1074 The Annals of Pharmacotherapy 2004 June, Volume 38 www.theannals.com

Tuberculosis (TB) is a great health problem worldwide,and the most effective control method is to cure the in-fection with antitubercular (anti-TB) drugs.1 Essential anti-TB drugs are isoniazid, rifampin, pyrazinamide, ethambutol,and streptomycin. Despite being effective chemotherapeuticagents, hepatotoxicity caused by some of these first-linedrugs is common, which may limit their use and lead to in-terruption of therapy. Isoniazid and pyrazinamide are themajor hepatotoxins, while rifampin, which is a powerful en-zyme inducer, may enhance the hepatotoxicity of isoniazid.2,3Ethambutol and streptomycin do not have hepatotoxic poten-

tial. The clinical, biochemical, and histopathologic featuresof drug-induced hepatitis are indistinguishable from thoseof viral hepatitis.4 Hepatotoxicity can range from asymp-tomatic elevation of serum transferases to hepatic failurerequiring liver transplantation.3

In the UK and US, 3 4% of patients taking anti-TBdrugs develop adverse hepatic reactions.5 These reactionsalso occur in Nepal, but the incidence is not known. InNepal, more than half of the adult population is infectedwith TB, and most of the patients are from poor socioeco-nomic areas.6 Apart from the disease worsening their con-dition, drug-induced hepatitis results in further complica-tions. The Centers for Disease Control and Prevention andthe American Thoracic Society have recommended edu-

Incidence of Hepatotoxicity Due to Antitubercular Medicines andAssessment of Risk Factors

Rajani Shakya, B Subba Rao, and Bhawana Shrestha

INTERNATIONAL REPORTS

BACKGROUND: Antitubercular drugs cause derangement of hepatic function revealed by clinical examination and abnormal liverfunction test results. Potential hepatotoxicity of some of the first-line antitubercular agents remains a problem, especially during theinitial period of treatment. OBJECTIVE: To determine the incidence of antitubercular druginduced hepatotoxicity in a Nepalese urban population and assessthe risk factors.METHOD: Fifty patients diagnosed with active tuberculosis infection with normal pretreatment liver function were monitored clinicallyas well as biochemically in a prospective cohort analysis. RESULTS: Antitubercular drugs were found to be associated with derangement of hepatic function, resulting in elevation of liver enzymesto a variable extent (t = 4.550, p < 0.01 for aspartate aminotransferase [AST]; t = 5.467, p < 0.01 for alanine aminotransferase [ALT] at95% CI). Thirty-eight percent of patients had 2 times and 30% had >3 times elevation of ALT. Similarly, 40% and 29% of patientsshowed 2 and >3 times elevation of the AST level, respectively. Four patients (8%) developed drug-induced hepatotoxicity.Jaundice was the presenting symptom in all patients. The time interval for onset of hepatotoxicity after initiation of therapy was1260 days (median 28). Antitubercular druginduced hepatotoxicity was found more often in younger patients (6% vs 2%; p =0.368, OR 2.75). Female gender was also a higher risk (p = 0.219, OR 4.2). Most patients who had developed hepatitis werediagnosed per sputum-smear positive reactions. Nutritional status, assessed by body mass index and serum albumin level, was thenext predisposing factor.CONCLUSIONS: A finding of an 8% incidence of hepatotoxicity is considerably high. Risk factors of hepatotoxicity included femalegender, disease extent, and poor nutritional status. Timely detection and temporary withdrawal of the offending agent cancompletely cure antitubercular druginduced hepatotoxicity.KEY WORDS: hepatotoxicity, isoniazid, liver function tests, Nepal, pyrazinamide, rifampin, tuberculosis.Ann Pharmacother 2004;38:1074-9.Published Online, 30 Apr 2004, www.theannals.com, DOI 10.1345/aph.1D525

Author information provided at the end of the text.

by guest on October 11, 2013aop.sagepub.comDownloaded from by guest on October 11, 2013aop.sagepub.comDownloaded from by guest on October 11, 2013aop.sagepub.comDownloaded from by guest on October 11, 2013aop.sagepub.comDownloaded from by guest on October 11, 2013aop.sagepub.comDownloaded from by guest on October 11, 2013aop.sagepub.comDownloaded from

cating patients about adverse effects and monitoring (clini-cal as well as biochemical) the patient closely during theentire treatment period. Recent studies have also suggestedthat, since the advent of routine monitoring, the risk of se-vere hepatotoxicity has been substantially reduced.

