Blackwell Science, LtdOxford, UKJGHJournal of Gastroenterology and Hepatology0815-93192005 Blackwell Publishing Asia Pty LtdNovember 2005201117451752Original Article Management

of ATT-hepatotoxicityS Agal
et al.

Journal of Gastroenterology and Hepatology (2005) 20, 17451752

DOI: 10.1111/j.1440-1746.2005.04048.x

HEPATOLOGY

Monitoring and management of antituberculosis drug induced hepatotoxicity

SUBHASH AGAL,* RAJIV BAIJAL,* SNEHANSHU PRAMANIK,* NIKHIL PATEL,* PARIJAT GUPTE,* PRAFUL KAMANI* AND DEEPAK AMARAPURKAR* ,

*Department of Gastroenterology, Jagjivanram Hospital and Department of Gastroenterology, Bombay Hospital, Mumbai, India

Abstract Background: Hepatotoxicity to antituberculosis therapy (ATT) poses a major challenge. This often results in inadequate therapy. The risk of fulminant hepatic failure and mortality is high once icteric hepatitis develops. There is no consensus on monitoring protocols and for the reintroduction of ATT. Methods: All patients (from the Department of Internal Medicine and Gastroenterology, Jagjivanram Hospital and the Department of Gastroenterology, Bombay Hospital, Mumbai, India) with a diagnosis of tuberculosis, who were to receive ATT during the study period, were included in the present study for prospective periodic laboratory monitoring for the development of hepatotoxicity. Those patients who developed hepatotoxicity formed Group A (n = 21), whereas those who did not develop hepatotoxicity were included in Group C (n = 179). For the purpose of comparison with Group A, all the patients who presented directly with ATT induced hepatotoxicity during the study period were categorized as Group B (n = 24). Group A and B were further studied after normalization of liver functions for sequential reintroduction with therapeutic doses at a weekly interval. Results: In Group A, 66.6% (14 patients) of the patients were diagnosed in the asymptomatic period. Seven patients had symptomatic hepatitis, but none had icteric illness. There were no mortalities in Group A. In contrast, all the patients in Group B had symptomatic hepatitis (75% icteric hepatitis). There was a mortality rate of 16.6% (four patients). Of the 41 patients from Groups A and B who survived, reintroduction was successful in 38/39 (97.4%). In the remaining two patients who were in Group B, reintroduction was not attempted because of decompensated liver disease. Conclusions: Periodic laboratory monitoring is important in detecting hepatotoxicity at an early stage, thereby preventing mortality. Sequential reintroduction is often successful. 2005 Blackwell Publishing Asia Pty Ltd Key words: antituberculosis drug-induced hepatotoxicity, monitoring of liver functions, reintroduction of antituberculosis therapy.

INTRODUCTION
Tuberculosis (TB) is highly prevalent in developing countries. It is re-emerging in developed countries as a result of the pandemic of HIV infection, increase in the organ transplant population and transglobal migration. The World Health Organization (WHO) has declared TB to be a global emergency.1 Antituberculous drug therapy (ATT) is associated with signicant hepatotoxicity. This often results in discontinuation of the most effective rst-line drugs

such as isoniazid (H), rifampicin (R) and pyrazinamide (Z). It is the dramatic difference in cost, efcacy and length of therapy between rst-line and second-line ATT that underlies positive efforts to maintain and reintroduce rst line drugs and this also prevents the problem of drug resistance. But there exist controversies related to prevention and management of hepatotoxicity that is the result of ATT. Is monitoring of liver enzymes useful? When should ATT be discontinued? Which drugs should be discontinued and for how long? When and how should they

Correspondence: Dr Nikhil D Patel, C1/30, Sainath Cooperative Society, Opposite Poonam Chambers, Dr Annie Besant Road, Worli, Mumbai 400018, India. Email: deepakn@bom3.vsnl.net.in Accepted for publication 5 June 2004.

