Tropical Medicine and International Health volume 15 no 2 pp 268272 february 2010

doi:10.1111/j.1365-3156.2009.02449.x

Antituberculosis drug-induced hepatotoxicity is uncommon in Tanzanian hospitalized pulmonary TB patients

Alma Tostmann1,2, Jossy van den Boogaard2,3, Hadija Semvua4, Riziki Kisonga5, Gibson S. Kibiki4, Rob E. Aarnoutse3 and Martin J. Boeree1,2
1 2 3 4 5 Department of Pulmonary Diseases, Radboud University Nijmegen Medical Centre, Nijmegen, The Netherlands University Centre for Chronic Diseases Dekkerswald, Radboud University Nijmegen Medical Centre, Nijmegen, The Netherlands Department of Clinical Pharmacy, Radboud University Nijmegen Medical Centre, Nijmegen, The Netherlands Kilimanjaro Clinical Research Institute, Kilimanjaro Christian Medical Centre, Moshi, Tanzania Kibongoto National Tuberculosis Hospital, Sanya Juu, Tanzania

Summary

Data on antituberculosis drug-induced hepatotoxicity in sub Saharan Africa are limited, probably because liver function tests are not carried out routinely during tuberculosis treatment in most African countries. We monitored the liver function of 112 Tanzanian hospitalized pulmonary tuberculosis patients during the rst 2 months (i.e. the intensive phase) of tuberculosis treatment. The rate of hepatotoxicity in our study was 0.9% (95% CI 0.044.3%). It is encouraging to nd a lower rate of antituberculosis drug-induced hepatotoxicity than one would expect based on the high prevalence of risk factors such as HIV and hepatitis B. keywords antitubercular treatment, adverse reactions, liver toxicity, Africa

Introduction With nine million new cases and almost two million deaths in 2007, tuberculosis (TB) remains a major cause of illness and death worldwide. The highest incidence rate is found in Africa, due to the HIV epidemic. TB is the leading cause of death among HIV patients (WHO 2009). Standard TB treatment consists of a 6-month course with isoniazid, rifampicin, pyrazinamide and ethambutol. In a clinical setting, this regimen can cure more than 95% of the patients with active TB caused by normally sensitive Mycobacterium tuberculosis. However, actual cure rates are lower. Adverse effects disrupt therapy adherence, which may cause treatment failure, relapse or drug resistance (Kaona et al. 2004; Wares et al. 2003; WHO/ IUATLD 2008). Isoniazid, rifampicin and pyrazinamide are potentially hepatotoxic drugs. They are metabolised in the liver, making this organ vulnerable for injury. Antituberculosis drug-induced hepatotoxicity (ATDH) is a serious adverse effect that can be fatal if therapy is not interrupted in time (Saukkonen et al. 2006; Tostmann et al. 2008) ATDH occurs in 228% of the tuberculosis patients; the variation is large due to different denitions of hepatotoxicity and differences between the populations studied (Tostmann

et al. 2008). Data on ATDH in sub Saharan Africa are scarce, probably because liver function tests are not carried out during tuberculosis treatment in most African countries. ATDH is reported to be <2%, but liver function was not closely monitored in these studies (Perriens et al. 1995; Johnson et al. 2000; Tostmann et al. 2007). It is important to have a valid estimation of the incidence of ATDH in this region. It may help to explain low cure rates and is particularly relevant considering the concurrent antiretroviral treatment, which exacerbates hepatotoxicity (Kwara et al. 2005). New tuberculosis treatment regimens containing moxioxacin and a high dose of rifampicin are currently being tested in subSaharan Africa. To estimate the rate of ATDH in an African population, we closely monitored the liver function in Tanzanian hospitalized patients during the intensive phase of TB treatment. Methods This study was conducted between April 2007 and July 2008 at the Kibongoto National Tuberculosis Hospital in Sanya Juu, northern Tanzania. All adults who were not on antiretroviral therapy before starting tuberculosis treatment were eligible for participation.

