Volume 64, Number 3Printed in the U.S.A.

(ISSN 0148-916X)

INTERNAHONAL JOURNAL Of IATROSY

Effect of rhulFN-yTreatment inMultibacillary Leprosy Patients'

Elizabeth P. Sampaio, Ana M. Malta, Euzenir N. Sarno, and Gilla Kaplan 2

It has long been known that interferongamma (IFN-'» can activate human mono-cytes in vitro and thereby stimulate killingof intracellular organisms ( 7 ). IFN-y has,therefore, been used in studies of patientswith lepromatous leprosy in whom there isa high bacillary load despite multidrug ther-apy (MDT). Our previous studies have de-scribed the responses of patients with lepro-matous leprosy to the intradermal adminis-tration of recombinant human IFN-y(rhulFN-y) (2 "' Intradermal injec-tion of rhuIFN-y for 6 days, as an adjunct toMDT, resulted in both a local reduction inthe number of acid-fast bacilli (AFB) in theskin at the site of injection and a reductionat distal sites. However, when rhulFN-ywas administered for the 6-day treatmentperiod and then monthly for 6-10 months,the cytokine was associated with the devel-opment of erythema nodosum leprosum(ENL) in 6 of 10 patients ( 1 "). Monocytesisolated from the blood of these patients se-creted elevated levels of tumor necrosis fac-tor alpha (TNF-a) in response to stimula-tion with mycobacterial components invitro. Thalidomide treatment of the patientsduring ENL resulted in a reduction in toxicsymptoms and elimination of the lesionsof the reactional state. The in vitro additionof thalidomide to cultured monocytesfrom these patients caused a suppression of

' Received for publication on 17 November 1995;accepted tOr publication in revised form on 22 Febru-ary 1996.

= E. P. Sampaio, M.D., Ph.D.; A. M. Malta, M.D.; E.N. Sarno, M.D., Leprosy Laboratory, Oswald() CruzInstitute, Fiocruz, Manguinhos, Rio de Janeiro, Brazil21045. G. Kaplan, Ph.D., Laboratory of Cellular Phys-iology and Immunology, The Rockefeller University,1230 York Avenue, New York, New York 10021,U.S.A.

Reprint requests to Dr. Kaplan at above address orFAX = 212-327-8875; entail = kaplang@ 2 rockvax.rockefeller.edu

268

TNF-a secretion. These findings suggestedthat ENL might be associated with en-hanced TNF-a production by monocytesand/or accelerated bacterial clearance.

To understand the relationship betweenrhulFN-y-induced bacillary clearance andrhulFN-y-induced ENL, multibacillary (MB)leprosy patients were treated with rhulFN-yfor 6 consecutive days and then monthly for10 months. Two groups of patients werestudied. In one group, the patients werenewly diagnosed and on MDT. These pa-tients received rhulFN-y and a single 100-mg dose of oral thalidomide at the sametime as each rhulFN-yinjection. In the sec-ond group, skin bacterial index (BO-nega-tive and BI-positive patients who had com-pleted 2 years of MDT received onlyrhulFN-y. Local cellular and systemic re-sponses, the frequency of ENL episodesand the rate of bacillary clearance wereevaluated in each group.

MATERIALS AND METHODSPatient population. Two groups of pa-

tients were evaluated at the leprosy clinic,Oswald() Cruz Institute, Fiocruz, Rio deJaneiro, Brazil. In the first group, eightnewly diagnosed patients classified, accord-ing to the Ridley Jopling classification ( 9 ),multibacillary patients on MDT (600 mg ri-fampin and 300 mg clofazimine monthly,supervised and 100 mg dapsone and 50 mgclofazimine daily, unsupervised) were in-chided in this study. Of these, 4 were classi-fied as borderline lepromatous (BL), 2 sub-polar lepromatous (LLs), I polar leproma-tous (LL), and 2 midborderline (BB) (Table1). The second group consisted of 12 pa-tients who had completed 24 months ofMDT and were under clinical surveillancein the absence of MDT. In this group, 6were classified as BL, 4 were LLs, I wasLL and I was 13B. Four of these patients(patients 9-12) had a negative skin BI (BI =

64, 3^Sampaio, et al.: IFN-yTreatment in MB Leprosy^269

TABLE 1. Newly diagnosed leprosy patients treated with MDT, multiple doses ofrhulFN- yand thalidomide.