1746 be restarted or should a modied regimen be used instead?2 Although most guidelines suggest the need for baseline liver function tests (LFT),1,35 there is no consensus on a monitoring schedule for patients with normal baseline LFT and without any risk factors for ATT-hepatotoxicity, such as alcohol abuse and chronic liver disease. The WHO recommendations for monitoring are that there is no need for regular LFT in addition to monthly clinical monitoring. In the case of ATT-hepatotoxicity, it recommends the reintroduction of the same regimen, but to avoid Z and R in case of severe ATT-hepatotoxicity. A suggested regimen in such patients is a 2-month initial phase of daily streptomycin (SM), H and ethambutol (E) followed by a 10-month continuation phase of H and E.1 There are a broad range of guidelines from various countries and scientic bodies such as the American Thoracic Society (ATS),5 the Centers of Disease Control (CDC),6 the British Thoracic Society (BTS)4 and French guidelines3,7 for monitoring of LFT during ATT, diagnosis of ATT-hepatotoxicity and reintroduction of ATT in the case of ATT-hepatotoxicity. These recommendations are based on evidence from expert committee reviews or opinions and/or clinical experience of respected authorities, but they indicate the absence of directly applicable prospective studies of good qualities.4 A consensus guideline for managing ATT-hepatotoxicity is yet to be developed.8,9 To address the issue of the monitoring and reintroduction of ATT after ATT-hepatotoxicity, a prospective study was planned (i) to identify ATT-hepatotoxicity at an early stage; (ii) to determine morbidity and mortality associated with ATT-hepatotoxicity; (iii) to compare the patients who presented as ATT-hepatotoxicity with the patients who developed ATT-hepatotoxicity on regular monitoring; and (iv) to dene the safety of sequential reintroduction.

S Agal et al. These were subclassifed as (1) Anicteric hepatitisif jaundice was absent. (2) Icteric hepatitisif jaundice was present. (3) Fulminant hepatic failure (FHF)if the timeperiod between the development of jaundice to encephalopathy was less than 14 days in the absence of pre-existing liver disease.14 (4) Subfulminant hepatic failureif the timeperiod between the development of jaundice to encephalopathy was more than 14 days in the absence of pre-existing liver disease.14 (5) Decompensated liver diseasethe presence of jaundice, coagulopathy and/or encephalopathy in pre-existing liver cirrhosis.14 (b) Serum transaminase (alanine aminotransferase [ALT] and/or aspartate aminotransferase [AST]) elevation was more than ve times the upper limit of normal (ULN) in any patient or more than twice the ULN in any patient with symptomatic hepatitis. (c) Serum bilirubin elevation more than twice the ULN in any patient. 2 Exclusion of other etiologies for the presence of hepatotoxicityviral, alcohol and drugs. Asymptomatic transaminase elevation up to ve times the ULN was not considered as ATT-hepatotoxicity and treatment was continued as per standard protocol.

Patient population
All the patients with a recent diagnosis of tuberculosis, who were to be treated by ATT, were included in the present study. These patients were evaluated in detail at baseline. Then these patients were monitored prospectively for the development of ATT-hepatotoxicity. Patients who developed ATT-hepatotoxicity formed Group A in the present study, whereas those who never experienced ATT-hepatotoxicity were included in Group C. For the purpose of comparison with Group A, Group B comprised of all the patients who had been referred from different institutions or from different physicians for suspected ATT-hepatotoxicity. There was no uniform monitoring of ATT carried out in these patients. All the patients in Group A and Group B were further monitored and studied for reintroduction of ATT. Diagnosis of TB in both the groups was based on the presence of caseating granulomas on histology and/or a positive smear and/or culture for acid-fast bacilli in clinical samples. Patients with one or more of the following were excluded from the study groups: 1 Any concomitant illness producing a life expectancy of less than a year. 2 Known immunosuppression by HIV infection (absolute CD4 lymphocyte counts less than 200 cells/ mm3) or patients with an organ transplantation. 3 Uncontrolled congestive cardiac failure (CCF) that might interfere with interpretation of LFT. 4 Hepatic involvement as a part of military TB or isolated hepatobiliary TB that might interfere with interpretation of LFT.