268

2010 Blackwell Publishing Ltd

Tropical Medicine and International Health A. Tostmann et al. Antituberculosis drug-induced hepatotoxicity

volume 15 no 2 pp 268272 february 2010

Tuberculosis was treated according to guidelines of the Tanzanian National Tuberculosis and Leprosy Programme. During the intensive phase of treatment, patients with a body weight <50 kg received 225 mg isoniazid, 450 mg rifampicin, 1200 mg pyrazinamide and 675 mg ethambutol daily (three xed-dose combination [FDC] tablets, Novartis) and patients >50 kg received 300 mg isoniazid, 600 mg rifampicin, 1600 mg pyrazinamide and 900 mg ethambutol daily (four FDC tablets). All patients received full facility-based directly observed therapy (DOT), as they were admitted to the hospital during the intensive phase of treatment. Demographic data were collected before treatment. Hepatitis B surface antigen (HbsAg) and hepatitis C virus antibodies were detected with a qualitative rapid chromatographic immunoassay (ACON Laboratories Inc, San Diego, CA, USA). Serum alanine aminotransferase (ALT), aspartate aminotransferase (AST) and total bilirubin were determined at baseline and after 2, 4, 6 and 8 weeks of TB treatment, using the Roche Reotron (Kerkhof Medical Service, the Netherlands). Hepatotoxicity was dened as serum ALT > 3 times the upper limit of normal (ULN) in the presence of hepatotoxicity symptoms or ALT > 5 ULN without symptoms of hepatotoxicity (Tostmann et al. 2008). The reference values were 41 U l for ALT, 40 U l for AST and 17 lmol l for bilirubin. Every 2 weeks, a research nurse screened the patients for symptoms of hepatotoxicity (jaundice, abdominal pain, nausea and vomiting). Parameters are presented as mean (standard deviation; SD) when normally distributed or median (inter-quartile range; IQR) if skewed. Statistical analysis were performed using spss for Windows (version 14.0; SPSS). Ethical considerations Ethical clearance was granted by Radboud University Nijmegen, the Institutional Review Board of the Kilimanjaro Christian Medical Centre (because the study was hosted by KCMC) and the Ministry of Health of Tanzania. All patients gave written informed consent. Results We included 117 patients in this study. Five patients were excluded from analysis: three died in the rst week of treatment, for reasons other than possible drug-induced hepatotoxicity, and two withdrew from the study immediately after baseline measurements. Of the 112 patients included in the analysis, 98 (87.5%) completed the 2-month study period. 11 (9.8%) were HIV positive and

6 of them started antiretroviral treatment during the intensive phase of TB treatment. Table 1 shows the patients characteristics. None of the patients developed hepatotoxicity during the rst 2 months of TB treatment. Seven patients (6.3%) developed slight ALT increases, all <3 times the ULN. The highest ALT observed was 87 U l; the median of peak ALT values during the study period was 12 U l (IQR 921 U l). The highest AST was 64 U l, with a median peak value of 17 U l (IQR 1222 U l). Table 2 shows the liver biochemistry at baseline and during the intensive phase of treatment. One patient developed hepatotoxicity symptoms in the sixth week of treatment: serious abdominal pain, nausea and vomiting and discoloration of the conjunctives after intake of the TB drugs which disappeared again after a few hours. Because serum ALT and bilirubin were completely normal, this patient did not have hepatotoxicity according to our denition. She had lost a lot of weight during treatment (from 54 to 34 kg) and was using four FDC tablets a day (for >50 kg). Treatment was interrupted and relieved the symptoms, after which she continued with three FDC tablets.
Table 1 Patient characteristics N = 112 Age, median years (IQR) Male sex, n (%) Weight, mean kg (SD) HIV, n (%) Positive Negative Unknown Hepatitis B, n (%) Positive Negative Unknown Hepatitis C, n (%) Positive Negative Unknown Study outcome, n (%) Study completed LTF between week 2 and 4 LTF between week 4 and 6 LTF between week 6 and 8 Concomitant drugs Antiretroviral drugs* Cotrimoxazole 32 (2740) 88 (78.6%) 52.4 (7.6) 11 (9.8%) 90 (80.4%) 11 (9.8%) 8 (7.1%) 88 (78.6%) 16 (14.3%) 3 (2.7%) 93 (83.0%) 16 (14.3%) 98 3 5 6 (87.5%) (2.7%) (4.5%) (5.4%)

6 (5.4%) 8 (7.1%)

LTF, Lost to follow-up; IQR, inter quartile range; SD, standard deviation. *Antiretroviral treatment in Tanzania: Combivir (zidovudine + lamivudine) + efavirenz.

2010 Blackwell Publishing Ltd

269

Tropical Medicine and International Health A. Tostmann et al. Antituberculosis drug-induced hepatotoxicity

volume 15 no 2 pp 268272 february 2010

Table 2 Liver biochemistry during the intensive phase of treatment Liver function parameter Baseline Median (IQR) Maximum Intensive phase* Median (IQR) Maximum

ALT 10 U l (715) 53 U l 12 U l (921) 87 U l

AST 13 U l (1018) 91 U l 17 U l (1222) 64 U l

Bilirubin 11 lmol l (916) 43 lmol l 13 lmol l (1019) 47 lmol l

ALT, alanine aminotransferase; AST, aspertate aminotransferase; Bilirubin, total bilirubin; IQR, inter quartile range. *Based on highest value measured during the intensive phase of TB treatment.