Hepatotoxic effects of anti-TB drugs are well known;however, further reports on the incidence and predisposingfactors will be helpful in minimizing these reactions. Tocontribute to the available information for healthcare prac-titioners regarding this adverse effect, we have attemptedto determine the incidence of anti-TB druginduced hepa-totoxicity in urban Nepalese people.

The present study was designed to elucidate the fact thatanti-TB treatment elevates the serum levels of liver en-zymes compared with the levels shown in the pretreatmentperiod. Identification of patients at increased risk for anti-TB druginduced hepatotoxicity is important because hep-atotoxicity causes significant morbidity and mortality andmay require modification of the therapeutic regimen.Therefore, another objective of this study was to assess therisk factors for anti-TB druginduced hepatotoxicity, thatis, to establish the relationship between age, gender, alco-hol intake, nutritional status, and disease extent of drug-in-duced hepatotoxicity.

Methods

PATIENTS

The study was conducted in the TB clinic of the German Nepal Tu-berculosis Project (GENETUP) from December 2001 to November2002. The study included 37 (74%) patients with active pulmonary TBand 13 (26%) with active extrapulmonary TB infection. Among those 13patients, 6 had tuberculous pleural effusion and 7 had tuberculous lym-phadenopathy.

The patients were negative for hepatitis B surface antigen, anti-HCVAb, and HIV. At the beginning of treatment, liver function testsshowed normal levels of serum aspartate aminotransferase (AST), ala-nine aminotransferase (ALT), bilirubin, alkaline phosphatase (ALP), al-bumin, and total protein. Patients receiving potentially hepatotoxic drugsin addition to the anti-TB drugs and patients with relapsed TB were ex-cluded. Fifty patients who fulfilled these criteria were selected from thetotal of 128 patients registered in the clinic during the study period. Pa-tients gave written informed consent after approval by the ethics com-mittee of the clinic.

DRUG REGIMENS

Treatment was planned as recommended by our National Tuberculo-sis Control Program. The total treatment period was 8 months, compris-ing an intensive phase of 2 months followed by a continuation phase of 6months (Table 1). Patients received directly observed treatment, shortcourse from the medical staff of the clinic (ie, pts. were observed takingeach dose of drug).

Doses were fixed according to the total body weight of the patient. Dai-ly doses were as follows: isoniazide 300 mg/day regardless of bodyweight; rifampin 300 mg (2539 kg), 450 mg (4054 kg), or 600 mg (55 kg); pyrazinamide 1000 mg (2539 kg), 1500 mg (4054 kg), or 2000mg (55 kg); and ethambutol 800 mg (2539 kg), 1000 mg (4054 kg), or1200 mg (55 kg).

DIAGNOSIS OF DRUG-INDUCED HEPATOTOXICITY

Anti-TB druginduced hepatotoxicity was defined as normalizationof liver enzyme levels and resolution of signs and symptoms of hepato-

toxicity after withdrawal of all anti-TB drugs and the presence of at leastone of the following criteria during use of anti-TB drug therapy7: (1) arise to 5 times the normal level of ALT and/or AST, (2) an increase to>1.5 mg/dL in the level of serum total bilirubin, and (3) any increase inAST and/or ALT levels above pretreatment values together with anorex-ia, nausea, vomiting, and jaundice. The normal maximum values in ourlaboratory are ALT 35 IU/L, AST 40 IU/L, and ALP 115 IU/L.

Drug-induced hepatitis is classified as hepatocellular and cholestatichepatitis. In hepatocellular hepatitis, there is marked increase in serumlevels of ALT and AST (>5 upper limit of normal [ULN] range), butmildly increased ALP and -glutamyl transferase. In cholestatic hepatitis,there is a mildly elevated serum transferase level (3 ULN) and bilirubin.

DESIGN

Liver function was monitored by measuring the serum levels of AST,ALT, ALP, bilirubin (total and direct), total protein, and albumin with thehelp of an auto-analyzer in the pathology laboratory of Dhulikhel Hospital.