Patient Diagnosis BI ' pre-rhulEN-y

1

BI BI' post-rhulFN-y

ENL duringrhuIFN-y

BL 4 4 — —

2 BL 4 3 + C3 BL 5 4 — —

4 LLs 4 4 — —

5 BB 4 4 — —6 LL 5 4 — +7 BL 4 4 +8 Lis 6 4 + —

'131 evaluated before rhulEN-yinjection (pre) and i 2 months after first rhulFN-ytherapy (post).h Induration observed at the site of rhulFN-yinjection 24 hr after cytokine administration.Patients developed neuritis; no ENL.

0), and in 8 patients (patients 13-20) the BIwas still positive (mean ± S.D. = 2.1 ± 0.8)(Table 2). Two patients in the treated group(patients 19 and 20) had a previous historyof reactional episodes (ENL). None of theother patients had any reactional manifesta-tion before the beginning of the study.

rhuIFN-y injectionsrhulFN-y(Boehringer-Ingelheim, Ingel-

heim am Rhein, Germany) was used forthese studies. The lyophilized cytokine (spe-cific activity 2 x 10 7 U/mg of protein) wasreconstituted in pyrogen-free sterile water.After written informed consent, patientswere injected intradermally on the backwith a single dose of 30 pg of rhulFN-ydaily for 3 days followed by 3 days of

respite and then three additional daily injec-tions of 30 pg rhulFN-y, for a total of sixinjections. One month after the last dose of30 pg, patients received a single dose of100 pg injected intradermally in the back;this continued monthly for a total of 10monthly injections of 100 pg of rhulFN-y.

Patient evaluationPatients were evaluated clinically and

bacteriologically before the injections, afterthe 1st, 6th, and 10th months of treatment,and then at 12 months after the initiation ofthe study. Induration at the site of injectionwas recorded 24 hr after each 30 pg injec-tion, and the patients were questioned forthe presence of any symptoms and/or in-duration at the injection site when they

TABLE 2. Leprosy patients following completion of MDT and subsequently treatedwith rhulFN-y.

Patient Diagnosis BP pre/postrhulFN-y

Previous ENL Induration' ENL duringrhulFN-y

9 131_ 0/0 — — —10 13L 0/0 — — —11 13L 0/0 — — —12 BB 00/0 — — —13 LLs 1/1 — — —14 LLs 1/1 — + —15 BL 2/2 -4- —

16 LL 2/2 + +17 LLs 2/2 +18 LLs 3/3 + +19 BL 33/2 + + +20 131. 3/3 + + +

131 evaluated before rImIEN-y injection (pre) and 12 months after first rhulFN-ytherapy (post).'' Induration observed at the site of rhulEN-y injection 24 hr after cytokine administration.

270^ International Journal of Leprosy^ 1996

came to receive the next monthly dose ofthe cytokine.

Thalidomide treatmentThe eight newly diagnosed patients (pa-

tients 1-8; Table I) who did not have ENLreceived a single 100-mg tablet of thalido-mide orally at the same time as each of theinjections of rhulFN-y.

Quantitation of bacterial loadThe B1 is a numerical microscopic esti-

mate of the bacterial burden in patients di-agnosed with leprosy (the number of AFBper 100x field oil immersion) (s). Slit-skinsmears for the 131 determination were takenfrom six anatomic sites (usually earlobes,elbows, a knee, and a lesion).

RESULTSNewly diagnosed MB leprosy patients

treated with adjunctive therapy withrhuIFN-y and thalidomide

Clinical response. When patients withnewly diagnosed MB leprosy were treatedwith rhuIFN-y and thalidomide as an ad-junct to MDT, there was a local erythema-tous and indurative response in only 3 ofthe 8 patients treated with the cytokine.None of the patients treated with rhulFN-yand thalidomide exhibited the mild sys-temic symptoms associated with rhulFN-ytreatment reported previously ( 10). Malaise,fatigue, low-grade fever, nausea or weightloss were not observed.

Changes in BI. There was at least aI -log reduction in the BI in 4 of 8 newly di-agnosed patients who received adjunctivetreatment with rhuIFN-y and thalidomide(Table 1). For the group as a whole, a mean0.6-log reduction in the BI was observedover the I 2-month period of treatment andevaluation. This is analogous to the reduc-tion in the BI usually observed in responseto MDT alone during 1 year of treatment.