METHODS
Study design
The present prospective study was carried out from June 2001 to December 2002 at Jagjivanram Western Railway Hospital and Bombay Hospital, Mumbai. The study protocol was approved by the institutional ethics review boards.

Criteria for ATT-hepatotoxicity

ATT-hepatotoxicity was dened in the present study by the presence of the following criteria:46,913 1 Presence of one or more of the following: (a) Symptoms suggestive of hepatitis (symptomatic hepatitis); nausea, vomiting, malaise, anorexia, right upper quadrant discomfort, jaundice, hemorrhagic spots/easy bruising (coagulopathy) and/or confusion/asterixis/unconsciousness (encephalopathy).

Management of ATT-hepatotoxicity 5 Presence of abnormal LFT before the start of ATT. 6 Those who refused to give informed written consent for recruitment in the study. 7 Presence of alcohol, virus or drugs other than ATT as a cause of abnormal baseline LFT. All the patients were recruited from the outpatient and inpatient sections of the Department of Internal Medicine and Gastroenterology, Jagjivanram Hospital and the Department of Gastroenterology, Bombay Hospital, Mumbai, India. Informed written consent was obtained from all the patients recruited in the present study.

1747 month, then fortnightly for the next 2 months and then monthly until the end of therapy; and (ii) whenever they developed symptoms of nausea, vomiting, anorexia, jaundice, malaise or lethargy. At each visit, patients were evaluated for compliance with ATT and alcohol abstinence, history suggestive of symptomatic hepatitis, clinical examination and LFT (serum bilirubin, AST and ALT).

Drug regimen
All the patients in Groups A and C were started on standard ATT; H 5 mg/kg/day up to a maximum of 300 mg/day PO for 612 months, R 10 mg/kg/day up to a maximum of 600 mg/day PO for 612 months, E 15 mg/kg/day up to a maximum of 1200 mg/day PO for an initial 2 months and Z 25 mg/kg/day up to a maximum of 2000 mg/day PO for an initial 2 months. In Groups B and A, when ATT-hepatotoxicity developed, drug regimens were modied. H, R and Z were withdrawn and E was continued. Ciprooxacin (C) 25 mg/ kg/day up to a maximum of 1500 mg in two divided doses PO and S 15 mg/kg/day up to a maximum of 1 g i.m. were added.1,5 Once liver function normalized on subsequent follow up, H 5 mg/kg/day followed by R 10 mg/kg/day followed by Z 25 mg/kg/day were added sequentially at weekly interval if LFT remained normal. After successful reintroduction of H, R and Z, C and S were omitted and then again the same monitoring procedure and reintroduction protocol were followed till end of the therapy. Patients with decompensation of underlying chronic liver disease were offered reintroduction of H only and not of R and Z.