Pretreatment liver function tests were conducted and, after drug thera-py was initiated, were performed a week later, then biweekly for at least 2months. The tests were repeated later whenever symptoms suggestive ofhepatotoxicity (eg, nausea, anorexia, malaise, vomiting, organomegaly,jaundice) occurred. Patients were observed closely and instructed to re-port any unusual signs and symptoms during their treatment period.

If a patient developed hepatotoxicity, medications were stopped im-mediately and serum enzymes were measured twice weekly until symp-toms resolved and the levels decreased to 2 times the ULN.8

Low body weight was considered as 10% below normal for gender andheight; values 6 units (48 g ethanol) per day for more than one year.

DATA ANALYSIS

The incidence of hepatotoxicity was determined from the rate of he-patic adverse reaction cases obtained from the population beginning anti-TB therapy. The incidence of hepatotoxicity was defined as the numberof hepatic adverse reactions causing at least one weeks interruption dur-ing a given treatment period divided by the number of patients registeredduring the same period.

Elevations in serum AST and ALT levels (pretreatment vs peak lev-els during the treatment period) were analyzed by paired t-test. Statisticalanalysis was performed using SPSS version 10.0. Rates of hepatotoxici-ty due to anti-TB drugs were also calculated for several subgroups of thecohort (male vs female, elder [>35 y] vs younger [1535 y], alcoholic vsnon-alcoholic). Comparisons of ratings were determined using themeans of the Fishers exact test (Epi Info, version 6.00).

The Annals of Pharmacotherapy 2004 June, Volume 38 1075www.theannals.com

Table 1. Treatment Regimen According to NationalTuberculosis Control Program of Nepal

RegimenTreatment Intensive Phase Continuation PhaseCategory (2 mo) (6 mo)

Ia isoniazid, rifampin, pyrazinamide, isoniazid, ethambutolethambutol

IIIb isoniazid, rifampin, pyrazinamide isoniazid, ethambutol

TB = tuberculosis.aNew sputum smearpositive pulmonary TB, newly diagnosed seri-ously ill patients with severe forms of TB.bSputum smearnegative pulmonary TB with limited parenchymal in-volvement, extrapulmonary TB (less severe form).

Results

The study was conducted in GENETUP, which is one ofthe first projects in South East Asia to use directly ob-served short-term therapy. It is situated in the heart ofNepals capital city (Kalimati, Kathmandu).

There were 22 (44%) female and 28 (56%) male pa-tients enrolled in the study (Table 2). Their ages ranged be-tween 15 and 57 years. Patients 1535 years of age wereconsidered younger, and those >36 years of age were con-sidered older. Thirty-seven (74%) patients had active pul-monary TB and 13 (26%) had active extrapulmonary TBinfection.

During the study period, 4 of 50 patients with active TBdeveloped hepatotoxicity, detected by clinical examinationand confirmed by liver function tests. Based on this find-ing, we project that 8% of the urban Nepalese populationis susceptible to hepatic adverse effects of anti-TB drugs.

Administration of anti-TB drugs was associated with anelevation of liver enzymes. The mean SD pretreatmentserum level of ALT was 23.50 7.58 IU/L, and for AST itwas 26.05 16.89 IU/L. These values were elevated to58.47 34.23 IU/L for ALT (t = 5.467; p < 0.01) and42.56 28.98 IU/L for AST (t = 4.55; p < 0.01 for AST).ALT was elevated to 2 times the pretreatment level in 38%

of the patients and >3 times in 30% of the patients. Similar-ly, 40% and 29% of patients showed 2 and >3 times eleva-tion of AST, respectively. Asymptomatic patients showingelevation in transferases (>3 but 3 ULN,total bilirubin >1.5 mg/dL). One patient showed hepatocel-lular hepatitis. He had significantly elevated transferaselevels (AST and ALT >5 ULN); the bilirubin level wasalso raised.

Symptoms shown by all patients developing drug-in-duced hepatitis were very similar. They had shown gastroin-testinal manifestations such as nausea, vomiting, abdominaldiscomfort, anorexia, and jaundice. Immediately after drug-induced hepatitis was diagnosed, anti-TB therapy was with-held temporarily until the patients clinical and biochemicalpicture normalized. The rest of the patients continued treat-ment without complications, and liver enzymes normalizedwithin a few days of continued treatment.

