVID STIRLING

trations of 0.1 to 1 Fmol/L. In the absence of antiCD3, thalidomide and the IMiDs did not stimulate T-cell proliferation, indicating that these agents are not mitogenic.ic The costimula- tory effect of the IMiDs predominantly reflected proliferation of the CD8+ subset, with a much smaller effect on the CD4+ subset. The CD4+/ CD8+ ratio declined from 2.7 in the absence of drug to less than 1 at maximal costimulatory con- centrations of IMiD. Therefore, in the presence of an IMiD, the CD8+ subset was the predominant T-cell.

The IMiDs provided a costimulatory signal for T-cell proliferation by increasing the production of the T-cell cytokines, interferon-gamma (IFNy) and IL-2. IFNy and IL-2 production were mea- sured in purified T cells and PBMCs, respectively, following stimulation with immobilized antiCD3. The IMiD CI-A increased IFNy and IL-2 at con- centrations of 0.01 to 0.1 pmol/L, whereas CI-B was approximately lo-times less potent. At very high concentrations, thalidomide was also able to increase IFNy and IL-2 production, but it was lOO- to l,OOO-fold less potent than the IMiDs. More- over, the maximum amount of IFNr and IL-2 was generally higher in the presence of an IMiD than with thalidomide.iO

DIFFERENTIAL EFFECTS OF THALIDOMIDE AND IMiDS

The application of thalidomide to cancer treat- ment is based on its antiangiogenic actions in vitro

Fig I. Differential costimula- tion of T-cells by IMiDs. Purified T ceils were cultured and stimu- lated with immobilized anti-CD3 in the absence or presence of the indicated concentrations of an

IMiD. C3H]-thymidine incorpora- tion was measured during the last I8 hours of l20-hour cul- tures. Cells were analyzed with a

cell sorter to determine incorpo- ration into the CD8+ and CD4+ subsets. (A) Effect of IMiD- on proliferation of the T-cell sub- sets. (8) CD41CD8 ratio of un- sorted T cells according to the concentration of IMiD- and IMiD-2.

and in animal models; however, there is still lime ited clinical data to support this hypothesis. From a safety vantage, several thousand cancer patients have received thalidomide, but normal wound healing has not been adversely affected. Moreover, thalidomide has been very effective in promoting healing of aphthous ulceration in patients with HIV or Behqet’s disease, as well as closing fistulas in patients with Crohn’s disease. Therefore, tha- lidomide appears to have antiangiogenic actions in certain clinical situations, but seems to promote wound healing in other conditions. Accordingly, there remains a need to rationalize these clinical observations with the drug’s molecular targets. Re- cent evidence suggests that the nature of the stim- ulus may determine whether thalidomide and the IMiDs produce an inhibitory or stimulatory effect.

The effect of the IMiDs and thalidomide on PBMCs depended on whether a T-cell-dependent or -independent stimulus was provided. When LPS, an inflammatory and Tcell-independent stimulus, was used, production of TNFa, granulo- cyte-macrophage colony-stimulating factor (GM- CSF), and IL-12 was inhibited, and IL-10 produc- tion was increased by the IMiDs and thalidomide (Fig 2). Conversely, in the presence of the T-cell- dependent stimulus anti-CD3, the IMiDs and tha- lidomide stimulated GM-CSF and IL-12 produc- tion without significantly affecting TNFor or IL- lO.iO As in the other studies, the IMiDs were much more potent than thalidomide. These observations were confirmed in purified populations of mono-

THALIDOMIDE, A NOVEL DRUG TEMPLATE 605

Fig 2. Differential modifica- tion of GM-CSF production by IMIDs. (A) GM-CSF production

by PBMCs was measured follow- ing stimulation with either LPS or anti-CD3 in the absence or presence of the indicated con- centration of IMiD. (8) GM-CSF was measured following stimula- tion of purified monocytes with

LPS or purified T cells with anti- CD3 in the absence or presence

of IMiD. GM-CSF production in the presence of IMiD is expressed as a percentage of that measured in its absence.

cytes and T cells. Thus, the IMiDs inhibited

A B PBMC Monocytes x T ceils

250- -+-cm3

-cl-LPS 200 .

g

$j 150.

v

5

loo i

50 :

0 0 0.001 0.01 0.1 1 0 0.001 0.01 0.1 1

mm (l-w mm (IJW

GM-CSF and IL-12 production by LPS-stimu- lated monocytes, but they stimulated GM-CSF and IL-12 production by anti-CD3-stimulated T cells. These results suggest that the nature of the stimulus and the types of cells that are present are important factors in determining which actions of the IMiDs and thalidomide are expressed.

