INTERNATIONAL JOURNAL OF LEPROSY

INTERNATIONALJOURNAL OF LEPROSY

And Other Mycobacterial Diseases

Volume 5_ Number I. \Printed in the LS.A. -

(ISSN 0148-916)6,\\

VOLUME 52, NUMBER 1^ MARCH 1 984

Hepatotoxicity of the Combination ofRifampin-ethionamide in the Treatment of

Multibacillary Leprosy'Stefaan R. Pattyn, Luk Janssens, Jacques Bourland, Turkun Saylan,

Elisabeth M. Davies, Saverio Grillone, Claude Feracci, and theCollaborative Study Group for the Treatment of Leprosy 2

For treatment of multibacillary leprosycombined chemotherapy must be admin-istered to prevent the selection of drug re-sistant mutants and possibly to improve re-sults in terms of relapse rates. It has beenour opinion ( 5 ') that a definite cure of mul-tibacillary leprosy can only be obtained bythe combined administration of powerfulbactericidal drugs against Mycobacterium

' Received for publication on 20 June 1983; acceptedfor publication on 15 September 1983.

2 S. R. Pattyn, M.D., Professor of Microbiology, Lep-rosy Laboratory, Department of Microbiology, Insti-tute for Tropical Medicine, Antwerp, Belgium. L. Jans-sens, M.D., Medical Officer, Museniene Hospital, Zaire.J. Bourland, M.D., Head, National Leprosy Service,Bujumbura, Burundi. T. Saylan, M.D., Professor ofDermatology, Head, Leprosy Service, Istanbul Uni-versity, Istanbul, Turkey. E. M. Davies, M.D., QuaIboe Church Leprosy Hospital, P.O. Box 46, Etinam,Cross River State, Nigeria. S. Grillone, Anthropol. D.,Head, National Leprosy Service, Comore Islands. C.Ferraci, M.D., former Medical Officer, Institute Mar-choux, Bamako, Mali. Collaborative Study Group:Zaire = N. Bossaer, L. Breugelmans, L. Couvreur, H.Cuyckens, J. Deverchin, G. Groenen, E. Nollet, R.Petit, A.-M. Passagez, B. Rosano, J. Van Boxstaele, J.Verlinden and R. Wattelet; Burundi = L. Desmet;Rwanda = M. Boen, E. Schacht and M. Valet; Ant-werpen = G. Hooft, A. Van Aerde and G. Van Loo.

lepaze, of which only two are presentlyknown—rifampin and ethionamide or pro-thionamide 6 ).

During prospective trials of combinedregimens including the combination of ri-fampin with ethionamide or prothionam-ide, we unexpectedly observed cases of hep-atitis which can only be interpreted as theresult of drug toxicity. These are discussedin the present report.

PATIENTS AND DRUG REGIMENSEthionamide and prothionamide are con-

sidered equivalent as far as antibacterial ac-tivity is concerned. We use here only theterm prothionamide (PRO), although insome areas ethionamide (ETH) has beenused.

Previously treated and untreated multi-bacillary patients [Bacterial Index (BI) >2at any of at least three skin sites] were in-cluded in the trials. After appropriate clin-ical, neurological, bacteriological and gen-erally histopathological evaluation, patientswere hospitalized and treated with a com-bination of rifampin (RMP), prothionam-ide (PRO), and either dapsone (DDS) or

International Journal qf Leprosy^ 1984

clofazimine (CLO). Previously untreatedpatients and patients treated previously fora maximum of five years with dapsone(monotherapy) received a combination ofRMP+ PRO+ DDS. Patients previouslytreated for more than five years with DDS,because they were at risk of having second-ary DDS resistance, were given CLO insteadof DDS. In Zaire and Rwanda, all femalepatients were treated with CLO-containingregimens. Treatment regimens varied in (lir-ferent countries but, during the introductoryphase, they always included the triple com-bination.

Since hepatitis was only very rarely ob-served after more than six months of treat-ment, only the first six months of the treat-ment regimens are given here.