Patients belonging to the younger age group were foundto be at higher risk for anti-TB druginduced hepatotoxici-ty than those in the older group (6% vs 2%; p = 0.368; OR2.75), respectively (Table 4). The transferase indexes ofpatients belonging to both age groups were almost thesame. Female gender was found to be another predispos-ing factor (p = 0.219; OR 4.2). The BMI of our patientswas low (

ly as well as radiologically. The extent of the disease isprobably another predisposing factor for anti-TB drugin-duced hepatotoxicity (Table 4).

Discussion

Isoniazid, rifampin, and pyrazinamide have been suc-cessful therapeutic agents for the treatment of TB becauseof their high therapeutic efficacy and good patient accep-tance. However, a variety of adverse reactions to thesedrugs has been reported. Liver toxicity is the most com-mon adverse effect, especially during short-course therapy,which often leads to interruption of treatment.6

In the present study, 8% of the patients developed hepa-totoxicity, which is higher than the figures reported in pre-vious studies from the US and UK.5 Wide variations havebeen found in the reported incidences of hepatotoxicity dur-ing anti-TB therapy. A Japanese study investigating 77 pa-tients indicated an incidence rate of 18.25% of patients de-veloping adverse hepatic reactions in the first month of iso-niazid and rifampin treatment.9 Asian patients haveincreased susceptibility to anti-TB druginduced hepatotox-icity. This is supported by an observation done in Taipei, inwhich an incidence of 14.7% of anti-TB druginduced hep-atotoxicity was reported.10 A recent study performed inHong Kong showed an incidence of 13% in Chinese pa-tients.11 The risk of hepatotoxicity, based on data from 4prospective Indian studies, was 11.5% compared with 4.3%in 14 published studies from the Western hemisphere.12

The incidence of hepatotoxicity due to anti-TB drugs ismuch higher in studies from developing countries com-pared with those from developed countries, despite use ofsimilar regimens.9,13 The reasons for the increased incidence

of hepatotoxic reactions in developing countries are un-clear. Perhaps poor nutrition, widespread parasitism, chron-ic infections, indiscriminate use of various drugs, ethnicfactors, severity of the disease, chronic alcoholism, or ge-netic predisposition may play roles individually or collec-tively.10,14 In a study from India, Kumar et al.15 highlightedanother reason for this disparity. They confirmed hepatitisA and B in 45% of 40 patients who developed acute hepati-tis during anti-TB therapy with isoniazid and rifampin. Theburden of TB worldwide lies heavily in Asia, and patientsrequiring anti-TB chemotherapy might have the aforemen-tioned concomitant predisposing factors for hepatotoxicity.Presence of viral hepatitis can lead to misdiagnosis of drug-induced hepatitis. To avoid such confusion and make thediagnosis of drug-induced hepatitis reliable, patients withpositive serologic tests for hepatitis B and C and HIV wereexcluded in our study. Although all of the patients showedelevation in hepatic enzymes, in most cases it was 3 ULN; bilirubin >1.5 mg/dL) in patients show-ing characteristics of cholestatic hepatitis. Cholestatic in-jury represents altered bile flow through inflammation, dis-ruption, or destruction of bile ductules or resulting from alarger bile duct. One patient had markedly elevated amino-transferase levels (AST, ALT >5 ULN), suggesting hepa-tocellular hepatitis. Hepatocellular hepatitis mimics viralhepatitis with markedly elevated aminotransferase val-ues.4,17 The actual cause of the hepatocyte injury or celldeath is direct damage to or destruction of cellular mem-branes or covalent binding of toxic metabolites to livermacromolecules, leading to impairment of calcium home-ostasis, mitochondrial dysfunction, or failure of other cel-lular systems.18-20

Jaundice was seen in our patients as the presentingsymptom of anti-TB druginduced hepatotoxicity.2,21,22Jaundice can result from either hepatocellular or cholestat-ic injury of the liver. All cases of drug-induced hepatotox-icity developed within 2 months of initiation of therapy.23-25

Patients enrolled in the study were taking a combinationof anti-TB drugs. Because of this, it is difficult to concludewhich drug was the main culprit for hepatitis. Althoughisoniazid is the major drug incriminated, the role of otherpossibly hepatotoxic drugs (eg, rifampin, pyrazinamide)can also be speculated. Previous studies have proven thatthe order of risk is isoniazid + rifampin > isoniazid > pyra-zinamide > rifampin > ethambutol.26