Evaluation of patients
In the study groups at baseline, detailed history taking and clinical examination were carried out noting age, sex, height, weight, site of tuberculosis, daily intake of alcohol, drug intake, presence of diabetes, presence of symptoms suggestive of symptomatic hepatitis and previous/present history suggestive of decompensation of liver disease. In Groups A and C, baseline LFT including serum bilirubin, ALT, AST, alkaline phosphatase (ALP), gamma glutamyltransferase (GGT), albumin, globulin, prothrombin time (PT), renal function test and blood sugar test (fasting and postprandial, using an automated analyzer, Hitachi 902, Tokyo, Japan) as well as serology tests for hepatitis B surface antigen (Micro well ELISA, third generation, EQUIPAR srl, Saronno, Italy), anti-hepatitis C antibodies (Micro well ELISA, 3rd generation, Biochem Immunosystem, Canada) and HIV antibodies (Micro well ELISA, fourth generation, Biochem Immunosystem, Montreal, Canada) were carried out. Ultrasonography of the abdomen for evidence of chronic liver disease, upper gastrointestinal endoscopy for esophageal varices and liver biopsy for histology were carried out whenever necessary to dene chronic liver disease. In Groups B and A, when liver function tests were abnormal, additional serology tests for IgM Anti-hepatitis A virus (HAV) (Micro well ELISA, EQUIPAR srl), HBsAg, IgM anti-hepatitis B core antigen (HBc) (Micro well ELISA, EQUIPAR srl) and IgM anti-hepatitis E virus (HEV) (Micro well ELISA, EQUIPAR srl) were carried out. History of alcohol intake and exposure to hepatotoxic drugs was reviewed. Ultrasonography of the abdomen was carried out to eliminate biliary obstruction. In Group B, the schedule of previous ATT regimen and its duration were noted. All these patients were evaluated for evidence of coagulopathy and encephalopathy. Patients who showed evidence of coagulopathy and/or encephalopathy, or of decompensation of pre-existing cirrhosis and those requiring intravenous therapy were admitted to the hospital, ATT was modied as per protocol (wide below) and supportive management was provided. Once included in the study, all patients were asked to abstain from alcohol.

Outcome measures
During the follow-up in Groups A and B, the following parameters were noted: (i) the time period for the development of ATT-hepatotoxicity from the start of the ATT (latency period); (ii) the duration of symptoms before the detection of ATT-hepatotoxicity; (iii) the time period for normalization of LFT after withholding ATT (H, R and Z); and the requirement for hospitalization and Intensive Care Unit (ICU) care as well as mortality. Morbidity was dened as the presence of serious hepatotoxicity (FHF and decompensation) and the need for hospitalization or ICU. The success rate of sequential reintroduction of the standard ATT as per protocol was calculated.

Statistical analysis
Statistical analysis was carriedout applying Students ttest and Z-test. A P-value less than 0.05 was considered signicant. A Z-value greater than 1.96 was considered signicant at 95% condence interval.

Visit schedule and monitoring

Once included in the study, all the patients were reviewed: (i) as a schedule, every week for the rst

1748

S Agal et al. Twenty-seven patients (for Group B) directly presented as suspected cases of ATT-hepatotoxicity during the study period. Two patients were found to be suffering from acute hepatitis E and one had concomitant phenytoin therapy, a potentially hepatotoxic drug. Therefore, these three patients were excluded and the remaining 24 patients formed Group B of the present study (Fig. 2).

RESULTS
Of the 220 patients initially evaluated, 200 patients were found eligible to be included in the present study for prospective monitoring after starting ATT. Twenty patients were excluded from the study; 14 patients had abnormal baseline LFT (six had isolated hepatobiliary or miliary TB, ve had alcohol related liver dysfunction, one had chronic hepatitis B [HBV] infection and two had CCF) and six patients had a pre-existing immunocompromised condition (three had HIV infection with absolute CD4 lymphocyte count less then 200 cells/ cmm and three had immunosuppressive therapy after renal transplantation). Of the 200 patients on prospective monitoring, 21 developed ATT-toxicity (Group A) and 179 patients (Group C) never developed ATThepatotoxicity (Fig. 1).
220 TB patients initially evaluated

Baseline characteristics
Amongst demographic features and mean body mass index, there was no signicant difference between the study groups (Table 1). There was no statistically signicant difference in the prevalence of signicant alcohol intake, diabetes, HIV infection, chronic HBV infection and chronic hepatitis C infection (Table 1). Prevalence of pre-existing liver cirrhosis with normal baseline LFT in each group was 9.5% (two patients) in Group A, 20.8% (ve patients) in Group B and 8.3% (15 patients) in Group C (Table 1).