Three patients developed hyperbilirubinemia (bilirubin levels > 2.5 times ULN). ALT and AST levels were within the normal range and jaundice was not reported in these patients. 14 patients (12.5%) dropped out during the study. Nine patients withdrew because they refused to give any more blood samples, due to cultural believes concerning blood; 3 after week 2, 4 after week 4 and 2 after week 6 of treatment. Discussion This is the rst study to date where the liver function is closely monitored during TB treatment in a sub-Saharan African population. None of our 112 Tanzanian pulmonary TB patients developed hepatotoxicity during the intensive phase of TB treatment. One patient developed symptoms of hepatotoxicity. Although her liver biochemistry remained normal, the symptoms could have been caused by the TB drugs. Therefore, the rate of hepatotoxicity in our study was 0.9% (95% CI 0.044.3%) at the most. Clinical trials with moxioxacin and a high dose of rifampicin are in preparation in the sub Saharan African region. Rifampicin-induced hepatotoxicity may be doserelated (Burman et al. 2001). It is encouraging to observe that the rate of TB drug-induced hepatotoxicity is lower than expected in an area with HIV and hepatitis B. Although case reports of moxioxacin-induced hepatotoxicity have been published, moxioxacin is considered safe and not expected to increase the risk of hepatotoxicity (Bertino & Fish 2000; Ho et al. 2009). Our ndings are in line with previous reports which indicate a low rate of ATDH in sub Saharan African TB patients. Recently, a study from South Africa reported that only 2 of 400 TB patients (0.5%) developed 270

hepatotoxicity during treatment (Marks et al. 2009). In a trial on cotrimoxazole prophylaxis in Malawian HIVpositive pulmonary TB patients, about 2% of the patients developed hepatotoxicity during TB treatment (Tostmann et al. 2007). Since cotrimoxazole can be hepatotoxic in patients with AIDS, the ATDH rate may even be slightly overestimated (Kovacs et al. 1984). Two studies from Congo and Uganda reported no hepatotoxicity (Johnson et al. 2000) and 1% hepatotoxicity (Perriens et al. 1995) during standard TB treatment. Even though liver function was not routinely monitored and treatment was not directly-observed, this indicates that hepatotoxicity during TB treatment is not a big problem. It is difcult to explain why the hepatotoxicity rate is low in sub-Saharan Africa despite the prevalence of risk factors such as HIV and hepatitis B. Hepatotoxicity rates of 1315% have been reported from India and Iran (Baghaei et al. 2009). Our study population was characterised by a low proportion of women, a relatively young age, 10% HIV and 7% hepatitis B carriers. Old age, female sex, malnutrition, HIV, alcohol usage, underlying liver disease such as hepatitis and several genetic polymorphisms are risk factors for ATDH (Saukkonen et al. 2006; Tostmann et al. 2008). Several genetic polymorphisms in drug metabolising enzymes have been associated with TB drug-induced hepatotoxicity, such as slow acetylator status (N-acetyltransferase 2), cytochrome P450 2E1 c1 c1 genotype and a glutathione S-transferase M1 homozygote null genotype (Sun et al. 2008). Slow and intermediate acetylators are highly prevalent in African populations (Sabbagh et al. 2008; Matimba et al. 2009). Almost 95% of the black South Africans had a CYP2E1 c1 c1 genotype (Chelule et al. 2006) and 33% of the Tanzanians have a GSTM1 null genotype (Dandara et al. 2002). These studies suggests that high risk genotypes are prevalent amongst Africans, which would contradict the low rate of hepatotoxicity found in our study. We monitored the liver function for 2 months only. Even though ATDH most often occurs in the rst 2 months of treatment (Tostmann et al. 2008), this may have slightly underestimated the hepatotoxicity rate. Our study was limited to hospitalised patients only, which could have overestimated the rate of hepatotoxicity, since generally higher rates of hepatotoxicity are seen in hospitalised patients. To estimate the rate of hepatotoxicity in sub Saharan African patients more accurately, a similar study could be performed in a larger group of TB patients that are not hospitalised. Ten percentage of our patients were HIV positive; six of them had started antiretroviral therapy during TB