Most patients enrolled in our study were of the youngerage group (median 28 y). This may be the reason for thehigh rate of hepatotoxicity shown in younger patients. Wefound that female gender was an independent predictor of

Hepatotoxicity Due to Antitubercular Medicines

The Annals of Pharmacotherapy 2004 June, Volume 38 1077www.theannals.com

Table 4. Demographic Characteristics of Patients with Hepatotoxicity

HepatotoxicityCharacteristic Cases (n) Incidence (%)

Total pts.(n = 50) 4 8

Gendermale (n = 28) 1 4female (n = 22) 3 14

Age (y)1535 (n = 27) 3 113660 (n = 23) 1 4

Disease extentsputum-positive (n = 37) 3 8sputum-negative (n = 13) 1 8

Alcohol intakeyes (n = 9) 0 0no (n = 41) 4 10

Albumin (g/dL)3.5 (n = 22) 1 4.5

anti-TB druginduced hepatotoxicity.4,17,21,27 Although thefrequency of drug-induced liver injury was found to behigher in females, the severity of hepatotoxicity was notrelated to gender. The difference in the incidence of drug-associated hepatotoxicity between males and females ismainly due to:

1. pharmacokinetic variations, probably slower biotrans-formation and subsequent clearance of exogenous mol-ecules due to lower levels of microsomal enzymes; and

2. women probably being acetylators (slow acetylatorenzymatic pattern shows male:female ratio of 4:1).28

The nutritional status (assessed by BMI and serum albu-min)29 of our patients seemed poor (mean BMI 18.7 kg/m2,serum albumin 2.8 mg/dL). This may be one of the riskfactors for drug-induced hepatotoxicity.11,22,27 We foundthat patients with pretreatment hypoalbuminemia had atwofold higher risk of developing hepatotoxicity. In mal-nutrition, glutathion stores are depleted, which makes onevulnerable to oxidative injury. In a malnourished person,the liver metabolizes drugs at a slower pace. In a studyconducted in India, the incidence of hepatotoxicity wasfound to be 3 times higher in malnourished patients.30

Drug-induced hepatotoxicity may be fatal if detectedlate. Prompt recognition of the condition with immediatewithdrawal of the offending agent is the mainstay of thera-py. The British Thoracic Society suggests that, if there is arise in ALT and/or AST to >3 times the ULN or an in-crease in bilirubin, or if the patient shows clinical symp-toms of hepatitis, then drugs should be stopped and rein-troduced sequentially when these parameters fall to withinnormal levels. In our study, abnormalities of liver bio-chemistries and symptoms shown by 4 patients suggesteddiscontinuing treatment. Within a few days after cessationof drug therapy, liver enzymes returned to normal levels.Normalization of liver enzyme levels once administrationof anti-TB drugs has been halted proves that all signs andsymptoms shown by the patient are related to the adminis-tration of anti-TB drugs.31

Once serum transferase levels have normalized and pla-teaued, reinstitution of anti-TB therapy was done. It is awell-accepted fact that the risk of adverse effects must bebalanced with the benefits of effective TB treatment. Pro-longed interruption of treatment may lead to undesired drugresistance and may prolong the therapy. In the present study,anti-TB therapy was reintroduced according to the severityof hepatotoxicity. In the case of mild hepatotoxicity, alldrugs were reintroduced simultaneously, initially in lowerdoses that were increased on subsequent days, with dailymonitoring of the patients clinical and biochemical condi-tions. If hepatotoxicity was severe, then after recovery, lowdoses of isoniazid and ethambutol were reintroduced andthe patient was monitored closely. If liver function test val-ues remained normal and the patient remained asymptomat-ic for a week, then other drugs were added, initially withlow doses that were increased on subsequent days.

All of our patients were closely monitored during theirtreatment period, and none showed a recurrence of hepato-toxicity. They completed their anti-TB treatment success-

fully without further complications. It can therefore beconcluded that it is possible to reintroduce potentially hep-atotoxic agents easily after recovery.22

Our study has resulted in some changes in the treatmentstrategy in the clinic in which the study was conducted. Wehave started monitoring patients more closely with theabove-stated risk factors. Pretreatment liver function testsare performed on all patients. Patient education (regardingadverse effects of anti-TB drugs) is now given highest pri-ority because most mishaps can be controlled by patientsinvolvement in their treatment.