Development of ENL. The systemicinflammatory reaction to rhuIFN-yinjectionobserved in the majority of those patientstreated with rhulFN-y in previous studies(I") was not observed when the cytokinewas administered together with 100 mg ofthalidomide. Only I patient of the 8 in thisgroup of newly diagnosed patients devel-oped ENL during the 12 months of this

study. Two patients developed neuritis inthe absence of any ENL manifestations.The neuritis was observed in 2 of the 3 pa-tients who had reported local induration atthe site of rhulFN-y injection.

Leprosy patients after MDT treated withrhuIFN-y therapy alone

Clinical response. In this group, pa-tients had completed 2 years of MDT he-fore entering the study. Following rhulFN-ytreatment according to the same protocoldescribed above, without concomitant MDT,two different types of clinical responseswere observed. Of the BI-negative patients(patients 9-12; Table 2), none reported lo-cal erythema or induration at the site of therhulFN-y injection. However, all of the BI-negative patients experienced systemicsymptoms including fever, headache, andnausea, 0 to 4 hr after the rhulFN-y injec-tions. One patient reported a transient or-chids (which could be an indication of mildENL) at the end of the study. In contrast, 6of the 8 13I-positive patients in this groupdeveloped erythema and induration within24 hr at the site of the rhulFN-y injection.During the study, all of the BI-positive pa-tients reported some mild clinical symp-toms including nausea, headache, low-grade fever, and fatigue following therhulFN-y injections.

Changes in HI. In contrast to the re-sults observed with the group of newly di-agnosed patients, 7 of 8 patients in group IIwho were 131 positive showed no changesin the BI during 12 months of treatmentwith rhulFN-y in the absence of MDT. Pa-tients in group II who were BI negative atthe inception of the study remained BI neg-ative.

Development of ENL. BI-negative pa-tients who received rhulFN-ywithout MDTfor 12 months did not develop ENL duringthe course of the study. However, 5 of 8 ofthe 13I-positive patients in this group did de-velop ENL following rhulFN-ytreatment.

DISCUSSIONTwo groups of patients with MB leprosy

were treated with rhulFN-y in this study.Group I consisted of newly diagnosed pa-tients on MDT who received rind FN-y si-multaneously withh thalidomide in conjunc-tion with continued MDT. Group ll con-

64, 3^Sampaio, et al.: IFN1/Treatment in MB Leprosy^271

A a6

ea1.5

c.2 nc

E-0 CV

00" v 0.5

o.oMDT* + +^+^-

IFN-;^-Thal^-

z

§

E0

-6 (2

1:1!MDT* +^+^+-^-IFN-y -^+^+^+^+Thal^-^-^+-^-B I^+^+^+_^+

THE FIGUR1:. Effect of treatment on rate of reduc-tion in skin bacillary load (A) and on development ofENL (B) in leprosy patients. Results are expressed asmean log reduction in BI during 12 months (A) andpercent of patients who developed ENL during 6months of treatment (B) with (+) or without (—) MDT,IF1S- y and thalidomide (Thal) in patients with (+) orwithout (—) bacilli (BI) in their skin at the start of thestudy. * = See Materials and Methods for doses andfrequency of treatment. Number of patients in eachgroup listed in Tables I and 2 and in Reference `'.

sisted of patients who had completed 2years of MDT and who now receivedrhulFN-y alone without either MDT orthalidomide.

Our studies demonstrate that IFN-y-in-duced bacillary clearance and IFN-y-in-duced ENL, which is putatively mediatedby TNF-a production, cannot be main-tained independently of each other. WhenrhulFN-y was administered to patients ingroup I, together with thalidomide whichpresumably inhibits TNF-a production, themean reduction in the bacterial load was notdifferent from the reduction usually ob-served with MDT therapy alone (The Fig.A). IFN-y combined with MDT has previ-ously been shown to induce an acceleratedclearance of bacteria ("). The present ob-servation suggests that thalidomide pre-vents the induction of enhanced bacterial

clearance by IFN-y treatment. Thalidomidedoes not interfere with bacterial clearancewhich occurs in response to MDT. Thus,the inflammatory component induced byIFN-ytreatment (presumably TNF-a) whichis associated with both ENL and the accel-erated reduction in bacterial load is inhib-ited by thalidomide.