Zaire-RwandaRegimen A: (26w RMP 600 7/7, PRO

500 7/7, DDS 100 7/7). During 26 weeks:RMP 600 mg, PRO 500 nig and DDS 100mg, daily.

Regimen B: same as A but DDS replacedby CLO.

Burundi-Comore IslandsRegimen C: (8w RMP 600 7/7, PRO 500

7/7, DDS 100 7/7 + 18w R 600 1/7, PRO500 7/7, DDS 100 7/7). Eight weeks: RMP600 mg, PRO 500 mg and DDS 100 mgdaily, followed during 18 weeks by RMP600 mg once a week, with daily PRO 500mg and DDS 100 mg.

Regimen D: (8w RMP 600 7/7, PRO 5007/7, CLO 100 7/7 + 18w RMP 600 1/7,PRO 500 7/7, CLO 100 7/7). Same as Cbut DDS replaced by CLO.

Since in Burundi a prevalence survey ofDDS resistance had been performed ('), pa-tients who had been treated for more thanfive years with DDS, but were shown notto be DDS resistant, were also given Regi-men C.

Turkey-NigeriaRegimen E: (2w RMP 600 7/7, PRO 500

7/7, DDS 100 7/7 + 24w RMP 600 1/7,PRO 500 7/7, DDS 100 7/7). Two weeks:RMP 600 mg, PRO 500 and DDS 100 mgdaily, followed by 24 weeks: RMP 600 mgonce weekly, PRO 500 mg and DDS 100mg daily.

Regimen F: identical to Regimen E but

DDS replaced by CLO first two wci.ks 100mg daily, next 24 weeks 300 mg onc; week-ly.

BamakoRegimen G: (90d RMP 600 2/7, PRO 500

7/7, DDS 100 7/7). Ninety days: RMP 600mg twice a week, PRO 500 mg daily andDDS 100 mg daily.

Regimen H: same as Regimen G but DDSreplaced by CLO.

RESULTSDiagnosis of hepatitis was almost always

clinical since most observations were donein rural hospitals. A few centers had labo-ratory facilities and based their diagnoseson increased ALAT (alanine-aminotrans-ferase) tests and bilirubin levels. When ic-terus or hepatitis were diagnosed, antilep-rosy treatment was stopped. In some casestreatment was restarted after the hepatitishad subsided. In most cases, however, treat-ment was continued with dapsone or clo-fazimine in monotherapy.

There were 105 patients in Regimen A(with four cases of hepatitis, 4%) and 184patients in Regimen B (with 11 cases of hep-atitis, 5.9%).

Since these differences are not significant(Fisher's exact test p = 0.11), patients onRegimens A and B are discussed as a singlegroup—Regimen A/B. The overall inci-dence of hepatitis in Regimen A/B was 5%.Table 1 shows the age distribution of thepatients, the duration of previous dapsonemonotherapy at the start of combined treat-ment, and the incidence of hepatitis. Thelatter increased with age—from less than 5 Tounder age 49, it reached 11.6% in patientsaged 50 years or older. The difference washighly significant. There was also an in-crease in hepatitis in those patients treatedpreviously for longer periods of time. butthe difference between the groups treated inthe past for 0-9 years and more than 10years was not significant.

The mean incubation time for hepatitiswas 88 days; median = 76 days (Table 2).Four patients in this group died (27%). Theirmean age was 48 years, not significantly dif-ferent from the mean age (45.6 years) ofthose who developed hepatitis and re-covered, but ten years higher than the meanage of all patients treated with Regimen A/

52, 1^Pau yn, et al.: Hepatotoxicity of Rifampin-ethionamide^3

TABLE 1. Age of patients, incidence of hepatitis . , and duration of previous treatment inpatients on Regimens ..-1/13, C/D, E/F and G/II.

Regimen A/13 Regimen C/D Regimen E/F RegimenG/HAge No. withNo. hepatitis (%) No. No. with

hepatitis ((lb) No. No. withhepatitis (%) No.