Summary

With the increasing incidence of TB worldwide, agreater number of patients are exposed to the risk of poten-tially serious hepatotoxic effects of anti-TB drugs. The fac-tor of greatest clinical importance in treatment of TB isprobably early recognition of hepatic dysfunction, which ispossible only by regular monitoring of the patient. Patientswho ultimately die or require liver transplants frequentlyhave a history of continued use of these medicines even af-ter symptoms of hepatotoxicity, including jaundice, haveappeared. Further study should be performed with largernumbers of patients from all areas of Nepal to generate pre-cise toxicity data that could be applied to evaluate medicalinterventions and assist in the development of health policy.

Rajani Shakya MPharm, Lecturer, Department of Pharmacy, Kath-mandu University, Dhulikhel, NepalB Subba Rao MPharm, Head of Department, Department of Phar-macy, Kathmandu University Bhawana Shrestha MD, Director, German Nepal TuberculosisProject (GENETUP), Kalimati, Kathmandu, NepalReprints: Rajani Shakya MPharm, Department of Pharmacy, Kath-mandu University, Dhulikhel, Nepal, fax 00977-11-61443, rajani@ku.edu

We thank the staff of GENETUP for their active participation in this study.

References1. A review of current epidemiological data and estimation of future tuber-

culosis incidence and mortality. WHO/TB/93,173. Geneva: World HealthOrganization, 1993.

2. Durand F, Jebrak G, Pessayre D, Fournier M, Bernuau J. Hepatotoxicityof antituberculosis drugs: rationale for monitoring liver status. Drug Saf1996;15:394-405.

3. Timbrell JA, Park BK, Harland SJ. A study of the effects of rifampicinon isoniazid metabolism in human volunteers. Hum Toxicol 1985;4:279-85.

4. Snider DE, Caras GJ. Isoniazid-associated hepatitis deaths: a review ofavailable information. Am Rev Respir Dis 1992;145:494-7.

5. Omerod LP, Skinner C, Wales J. Hepatotoxicity of antituberculosisdrugs. Thorax 1996;51:111-3.

6. His Majestys Government of Nepal Ministry of Health and the WorldHealth Organization. Tuberculosis Control in Nepal 19951999. A de-velopment plan for the National TB Programme. Kathmandu, Nepal,1995.

7. Tanaoglu K, Atac G, Sevim T, Tarun T, Yazicioglu O, Horzum G, et al.The management of anti-tuberculosis druginduced hepatotoxicity. Int JTuberc Lung Dis 2001;5:65-9.

8. Guidelines for the management of adverse drug effects of antimycobac-terial agents. Lawrence Flick Memorial Tuberculosis Clinic, Philadel-phia Tuberculosis Control Program, November 1998.

1078 The Annals of Pharmacotherapy 2004 June, Volume 38 www.theannals.com

R Shakya et al.

9. Ohno M, Yamaguchi I, Yamamoto I, Fukuda T, Yokota S, Maekura R, etal. Slow N-acetyltransferase 2 genotype affects the incidence of isoniazidand rifampicin-induced hepatotoxicity. Int J Tuberc Lung Dis 2000;4:256-61.

10. Turktas H, Unsal M, Tulek N, Oruc O. Hepatotoxicity of antituberculosistherapy or viral hepatitis. Tubercle Lung Dis 1994;75:58-60.

11. Huang YS, Chern HD, Su WJ, Wu JC, Lai SL, Yang SY, et al. Polymor-phism of the N-acetyltransferase 2 gene as a susceptibility risk factor forall antituberculosis drugsinduced hepatitis. Hepatology 2002;35:883-9.

12. Steele MA, Burk RF, DesPrez RM. Hepatitis with isoniazid and rifam-pin. A meta-analysis. Chest 1991;99:465-71.

13. Snider DE Jr, Long MW, Cross FS, Farer LS. Six-months isoniazidrifam-pin therapy for pulmonary tuberculosis. Report of a United States HealthService Co-operative Trial. Am Rev Respir Dis 1984;129:573-9.

14. Dutt AS, Moers D, Stead WW. Short course chemotherapy for tubercu-losis with mainly twice weekly isoniazid and rifampin. Communityphysicians seven-year experience with mainly outpatients. Am J Med1984;77:233-42.