2010 Blackwell Publishing Ltd

Tropical Medicine and International Health A. Tostmann et al. Antituberculosis drug-induced hepatotoxicity

volume 15 no 2 pp 268272 february 2010

treatment (zidovudine, lamivudine and efavirenz). None developed any increase in ALT or AST. Concurrent TBHIV therapy is often complicated by overlapping toxicities and drug-drug interactions (Kwara et al. 2005). Drug toxicity, including hepatotoxicity, is a major cause of TB or HIV treatment interruption during combined TB-HIV treatment (Dean et al. 2002). Nevirapine is the most hepatotoxic non-nucleoside reverse transcriptase inhibitor (Sanne et al. 2005). The majority of the nucleoside reverse transcriptase inhibitors (e.g. didanosine and stavudine) are potentially hepatotoxic, and hepatotoxicity has been reported for some protease inhibitors (e.g. ritonavir, indinavir and saquinavir). The worldwide incidence of hepatotoxicity during antiretroviral treatment ranges from 2% to 18% (Nunez 2005). In sub Saharan Africa, data are conicting. In the study of Marks et al. (2009) the rate of hepatotoxicity during TB treatment in HIV positive (n = 141) and HIV negative (n = 240) TB patients was low (0.5%), while 16.3% of the HIV positive patients received antiretroviral treatment. These ndings are in contrast with another South African study by Hoffmann et al. (2007), where 4.6% of the 868 patients who took antiretroviral drugs (zidovudine, lamivudine and efavirenz) developed hepatotoxicity (grade 3 or 4; i.e. ALT > 5 times the ULN) and where the use of TB therapy increased the risk of hepatotoxicity eightfold. More knowledge on hepatotoxicity during combined HIV-TB treatment is warranted, since a signicant proportion of the TB patient population in sub-Saharan Africa is HIV positive and would need combined TB-HIV treatment (Corbett et al. 2003). A suggestion for future research would be to get more insight in country specic rates of hepatotoxicity during TB treatment and the effect of HIV and antiretroviral therapy by using data of the national TB programs. Our study and other studies (Tostmann et al. 2007; Marks et al. 2009) suggest that in the absence of symptoms, the risk of missing signicant hepatotoxicity in this population is encouragingly low. If further data is conrmatory, routine liver function monitoring during TB treatment may not be necessary in sub-Saharan Africa. Nevertheless, it remains important that TB patients receive clear instructions on possible signs and symptoms of hepatotoxicity and that they are instructed to visit their doctor in case such symptoms occur. Acknowledgements We thank the patients for participating in this study, the staff at Kibongoto National Tuberculosis Hospital in Sanya Juu, Tanzania for their cooperation and effort, and the laboratory technicians Liselotte Wolters and Arnold Ndaro for their technical support. This study was sup-

ported by a research grant from the KNCV Tuberculosis Foundation (Den Haag, the Netherlands) and Poverty Related Infection Oriented Research (PRIOR), a research network sponsored by the Netherlands Foundation for the Advancement of Tropical Research (NWO-WOTRO). References
Baghaei P, Tabarsi P, Chitsaz E et al. (2009) Incidence, Clinical and Epidemiological Risk Factors, and Outcome of DrugInduced Hepatitis Due to Antituberculous Agents in New Tuberculosis Cases. American Journal of Therapeutics [Epub ahead of print]. Bertino J & Fish D (2000) The safety prole of the uoroquinolones. Clinical Therapeutics 22, 798817. Burman WJ, Gallicano K & Peloquin C (2001) Comparative pharmacokinetics and pharmacodynamics of the rifamycin antibacterials. Clinical Pharmacokinetics 40, 327341. Chelule PK, Pegoraro RJ, Gqaleni N & Dutton MF (2006) The frequency of cytochrome P450 2E1 polymorphisms in Black South Africans. Disease Markers 22, 351354. Corbett EL, Watt CJ, Walker N et al. (2003) The growing burden of tuberculosis: global trends and interactions with the HIV epidemic. Archives of Internal Medicine 163, 10091021. Dandara C, Sayi J, Masimirembwa CM et al. (2002) Genetic polymorphism of cytochrome P450 1A1 (Cyp1A1) and glutathione transferases (M1, T1 and P1) among Africans. Clinical Medicine and Laboratory Medicine 40, 952957. Dean GL, Edwards SG, Ives NJ et al. (2002) Treatment of tuberculosis in HIV-infected persons in the era of highly active antiretroviral therapy. AIDS 16, 7583. Ho CC, Chen YC, Hu FC et al. (2009) Safety of uoroquinolone use in patients with hepatotoxicity induced by anti-tuberculosis regimens. Clinical Infectious Diseases 48, 15261533. Hoffmann CJ, Charalambous S, Thio CL et al. (2007) Hepatotoxicity in an African antiretroviral therapy cohort: the effect of tuberculosis and hepatitis B. AIDS 21, 13011308. Johnson JL, Okwera A, Nsubuga P et al. (2000) Efcacy of an unsupervised 8-month rifampicin-containing regimen for the treatment of pulmonary tuberculosis in HIV-infected adults. Uganda-Case Western Reserve University Research Collaboration. International Journal of Tuberculosis and Lung Disease 4, 10321040. Kaona FA, Tuba M, Siziya S & Sikaona L (2004) An assessment of factors contributing to treatment adherence and knowledge of TB transmission among patients on TB treatment. BMC Public Health 4, 68. Kovacs JA, Hiemenz JW, Macher AM et al. (1984) Pneumocystis carinii pneumonia: a comparison between patients with the acquired immunodeciency syndrome and patients with other immunodeciencies. Annals of Internal Medicine 100, 663671. Kwara A, Flanigan TP & Carter EJ (2005) Highly active antiretroviral therapy (HAART) in adults with tuberculosis: current status. International Journal of Tuberculosis and Lung Disease 9, 248257.