CDC-501: AN IMiD IN CLINICAL DEVELOPMENT

As evidenced by these in vitro studies, the IMiDs represent useful tools for identifying poten- tial molecular targets of thalidomide. By gaining this understanding, it may then be possible to rationalize some of the biological activities that have been observed and hopefully explain why thalidomide is effective in the clinic. Based on what we know today, duration of treatment is important, as disease modification may take sev- eral weeks to develop. However, the major goal of this research is to identify new therapeutics that will be effective in patients with multiple my- eloma. One IMiD, CDC-501, was selected for clin- ical development based on its activity in vitro as well as its pharmacokinetic profile. The bioavail- ability of this agent was greater than 50% in three species: rats, dogs, and cynomolgus monkeys. In animals, it did not cause any cardiovascular or pulmonary toxicity, and it was devoid of muta- genic potential. Like thalidomide, CDC-501 was not very effective in animal models, and conse-

quently pharmacokinetic/pharmacodynamic (PK/ I’D) relationships in animals may not be predictive of the clinical situation. If CDC-501 behaves like thalidomide, and assuming that the parent drug is the active moiety, then it may be difficult to cor- relate serum drug levels and duration of biological activity in humans. It is well recognized that lep- rosy may still be controlled biologically by thalid- omide after the patient has discontinued therapy for some time. However, if CDC-501 is not the active moiety, then it must be identified before any PK/PD correlations can be attempted.

The safety of CDC-501 has been evaluated in blinded, placebo-controlled, single- and multiple- dose phase I studies in healthy volunteers. Single doses of 50 to 400 mg and multiple doses of 100 mg for 7 days have been successfully administered. In the multiple dose study of 100 mg, the only note- worthy adverse events were dermatologic in na- ture, including pruritis and dermatitis, leading two subjects to discontinue treatment. Based on these phase I studies, two pilot phase I/II studies have been designed in multiple myeloma patients. These studies are currently ongoing at the Dana Farber Cancer Center and at The University of Arkansas. Evidence of stable disease would be sig nificant in this refractory patient population. In addition, two phase I studies will be initiated at the National Cancer Institute to further explore this agent’s activity at the biological level as well as at the molecular target level.

606 DAVID STIRLING

REFERENCES 1. Mellin GW, Katzenstein M: The saga of thalidomide

(concluded): Neuropathy to embryopathy, with case reports of congenital anomalies. N Engl J Med 267:1238-1244, 1962

2. Sampaio El?, Kaplan G, Miranda A, et al: The influence of thalidomide on the clinical and immunologic manifestations of erythema nodosum leprosum. J Infect Dis 168:408-414, 1993

3. Clark TE, Edom N, Larson J, et al: Thalomid (thalido- mide) capsules: A review of the first 18 months of spontaneous postmarketing adverse event surveillance, including off-label prescribing. Drug Safety 24:87-117, 2001

4. Calabrese L, Fleischer AB: Thalidomide: Current and potential clinical applications. Am J Med 108487-495, 2000

5. Peuckmann V, Fisch M, Bruera E: Potential novel uses of thalidomide: Focus on palliative care. Drugs 60:273-292, 2000

6. Rajkumar SV: Thalidomide in multiple myeloma. Oncol- ogy 14:11-16, 2000 (suppl 13)

7. Thomas DA, Kantarjian HM: Current role of thalido- mide in cancer treatment. Curr Opin Oncol 12564-573, 2000

8. Singhal S, Mehta J, Desikan R, et al: Antitumor activity of thalidomide in refractory multiple myeloma. N Engl J Med 341:1565.1571, 1999

9. Oliver SJ, Freeman SL, Corral LG, et al: Thalidomide analogue CC1069 inhibits development of rat adjuvant arthri- tis. Clin Exp Immunol 118:315-321, 1999

10. Corral LG, Haslett PAJ, Muller GW, et al: Differential cytokine modulation and T cell activation by two distinct

classes of thalidomide analogues that are potent inhibitors of TNF-a. J Immunol 163:380-386, 1999

11. Muller GW, Shire MG, Wong LM, et al: Thalidomide analogs and PDE4 inhibition. Bioorg Med Chem Lett 8:2669- 2674, 1998

12. Marriott JB, Westby M, Cookson S, et al: CC-3052: A water-soluble analog of thalidomide and potent inhibitor of activation-induced TNF-(U production. J Immunol 161:4236- 4243, 1998

13. Muller GW, Chen R, Huang S-Y, et al: Amino-substi- tuted thalidomide analogs: potent inhibitors of TNF-(Y produc- tion. Bioorg Med Chem Lett 9:1625-1630, 1999

14. Sampaio EP, Sarno EN, Galilly R, et al: Thalidomide selectively inhibits tumor necrosis factor 01 production by stim- ulated human monocytes. J Exp Med 173:699-703, 1991

15. Moller DR, Wysocka M, Greenlee BM, et al: Inhibition of IL-12 production by thalidomide. J Immunol159:5157-5161, 1997

16. Rowland TL, McHugh SM, Deighton I, et al: Differen- I . tial regulation by thalidomide and dexamethasone of cytokine expression in human peripheral blood mononuclear cells. Im- munopharmacology 4O:l l-20, 1998

17. Haslett PAJ, Corral LG, Albert M, et al: Thalidomide costimulates primary human T lymphocytes, preferentially inducing proliferation, cytokine production, and cytotoxic responses in the CD8+ subset. J Exp Med 187:1885-1892, 1998