Thalidomide has been shown to inhibitthe production of TNF-a ( 1 . "). "). It may bethat bacterial clearance depends in part onTNF-a production by monocytes activatedby either endogenous or exogenous IFN-y('). TNF-a may render infected macro-phages in the skin better capable of bacte-riocidal activity. It appears that the effect ofexogenous IFN-yoccurs only when TNF-ais induced. However, because TNF-a pro-duction causes significant toxicities (ENL)in these patients, the therapeutic use ofIFN-ymay he limited. ENL may have moredeleterious clinical consequences than aslower bacterial clearance obtainable withMDT alone.

In the group of newly diagnosed patients,thalidomide treatment in conjunction withMDT and IFN-y was associated with lowfrequency in the occurrence of ENL (TheFig. B). Previous studies have shown thatwhen MB leprosy patients are treated withrhuIFN-y and MDT, approximately 60%develop ENL during the first 6 months ofthe treatment period ( 1 "). In contrast, the ad-dition of thalidomide to this regimen re-duced the frequency of ENL episodes to12% in the same time period. This lowerfrequency of ENL is comparable to the fre-quency observed in patients treated withMDT alone. The present results provide ad-ditional evidence for a role for TNF-a inthe induction of ENL.

In the second group of patients, who hadalready completed MDT before the start ofthis study, the presence or absence of Mr-cobacterium leprae in the skin (BI) ap-peared to be associated with the frequencyof development of ENL. In BI-negative pa-tients, ENL did not occur during rhulFN-ytreatment despite the appearance of sys-temic symptoms following rhulFN-y injec-tions. These patients did not develop ery-thema or induration at the site of IFN-y in-jections. In 131-positive patients, thefrequency of ENL was much higher as wasthe frequency of local erythema and indura-

272^ International Journal of Leprosy^ 1996

tion. The local cutaneous response appearedto precede development of ENL, while themild systemic symptoms did not necessar-ily indicate development of ENL. Thus, itappears that the presence of M. leprae inthe skin is required for ENL induction.

It is of interest to note that rhulFN-ytreatment without concomitant MDT didnot result in a reduction in the bacterial loadin the skin of BI-positive patients. This find-ing suggests that IFN-y does not, by itself,accelerate bacterial clearance. rhulFN-ytreatment together with MDT and thalido-mide also does not accelerate bacterial clear-ance. Since thalidomide inhibits TNF-a pro-duction (") it appears that TNF-a is re-quired for accelerated bacterial clearance.Treatment with rhulFN-y, although it canlead to reduction in the bacillary load whenadministered together with MDT, has a sig-nificant risk of causing ENL. Therefore,further large-scale patient studies whichwould be required to expand and to confirmthe role of exogenous IFN-y in the possiblemechanism of ENL induction are not rec-ommended.

SUMMARYPrevious studies have shown that when

multibacillary leprosy patients were treatedwith recombinant human interferon gamma(rhulFN-y) for 6-10 months there was anaccelerated reduction in the number of acid-fast bacilli in the skin at the site of injectionas well as an accelerated bacillary reductionat distal sites. However, this favorable out-come of IFN-y treatment was associatedwith the development of erythema nodosumleprosum (ENL). The present study was un-dertaken to investigate whether rhulFN-y-induced bacillary clearance could be disas-sociated from the induction of ENL.

rhulFN-y was administered together withthalidomide and conventional multidrugchemotherapy to newly diagnosed leprosypatients. During treatment with this combi-nation of drugs, the mean reduction in bac-terial load was the same as the reductionobserved with chemotherapy alone. More-over, the inclusion of thalidomide in thetreatment regimen was associated with alow frequency of ENL episodes. A secondgroup of leprosy patients, who had alreadycompleted 2 years of chemotherapy, were

treated with rhulFN-y only. In those pa-tients who were skin bacilli negative, ENLdid not occur during rhulFN-y treatment.In contrast, in bacilli-positive patients thefrequency of ENL during rhulFN-y treat-ment was higher, as was the occurrence oflocal erythema and induration. However,rhulFN-y treatment without concomitantchemotherapy did not result in a reductionin the bacterial load in the skin of bacilli-positive patients. These findings, takentogether, indicate that rhulFN-y does not,by itself, accelerate bacterial clearance, butrequires concomitant chemotherapy toachieve the accelerated reduction in bacil-lary load. Thalidomide reduces the fre-quency of IFN-y-induced ENL, but alsoeliminates the IFN-y-induced bacillaryclearance.