<10^4 II10-19^14 37 3 (8) 620-29^48^2 (4.1) 55 3 (5.4) 15 1730-39^84^3 (3.5) 49 1^(2)" II 4 2640-49^79^3 (4) 31 3 4>50^60^7(11.6) 43 I^(2.3) 5 5

Adults 41Total^289^15 (5) 226 8 (3.5) 81 5 (6.2) 53

Duration previous treatment (yr)118 2 (1.6) 37 2 30^119^3 (2.5)

1-4^28^2(11) 65 5 (7.6) 16 415-9^33^1 (3) 12 2 6

10-14^26^2 (7.6) 8 1 315-20^40^2 (5) 17 3>20^43^4 (9.3) 6 1

Unknownh 20

" Patient treated with Regimen D.b Unknown, but for many years.

13. In two patients treatment was restarted;hepatitis reappeared within a month in onepatient, while in the second treatment hadto be stopped because of gastrointestinalcomplications.

There were 205 patients in Regimen Cwith seven cases of hepatitis (3.4%) and 21patients in Regimen D with one case of hep-atitis (4.7%). Since these incidences werenot significantly different (p = 0.36), pa-tients in Regimens C and D are discussedas one group (Regimen C/D). The 3.5%overall incidence of hepatitis in this groupwas not significantly different from the in-cidence in Regimen A/13 (p = 0.13). An in-crease of hepatitis with increasing age wasnot observed, although this may have beendue to the smaller number of older patientsincluded in this regimen. The mean age forgroup A/I3 was 38.8 years compared with27.6 years for group C/D; the 50'11) percen-tile for Regimen A/I3 was nearly 40 years,while it was nearly 30 years for Regimen C/D. Hepatitis in group A/13 appeared after amean incubation time of 108 days: medi-an = 118-152 days (Table 2). The mean ageof these patients was 24 years. not signifi-cantly different from the mean of all thosetreated with Regimen C/D. There were two

deaths. The mortality of hepatitis in Regi-men C/D was 25%.

Treatment was restarted cautiously inthree patients by the administration for twoweeks of the combination DDS+RMP;when results of alkaline phosphatase andALAT tests remained normal, ETH wasadded at a reduced dose of 375 mg withoutany recurrence of hepatitis.

There were only 81 patients in RegimenE/F with five cases of hepatitis (one wassubchnical); one patient died. There was noclear correlation with age or duration of pre-vious treatment, probably because thegroups were too small (Table 1). The meanincubation time for hepatitis in this groupwas 80 days (Table 2).

No cases of hepatitis were observed amongthe 47 patients included in Regimen G orthe six patients in Regimen H (Table 1). Themajority of these patients were young; theirmean age was 32 years and the 50% per-ce,aile was between 30 and 34 years. Thegreat majority had been previously (irreg-ularly) treated for periods not exceeding fiveyears.

Acute phase serum from seven hepatitiscases was examined for the presence of hep-atitis 13 virus markers. In no instance was

No.^Sex

^

I^M

^

2^M

^

3^M

^

4^M

^

5^F

^

6^F

^

7^M

^

8^M

^

9^F

^

10^M

^

II^F

^

12^F13

^

14^F

^

15^F161718192021

232425262728

4^

International Journal of Leprosy^ 1984

TABLE 2. Se.v, age, and mean BacterialIndex (131) at start of treatment and incu-bation time for hepatitis during dillerentreginiens.

Age BI Regi-men

Incub.time

(days)

21 3 A 5142 3 A 6444 A 7233 4 A 14250 4 850 2.5 6662 4.5 7 2 "61 4.5 7253 3 II 7656 5 77-109"65 2.5 102"61 106 29 3.5 12130 2 13 142"36 3 13 150"50 3616 3.5 C 46II 3.5 C 118"

15220 2.5 C 15917 4.5 C 16725 2 C 186"30 520 2 E 7230 3.5 F 48"34 5 F 6726 3 F 8034 5 F 134

Relapse on retreatment.b Died.

13 antigen detected but antibodies againstthe surface antigen were present in all sera.