15. Kumar A, Misra PK, Mehotra R, Govil YC, Rana G. Hepatotoxicity ofrifampin and isoniazid. Is it all drug-induced hepatitis? Am Rev RespirDis 1991;143:1350-2.

16. British Thoracic and Tuberculosis Association. Short course chemothera-py in pulmonary TB. Lancet 1975;1:119-24.

17. Mitchell JR, Zimmerman HJ, Ishak KG, Thorgeirsson UP, Timbrell JA,Snodgrass WR, et al. Isoniazid liver injury: clinical spectrum, pathology,and probable pathogenesis. Ann Intern Med 1976;84:181-92.

18. Altman C, Biour M, Grange J. [Hepatotoxicity of antitubercular agents.Role of different drugs. 199 cases] French. Presse Med 1993;22:1212-6.

19. Schenker S, Martin RR, Hoyumpa AM. Antecedent liver disease anddrug toxicity. J Hepatol 1999;31:1098-105.

20. Sturgill MG, Lambert GH. Xenobiotic-induced hepatotoxicity: mecha-nism of liver injury and methods of monitoring hepatic function. ClinChem 1997;43:1512-26.

21. Askgaard DS, Wilcke T, Dossing M. Hepatotoxicity caused by the com-bined action of isoniazid and rifampicin. Thorax 1995;50:213-4.

22. Singh J, Garg PK, Tandon R. Hepatotoxicity due to antituberculosis ther-apy. Clinical profile and reintroduction of the therapy. J Clin Gastroen-terol 1996;22:211-4.

23. Koponoff DE, Snider DE Jr, Caras GJ. Isoniazid-related hepatitis: a USPublic Health Service Cooperative surveillance study. Am Rev RespirDis 1978;117:991-1001.

24. Donald PR, Schoeman JF, OKennedy A. Hepatotoxicity during chemo-therapy for severe tuberculosis meningitis. Am J Dis Child 1987;141:741-3.

25. Pessayre D, Bentata M, Degott C, Nouel O, Miguet JP, Rueff B, et al.Isoniazidrifampin fulminant hepatitis. A possible consequence of en-hancement of isoniazid hepatotoxicity by enzyme induction. Gastroen-terology 1977;72:284-9.

26. OBrien RJ. Hepatotoxic reaction due to antituberculosis drugs. Adjust-ment to therapeutic regimen. JAMA 1991;265:3323.

27. Gronhagen-Riska C, Hellstrom PE, Froseth B. Predisposing factors inhepatitis induced by isoniazidrifampin treatment of tuberculosis. AmRev Respir Dis 1978;118:461-6.

28. Marvin W. Impacts of gender on drug responses. Drug Topics 1998;5:91-100.

29. Krishnaswamy K, Prasad CE, Murthy KJ. Hepatic dysfunction in under-nourished patients receiving isoniazid and rifampicin. Trop Geogr Med1991;43:156-60.

30. Mehta S. Malnutrition and drugs: clinical implications. Dev PharmacolTher 1990;15:159-65.

31. de Souza AF, de Oliveira e Silva A, Baldi J, de Souza TN, Rizzo PM.[Hepatic functional changes induced by the combined use of isoniazid,pyrazinamide, and rifampicin in the treatment of pulmonary tuberculo-sis] Portuguese. Arq Gastroenterol 1996;33:194-200.

EXTRACTO

TRASFONDO: Los medicamentos antituberculosos alteran la funcinheptica, y sta se demuestra mediante un examen clnico y las pruebasde funcin heptica. El potencial de hepatotoxicidad de algunos de losagentes antituberculosos de primera seleccin sigue siendo un problemahoy da, especialmente durante la etapa inicial del tratamiento.