RESUMENPreviamente se ha demostrado que los pacientes

con lepra multibacilar tratados durante 6 a 10 mesescon interferón gamma human() recombinante (rhuIFINy),muestran una reducción acelerada en el ntimero debacilos acido-resistentes en Ia piel des sitio de inyec-cion asi como una reducción bacilar acelerada en sitiosdistantes. Sin embargo, gists resultado favorable deltratamiento con IFN yestuvo asociado con el desarrollode eritema nodoso leproso (ENL). Este estudio se hizocon el tin de investigar si la eliminación de bacilos in-ducida por el rhulENy pudiera desasociarse de Ia in-ducción de ENL. Para esto, el rhulFNyse administrójunto con talidomida y con la poi quimioterapia con-vencional a pacientes con lepra recien diagnosticada.Durante el tratamiento con esta combinación de drogas,le reducción promedio en Ia carga bacteriana foe Iamisma que la reducción observada con la quimioterapiacola. Ademtis, Ia inclusión de talidomida en el regimende tratamiento estuvo asociada con una baja frecuenciade episodios ENL. Un segundo grupo de pacientes quehabian completado 2 aiios de quimioterapia fuerontratados solo con rhuI EN y. En aquellos pacientes queno tuvieron bacilos en la piel no hubieron episodios deENL durante el tratamiento con rholFNy. En contraste,en los pacientes con bacilos la frecuencia de ENL du-rante el tratamiento con rhoIENy fue alta, como tam-biCn fue alta Ia frecuencia de eritema e induración lo-cales. Sin embargo, el tratamiento con rhulFNy sinquimioterapia no cause ninguna reducción de la cargabacteriana en la piel de Iso pacientes con bacilos. Enconjunto, los resultados indican que el rholENypor sisolo no acelera la eliminación de bacilos sino que re-quiere pant esto del efecto de la quimioterapia con-cornitante. La talidomida reduce la frecuencia de ENLinducido por el 1FN y pen) tambien bloquea le elimi-nación de bacilos inducida por esta citocina.

64, 3^Sampaio, et al.: IFN-yTreatment in MB Leprosy^273

RÉSUMÉ

Des etudes anterieures ont montré que quand des'naiades lCpreux multibacillaires étaient traites avec de[interferon-y humain recombinant (IEN-y hurt pen-dant 6-10 moil, it y avait une reduction accelerée dunombre de bacilles acidos-résistants clans lit peau auniveau du site d'injection, ainsi qu'une diminution ac-célérée de Ia charge bacillaire dans des sites plus dis-tants. Cependant, cet effet favorable du traitementl'IFN-y etait associe au développement d'erythèmenoueux lépreux (ENL). L'etude présente a etC entre-prise pour rechercher si Ia diminution de Ia chargebacillaire induite par l'IFN-yhur pouvait titre dissociéede l'induction de l'ENL.

L'IFN-yhur a etc administre en mane tenips que Iathalidomide et une polychiniiotherapie convention-nelle à des 'naiades de lit lepre nouvellement diagnos-tiques. Au cours du traitement avec cette combinaisonde medicaments, la reduction moyenne de la chargebactérienne était la meme que cello observée avec lachimiotherapie settle. De plus, [inclusion de Iathalidomide dans le regime therapeutique etait asso-ciée a une foible frequence d'episodes ENL. Un sec-ond groupe de malades de la lepre, qui avait déjà ter-mine deux ans de chimiotherapie, a etC traits avec del'IFN-yhur seul. Cher les patients qui avaient des frot-tis cutanes negatifs, l'ENL nest pas apparu au coursdu traitement a l'IFN-yhur. Par contre, chez les pa-tients avec des frottis cutanes positifs, la frequence del'ENL an cours du traitement était plus élevee, ainsique la survenue d'erythème local et d'induration.Cependant, le traitement ii l'IFN-yhur sans chimio-therapie concommitante n'a pas resulté dans tine re-duction de Ia charge bactérienne dans la peau des pa-tients à frottis positifs. Ces observations indiquent que

1FN- y hur n'accélere pas par liu-meme Ia disparitiondes bactéries, ma's requiert une chimiotherapie con-commitante pour obtenir une reduction accélérée dela charge bactérienne. La thalidomide réduit Ia frC-quence de l'ENL provoque par l'IFN-y, mail supprimeegalement l'élimination des bacilles provoquée par'1FN-y.