DISCUSSIONHepatitis, mostly with jaundice, appeared

in 4.5% of a group of 596 multibacillaryleprosy patients treated for six months witha combination of RMP 600 mg, ETH 500mg, and either DDS or CLO 100 mg. In-cubation time was 5-186 days, with a meanof 93 and a median of 76 days. Mortalityfrom this hepatitis was 26%.

Hepatitis does occur rarely in leprosy pa-tients in the absence of combined therapy;in our studies, hepatitis was observed insome patients several months after therapyhad been stopped, and in others taking DDS,many months after they had taken a singledose of RMP of 25 mg/kg body weight. Theetiology of this low-incidence hepatitis,

which could be identical to that occurringin the nonleprosy population, remains un-known.

The hepatitis cases described in this paperare probably not caused by the hepatitis 13virus, since the hepatitis B virus antigen wasnot found in any of the acute phase seraexamined. Hepatitis virus surface antibod-ies were present in all sera. (This is a com-mon observation in Africans, particularlyin multibacillary leprosy patients 4 .) There-fore the hepatitis cases observed must beinterpreted as the result of drug toxicity.

In all of the regimens used, ET1.1 was giv-en daily in a 500 mg dose. RMP. 600 mg,was given either daily (Regimen A/I3) ordaily during the first eight weeks followedby a once-weekly dose (Regimen C/13), ordaily during the first two weeks only, alsofollowed by a once-weekly dose (RegimenE/F). There was no difference in the inci-dence of hepatitis for these alternatives ofRMP administration. Whether DDS or CLOwere included in the regimens did not affectthe incidence of hepatitis.

Cases of icterus were described during thetreatment of tuberculosis with RMP. It waslater discovered that these were not due toRMP per se, but that RMP increased thetoxicity of isoniazid (INH) administered si-multaneously. There were only exceptionalcases of jaundice when RMP was admin-istered in monotherapy ("). ETH has beenclaimed to induce liver damage, but the in-cidence of this is difficult to evaluate be-cause it was generally given in associationwith other drugs, particularly INH. Fromseveral series of observations, the incidencevaries from 0-5% ( 2 . ")). Rollier and Rollier( 8 ) reported one case of hepatitis among 102multibacillary leprosy patients treated withETH 500 mg daily in monotherapy. Manyreports from leprosy treatment centers men-tion that they administer the combinationRMP+DDS or RMP+CLO or evenRMP+DDS+CLO for periods of from 1-3 months, but hepatitis is never mentioned.The combinations ETH + DDS andETH +CLO are also not hepatotoxic be-cause during a second "continuation" phaseof treatment, we have administered eitherDDS or CLO or DDS+ ETH or CLO+ ETH,without RMP, to most of the above-men-tioned patients (for six months). Only twocases of hepatitis appeared among them

52, 1^Pattyn, et al.: Hepatotavicity of Rifampin-ethionamide^5

(both on the combination ETH+CLO)which might not be different from the lowincidence hepatitis in the nonleprosy pop-ulation (unpublished data).

It, therefore, seems that in our observa-tions the combination of RMP and ETH-in the dosages and frequencies used—is theetiologic factor for the hepatic toxicity. Itcould be that in analogy with what occursfor INH, RMP induces the synthesis of hostenzymes resulting in the production of toxicmetabolites of ETH.

The correlation of hepatitis with the olderage group in Regimen A/B may be the resultof - pre-existin g liver damage, either throughthe longstanding leprosy infection with con-current amyloidosis, or liver damage fromother causes.

There are four possible reasons for theabsence of hepatitis in Regimen G/H: 1) thepatients in this regimen were considerablyyounger than those in Regimen A/B, al-though they were ofcomparable age to thosein Regimen C/D; 2) since in all regimensapproximately 50% of hepatitis cases ap-pear during the first trimester, shorteningthe period of administration of the combi-nation RMP+ ETH should reduce the in-cidence of hepatitis by about one half; 3)the number of patients in Regimen G/H wassmall; and 4) it is also possible that RNIPadministered twice a week instead of daily(in combination with daily ETH) avoidshepatotoxicity.