200 Eligible for the study

20 Excluded

21 Had hepatotoxicity (Group A)

179 Never had hepatotoxicity (Group C)

6 Immunocompromised: 3 HIV with CD4<200 3 Immunosuppressive drugs

27 Suspected hepatotoxicity

24 Eligible for the study (Group B)

3 Excluded: 2 Acute HEV infections 1 Phenytoin therapy

21 Reintroduction tried: 20 Successful 1 Rifampicin intolerant

14 Abnormal baseline LFT: 6 Miliary/hepatic TB 5 Alcoholic liver disease 1 Chronic HBV

18 Reintroductions carried out: 18 Successful

6 Reintroductions not carried out: 4 Deaths 2 Decompensated liver disease

2 CCF

Figure 1 Trial prole of patients who were prospectively monitored for antituberculosis therapy hepatotoxicity. CCF, congestive cardiac failure; HBV, hepatitis B virus; TB, tuberculosis. Table 1 Baseline characteristics of all the study groups Group A (n = 21) 37.9 10.6 1645 1:1.3 20.4 0.6 0 (0%) 0 (0%) 0 (0%) 2 (9.5%) 3 (14.2%) 2 (9.5%) 21 (100%)

Figure 2 Trial prole of patients who presented directly with suspected antituberculosis therapy hepatotoxicity. HEV, hepatitis E virus.

Baseline characteristics Age, mean SD (years) Range (years) Sex (male : female) Body mass index (kg/m2) HBsAg positive Anti-HCV positive HIV positive Diabetes Signicant alcohol intake Pre-existing cirrhosis Weight based drugs

Group B (n = 24) 41.2 12.5 567 1:1.4 19.8 0.3 2 (8.3%) 0 (0%) 1 (4.1%) 4 (16.6%) 3 (12.5%) 5 (20.8%) 12 (50%)

Group C (n = 179) 37.3 11.5 670 1.1:1 20.1 0.4 8 (4.4%) 0 (0%) 1 (0.5%) 15 (8.3%) 24 (13.4%) 15 (8.3%) 179 (100%)

Indicates signicant statistical difference between groups A and B as well as groups B and C. HBsAg, hepatitis B surface antigen; HCV, hepatitis C virus.

Management of ATT-hepatotoxicity There was no statistically signicant difference in the three groups in the site of tuberculosis. In Groups A, B and C, different sites of involvement were chest TB (52.3%, 62.5% and 51.3%), abdominal TB (28.5%, 20.8% and 31.9%), lymph node TB (9.5%, 4.1% and 3.3%), skin TB (0%, 0% and 1.1%) and neurological TB (4.7%, 8.3% and 1.1%). All the patients in study Group A and C had received weight based standard ATT as per protocol, whereas in Group B, only 12 patients (50%) had received weight based ATT regimens before the development of ATThepatotoxicity (Table 1).

1749 In the present study, asymptomatic transaminase elevation of 25 times the ULN was seen in 14 patients (7.8%) in Group C that resolved on continuation of the same ATT within 20.1 2.6 days (1025 days). In Group A, 14 patients (66.6%) experienced transaminase elevation more than ve times the ULN without any symptoms and were diagnosed as ATThepatotoxicity. Of those who were symptomatic (in Group A and B), 2/7 (28.6%) in Group A and 5/24 (20.8%) in Group B experienced a rise in transaminases between two and ve times the ULN and in the remaining patients, a rise in transaminases was more than ve times ULN (Table 2). Of those patients in Group B with icteric illness, 5/24 (20.8%) had an enzyme elevation less then ve times the ULN. Nineteen patients (79.1%) had an elevation of transaminases more than ve times the ULN.