OBJETIVO: Determinar la incidencia de la hepatotoxicidad inducida pormedicamentos antituberculosos en la poblacin urbana nepalesa yevaluar los factores de riesgo. MTODOS: Cincuenta pacientes diagnosticados con tuberculosis activa yfuncin heptica normal antes del tratamiento fueron monitorizadosclnica y bioqumicamente en un anlisis de cohorte prospectivo.RESULTADOS: Se encontr que los medicamentos antituberculosos estnasociados con una alteracin de la funcin heptica que causa un aumentovariable de las enzimas del hgado (t = 4.550, p = 0.00 para el aspartatoaminotransferasa, y t = 5.467, p = 0.00 para la alanina aminotransferasa,95% CI). Treinta y ocho por ciento de los pacientes tuvieron unaelevacin de 2 veces ms de la ALT, y 30 por ciento tuvo una elevacinmayor de 3 veces ms de la ALT. Similarmente, el 40 por ciento y el 29por ciento demostraron una elevacin de la AST de 2 veces ms y de 3veces ms, respectivamente. Cuatro pacientes (8 por ciento) desarrollaronhepatotoxicidad inducida por medicamentos. Todos presentaron ictericia.El intervalo de tiempo del inicio de la hepatotoxicidad fue de 12 a 60 das(28 das, mediana). Se encontr que la hepatotoxicidad inducida pormedicamentos antituberculosos fue mayor en pacientes jvenes (6 porciento vs 2 por ciento, p > 0.05, p = 0.368, Odds Ratio [OR]; 2.75). Lasmujeres demostraron un riesgo mayor (p > 0.05, p = 0.219, Odds ratio[OR]; 4.2). La mayora de los pacientes que desarrollaron hepatitistuvieron una prueba de esputo positiva. El prximo factor depredisposicin fue la condicin nutricional, evaluada por el ndice demasa corporal (IMC) y el nivel de albmina srico. CONCLUSIONES: El hallazgo de una incidencia de un 8 por ciento de latasa de hepatotoxicad es considerablemente alta. Los factores de riesgode la hepatotoxicidad incluyeron el sexo femenino, la extensin de laenfermedad, y una condicin nutricional deficiente. La deteccinoportuna y el retiro temporal del medicamento causante pueden curar lahepatotoxicidad inducida por los medicamentos antituberculosos.

Rafaela Mena

RSUMINTRODUCTION: Les agents antituberculeux peuvent entraner unedysfonction hpatique qui se traduit par des signes cliniques et des tests dela fonction hpatique anormaux. Lhpatotoxicit potentielle de quelquesagents antituberculeux de premire intention est toujours un problme lheure actuelle, particulirement durant la priode initiale de traitement.OBJECTIF: Dterminer lincidence de lhpatotoxicit induite par lesagents antituberculeux chez la population urbaine du Npal et en tablirles facteurs de risque.MTHODOLOGIE: Cinquante patients avec un diagnostic de tuberculoseactive et une fonction hpatique normale avant le traitement ont t inclusdans une cohorte prospective. Des paramtres cliniques et biochimiquesont t compils et analyss pour lensemble de ces patients.RSULTATS: Les agents antituberculeux ont t associs une dysfonctionhpatique rsultant en une lvation des enzymes hpatiques un degrvariable (t = 4.55, p = 0.00, p < 0.05 pour lAST et t =-5.467, p = 0.00,p < 0.05 pour lALT avec un intervalle de confiance 95%). Trentehuitpour cent des patients ont eu 2 reprises et 30% plus de 3 reprises unelvation de lALT. De faon similaire, 40% et 29% des patients ont eudes lvations dAST 2 et plus de 3 reprises, respectivement. Quatrepatients (8%) ont dvelopp une hpatotoxicit mdicamenteuse, dontlictre tait le symptme principal. Lhpatotoxicit est apparue en 12 60 jours, avec une mdiane de 28 jours. Lhpatotoxicit induite par lesagents antituberculeux a t plus frquente chez les jeunes patients (6%vs 2%, p > 0.05, p = 0.368, [OR]: 2.75). Le sexe fminin tait aussi unfacteur de risque (p > 0.05, p = 0.219, [OR]: 4.2). La majorit despatients ayant dvelopp une hpatite taient ceux avec des crachatspositif pour la tuberculose, donc ceux avec une maladie plus svre. Lestatut nutritionnel, dtermin par lindice de masse corporelle et leniveau srique dalbumine, tait aussi un facteur prdisposant.CONCLUSIONS: Un taux dincidence dhpatotoxicit de 8% estconsidrablement lev. Les facteurs de risque sont le sexe fminin,ltendue de la maladie, et un mauvais tat nutritionnel. La dtectionprcoce et la suspension temporaire de lagent en cause peut gurircompltement lhpatotoxicit induite par les agents antituberculeux.

Esthel Rochefort

Hepatotoxicity Due to Antitubercular Medicines

The Annals of Pharmacotherapy 2004 June, Volume 38 1079www.theannals.com