In five cases treatment was restarted. Inone patient retreated with RMP 600 mg andETH 500 mg hepatitis recurred; in a second,treatment had to be abandoned because ofgastric intolerance. In three other cases noharm was done when the drugs were rein-troduced cautiously with a lower dose of375 mg ETH.

When this high incidence of hepatitis wasobserved, the dosage of ETH in the regi-mens was lowered to 5 mg/kg body weight.Since then about 40 patients have beentreated with this reduced ETH dosage, andno further cases of hepatitis have been ob-served.

This measure might not avoid entirely theappearance of hepatitis. For example, Stingland Stingl CI observed one case of hepatitisamong 18 patients treated with the com-bination RMP 600 mg, ETH 175 mg, DDS50 mg, and INH 175 mg daily; however,

they did not present details on the age ofthe patients or the length of their previoustreatment, although "50"/o of [the group] hadreceived DDS treatment for varying dura-tion and regularity."

If a definite cure for multibacillary lep-rosy can only be obtained by the adminis-tration of an association of bactericidaldrugs, our observations illustrate once morethe great need for new, potent bactericidaldrugs against IL leprac, since the combi-nation of the only two powerful bacteri-cidals presently known ( 3 ) gives rise to 5–10% hepatotoxicity.

Ongoing studies will have to show if areduction in the dosage of ETH or a reduc-tion in the duration of the treatment withcombination RMP+ ETH will be able to re-duce the hepatotoxicity while keeping itsantibacterial effect.

SUMMARYDuring treatment of multibacillary lep-

rosy with the combination rifampin (RMP)600 mg, ethionamide (ETH) 500 mg, andeither dapsone (DDS) or clofazimine (CLO)100 mg. hepatotoxicity was observed in4.5% of 596 patients. Hepatitis appearedafter 5-186 days, with a mean of 93 daysand a median of 76 days. Mortality was2 6%. ETH and DDS or CLO were admin-istered daily in all regimens in which hep-atitis occurred. RMP was given either dailyor daily during the first two weeks or eightweeks, followed by a once-weekly dose.It is concluded that the combinationRMP+ETH is the toxic component. In somepatient groups there was a high correlationof toxicity with age. A regimen in whichRMP was administered only twice a weekduring three months was not accompaniedby hepatotoxicity. Future studies shouldshow if reduction of the daily dose of ETHor reduction of the duration of the admin-istration of RMP+ETH might reduce theincidence of hepatotoxicity while conserv-ing the efficacy.

RESUMENSe observe que durante el tratamiento de 596 pa-

cientes con lepra multibacilar con una combinaciOn de600 mg de rilampicina (RMP). 500 mg de etionamich(ETH) y 100 mg de dapsona (DDS) o clofazimina (CLO).el 4.5°,4) de ellos mostraron hepatotoxicidad. La hepati-tis apareciO despues de 5 a 186 Was, con una media

6^ International .10111 . 11(11 (>1 . Leprosy^ 1984

de 93 dins y una mediana de 76 Bias. La mortalidadfur de 26%. En los casos en los que hobo hepatitis. IaET11 y la DDS o la CLO se administraron dianamente.La RM P se administrO diariamente, o (liana men te solodurante las primeras 2 a 8 semanas continuando des-plies con una dosis una vet cada semana. Se concluyOque la combinaciOn RMP-4-ET11 Me el componentetOxico. En algunos grupos de pacicntcs bubo una altacorrelation entre Ia toxicidacl v Ia edad. ('uando IaRMP se administr6 solamente 2 veccs por semanadurante 3 meses. no ocurriO la hepatotoxicidad. Losestudios futuros demostraran si la reducciOn de Ia dosisdiaria de ETH o Ia reducciOn en el tiempo de admi-nistraciem de RMP+ ETH pueden reducir Ia incidenciade hepatotoxicidad sin disminuir la eficacia observada.