Acknowledgment. These studies were supportedin part by NIII grant Al-33124 and by the CelgeneCorporation. We thank Dr. Victoria Freedman for helpin writing the manuscript, Marguerite Nulty for typingthe manuscript, and Judy Adams for help with prepa-ration of the figures.

REFERENCESI. Bi n ^L. E. and KAI'l.AN, G. Recombinant

cytokines for controlling mycobacterial infections.Trends Microbiol. 3 (1995) 22-26.

2. KAI'l.AN, G., MATtIUR, N. K., Jon, C. K., NATI', I.and COHN, Z. A. The effect of multiple IFN-y in-jections on the disposal of Mycobacterium leprae.INAS 86 (1989) 8073-8077.

3. KAPLAN, G., NUSRAT, A., SARNO, E. N., Jon, C. K.,MCEt,RATII, J., Poiero, J. A., NNITIAN, C. F. andCOHN, Z. A. Cellular responses to the intradermalinjection of recombinant human interferon-)' inleproniatous leprosy patients. Am. J. Pathol. 128(1987) 345-353.

4. MAKoNKAwKEYooN, S., LintsoN-PonRE, R. N. R.,MokiaRA, A. L., SCHAUF, V. and KAPLAN, 0.Thalidomide inhibits the replication of human im-munodeficiency virus type I. PNAS 90 (1993)5974-5978.

5. NATHAN, C. F., KAPLAN, G., Lnvis, W. R., NUSRAT,

A., WITMER, M. 1)., SHERWIN, S. A., Jon, C. K.,HOROWITZ, C. R., STEINNIAN, R. M. and ConN, Z.A. Local and systemic effects of low doses of re-combinant interferon-yafter intradermal injectionin patients with lepromatous leprosy. New Engl. J.Med. 315 (1986) 6-15.

6. NATHAN, C. F., SQUIRES, K., GRirm, W., LEvis,W., VARGI.ESE, M., Jon, C. K., NUSRAT, A. R.,SHERWIN, S., RAPPAPORT, S. SANDES, E., Bum:-HARDT, R. A. and KAPLAN, G. Widespread jinni-dermal accumulation of mononuclear leukocytesin leproniatous leprosy patients treated systemi-cally with recombinant interferon-y. J. Exp. Med.172 (1990) 1509-1512.

7. NATHAN, C. F. and Yostium, R. Interferon-y. In:inflammation: Basic Principles and Clinical Cor-relates. Gallin, J., Goldstein, I. and Snyderman,R., eds. New York: Raven Press, Ltd., 1988, pp.229-25 1 .

8. RtutiA, I). S. and iln.soN, 0. R. F. A logarithmicindex of bacilli in biopsies I. Method. Int. J. Lepr.35 (1967) 184-186.

9. RitiN, D. S. and Joil.ING, W. IL Classification ofleprosy according to immunity; a five-group sys-tem. Int. J. Lepr. 34 (1966) 255-273.

10. SAmmio, E. P., MoREIRA, A. L., SARNO, E. N.,MALTA, A. M. and KAI'l.AN, G. Prolonged treat-ment with recombinant interferon gamma induceserythema nodosum leprosum in leproniatous lep-rosy patients. J. Exp. Med. 175 (1992) 1729-1737.

I I. SANtrAio, E. P., SARNO, E. N., GAiiii.tx, R., CottN,Z. A. and KAPLAN, G. Thalidomide selectively in-hibits TNEa production by stimulated humanmonocytes. J. Exp. Med. 173 (1991) 699-703.

12. SAMUEL, N. M., GRANGE, J. M., SAMUEL, S., LU-CAS, S., OWII.1.1, 0. M., ADAI.I.A, S., LEIGH, I. M.and NAVARRETTA, C. A study of the effects of in-tradermal administration of recombinant gammainterferon in leproniatous leprosy patients. Lepr.Rev. 58 (1987) 389-4(X).