2010 Blackwell Publishing Ltd

271

Tropical Medicine and International Health A. Tostmann et al. Antituberculosis drug-induced hepatotoxicity

volume 15 no 2 pp 268272 february 2010

Marks DJ, Dheda K, Dawson R, Ainslie G & Miller RF (2009) Adverse events to antituberculosis therapy: inuence of HIV and antiretroviral drugs. International Journal of STD and AIDS 20, 339345. Matimba A, Del-Favero J, Van BC & Masimirembwa C (2009) Novel variants of major drug-metabolising enzyme genes in diverse African populations and their predicted functional effects. Human Genomics 3, 169190. Nunez M (2005) Hepatotoxicity of antiretrovirals: incidence, mechanisms and management. Journal of Hepatology 44, S132 S139. Perriens JH, St Louis ME, Mukadi YB et al. (1995) Pulmonary tuberculosis in HIV-infected patients in Zaire. A controlled trial of treatment for either 6 or 12 months. New England Journal of Medicine 332, 779784. Sabbagh A, Langaney A, Darlu P et al. (2008) Worldwide distribution of NAT2 diversity: implications for NAT2 evolutionary history. BMC Genetics 9, 21. Sanne I, Mommeja-Marin H, Hinkle J et al. (2005) Severe hepatotoxicity associated with nevirapine use in HIV-infected subjects. Journal of Infectious Diseases 191, 825829. Saukkonen JJ, Cohn DL, Jasmer RM et al. (2006) An Ofcial ATS Statement: Hepatotoxicity of Antituberculosis Therapy. American Journal of Respiratory and Critical Care Medicine 174, 935952.

Sun F, Chen Y, Xiang Y & Zhan S (2008) Drug-metabolising enzyme polymorphisms and predisposition to anti-tuberculosis drug-induced liver injury: a meta-analysis. International Journal of Tuberculosis and Lung Disease 12, 9941002. Tostmann A, Boeree MJ, Harries AD et al. (2007) Short communication: antituberculosis drug-induced hepatotoxicity is unexpectedly low in HIV-infected pulmonary tuberculosis patients in Malawi. Tropical Medicine and International Health 12, 852855. Tostmann A, Boeree MJ, Aarnoutse RE et al. (2008) Antituberculosis drug-induced hepatotoxicity: concise up-to-date review. Journal of Gastroenterology and Hepatology 23, 192202. Wares DF, Singh S, Acharya AK & Dangi R (2003) Non-adherence to tuberculosis treatment in the eastern Tarai of Nepal. International Journal of Tuberculosis and Lung Disease 7, 327335. World Health Organization (2009) Global Tuberculosis Control 2009: Surveillance, Planning, Financing: WHO Report 2009. WHO, Geneva. http://www.who.int/tb/publications/global_ report/2009/pdf/full_report.pdf World Health Organization IUATLD Global project on antituberculous drug Resistance Surveillance (2008) Anti-Tuberculosis Drug Resistance in the World. Fourth Global Report. WHO, Geneva. http://whqlibdoc.who.int/hq/2008/ WHO_HTM_TB_2008.394_eng.pdf

Corresponding Author Martin J. Boeree, University Centre for Chronic Diseases Dekkerswald, Radboud University Nijmegen Medical Centre, PO Box 66 6560 AB Groesbeek, Nijmegen, The Netherlands. Tel.: +31 (0)24 6859 563; Fax: +31 24 3610324; E-mail: m.boeree@ulc.umnc.nl

272

2010 Blackwell Publishing Ltd

Copyright of Tropical Medicine & International Health is the property of Blackwell Publishing Limited and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.