RESUMEAu cours du traitement de Ia lepre multibacillaire

par line combinaison rifampicine (RMP) 600 nig,ethionamide (ET11) 500 mg, complete par la dapsonc(DDS) ou Ia clolatimine ((MO), 100 mg. on a observeone toxicite hepatique chez 4.5%de 596 malades. L'he-pante est apparue apres 5-186 jours, avec one moyennede 93 jours et une mediane de 76 jours. La mortalitea ete de 26%. L'ETI I et Ia DDS ou la CLO avaicnt eteadministrees quotidiennement dans tomes les poso-logics a la suite clesquelles une hepatite est survenue.,La rifampicine elan administree soit quotidicnnement,ou bien quotidiennement initialement sculcment du-rant deux semaines. ou durant butt semaines, suivieensuite par une close hebdornadaire. On en conclut quela combinaison RNIP+ ETH constitue Pi:lenienttoxiquc. Dans certains groupes de malades, on a releveune correlation elevee de Ia toxicite avec rage. Uneposologie faisant appel a ('administration de RNIP uni-quement deux fois par semaine au cours des trois pre-miers mois, n'a pas ete accompagnee de toxicitepatique. Des etudes ulterieures pourraient montrer sila reduction de Ia dose quotidienne d'ETH, ou la re-duction de la duree d'administration de la combinaisonRMP+ETH est susceptible de reduire ('incidence dela toxicite hepatique, tout en conservant son ellicacitetherapeutique.

Ackno%%ledgments. The clinical centers in Zaire-Rwanda, Burundi-Comore Islands, and the LeprosyLaboratory, Institute for Tropical Medicine, Antwerp,Belgium, were financed by the Damicn Foundation,Brussels, Belgium. The Antwerp laboratory also had a

grant from the Nationaal Fonds your Wetenschappel-ijk Ondertock (National Foundation for Scientific Re-search), Brussels, Belgium.

REFERENCESI. 110IIRLAND„1., VAN Loo, G. and PATTYN, S. R.

Dapsone resistant leprosy in Burundi. Lepr. Rev.54 (1983) 239-242.

2. C.. KTRNIAREc, J., HAGIIENAIIER, 0.,PAGE, G. and 13Ac n, J. F. Essais cliniques duthiamide de racide-alpha-prophylisonicotinique(1321 Th) dans le traitement de la tuberculosehumaine. Tolerance, toxicite viscerale comparees

celles du 1314 Th. Rev. Tuberc. Pneumol. 29(1965) 33-48.

3. CoLsToN, M. J., ELLARD, G. A. and GAMMON, P.T. Drugs for combined therapy: Experimentalstudies On the antileprosy activity of ethionamideand prothionamide and a general review. Lepr.Rev. 49 (1978) 115-126.

4. MAUPAS, P., CI incoN, J. P., GotimAti, A., COUR-SAGET, P., PERRIN, J., IJARIN, F., YVONNET, B.,Dr mns, F., Di , [To, B., DuFt.o-NloiciAti, 13., Si-Duni, S. and DIALic), A. N. Epidemiologie et con-sequences pathologiqucs du portage chroniquevirus de l'hepatite B au Mali. 131111. Soc. Pathol.Exot. Filiales 74 (1981) 722-732.

5. PATTYN, S. R. Comments on the chemotherapyof leprosy as influenced by the present knowledgeon Mycobacterium leprae. Lepr. Rev. 43 (1972)126-136.

6. PA . FTYN, S. R. Further data on the effect of ethion-amide and prothionamide in experimental lep-rosy. Lepr. Rev. 49 (1978) 199-202.

7. PATTYN, S. R. The strategy of leprosy treatment:A personal view. Ann. Soc. Belg. Med. Trop. 60(1980) 253-262.

8. ROLLER, R. and ROLLIER, M. Traitement de leprelepromateuse par Pethionamide. Maroc. Med. 52(1972) 148-166.

9. P. and SI INGi., M. Results of a combinedleprosy therapy in Sierra Leone. Castellanica 3(1977) 53-54.

10. ZANKI., I. and WricifERT, E. Tolerance hepatiquedu prothionamide. Prax. Pneumol. 26 (1972) 450-454.

I 1 . ZINIMERMAN, H. J. hrepatotavicay. The AdverseOleas of Drugs and Other Chemicals on the Liv-er. New York: Appleton-Century-Crofts. 1978.