Clinical features and laboratory features

In 200 patients (Group A + C), a total of 21 patients (10.5%) developed ATT-hepatotoxicity (Group A), of which seven (33.3%) had developed the symptoms of nausea (seven patients, 100%), vomiting (three patients, 42.8%) and anorexia (six patients, 85.7%). None of the patients developed icterus, encephalopathy, FHF or decompensation of liver cirrhosis (Table 2). Of the 24 patients in Group B, all were symptomatic (100%). Nausea was experienced by all the patients in Group B (100%), 22 (91.6%) experienced vomiting and all had anorexia (100%). Eighteen patients (75%) developed icterus and seven patients (29.1%) developed encephalopathy. Of these seven patients, three (12.5%) patients experienced FHF and four (16.6%) had decompensation of pre-existing liver cirrhosis (Table 2).
Table 2

Outcome and reintroduction

In Group A, the latency period for development of ATThepatotoxicity was 12.6 3.6 days (736 days), whereas in Group B it was signicantly more 29.8 5.3 days (7 65 days). The symptomatic period before detection of ATT-hepatotoxicity was signicantly less in Group A (2.3 0.7 days with range of 16 days) than in Group B (11.5 1.2 days with range of 715 days) (Table 3). In Group A, none of the patients required hospitalization or ICU stay and there was no mortality, whereas

Clinical and laboratory features of antituberculosis therapy hepatotoxicity in the study groups Group A (n = 21) 7 (33.3%) 0 (0%) 0 (0%) 0 (0%) 2/7 (28.6%) 5/7 (71.4%) Group B (n = 24) 24 (100%) 18 (75%) 7 (29.1%) 24 (100%) 5/24 (20.8%) 19/24 (79.2%) P or Z values S S S S NS NS

Clinical or laboratory features Symptomatic hepatitis Icteric hepatitis Serious hepatotoxicity (FHF, decompensation) Serum bilirubin >2 ULN Symptomatic hepatitis (a) Enzymes 25 ULN (b) Enzymes >5 ULN

FHF, fulminant hepatic failure; NS, not signicant; S, signicant; ULN, upper limit of normal. Table 3 Outcome and reintroduction in the study groups with antituberculosis therapy hepatotoxicity Group A (n = 21) 12.6 3.6 (730) 2.3 0.7 (16) 0 (0%) 0 (0%) 0 (0%) 21 (100%) 20.2 6.2 (740) 21 (100%) 20 (95.2%) Group B (n = 24) 29.8 5.3 (765) 11.5 1.2 (715) 10 (41.6%) 7 (29.1%) 4 (16.6%) 20 (83.3%) 24.8 5.4 (1078) 18 (90%) 18 (100%) P or Z values S S S S S NS NS NS NS

Outcome measures and reintroduction Latency period, days (range) Symptomatic period before detection of toxicity (range) Requirement of hospitalization Requirement of ICU Mortality Normalization of LFT Time for normalization of LFT after ATT withholding (range) Reintroduction attempted Successful reintroduction

ATT, antituberculosis therapy; ICU, Intensive Care Unit; LFT; liver function tests; NS, not signicant; S, signicant.

1750 in Group B a total of 10 patients (41.6%) were hospitalized, seven patients (29.1%) required intensive treatment in ICU (four with decompensation and three with FHF). Of these patients who required intensive treatment in ICU, four (16.6%) patients died (two from FHF and two from decompensation). So the mortality rate among patients who developed FHF was 2/3 (66.6%) and among those who developed decompensation, it was 50% (2/4). Morbidity and mortality were signicantly higher in Group B in the present study (Table 3). After withholding ATT, all the 21 patients in Group A had normalization of LFT after 20.2 6.2 days (740 days), whereas in Group B all patients who survived (20/24) had normalization of LFT within 24.8 5.4 days (1048 days) (Table 3). In all the 21 patients of Group A, sequential reintroduction of ATT was attempted. It was successful in 20 patients (95.2%). One patient was intolerant to R as demonstrated during deliberate re-reintroduction. In all the 20 survivors of Group B, reintroduction was not offered in those 2/20 patients who had decompensation of underlying chronic liver disease. In Group B, reintroduction was attempted in 18/20 (90%) patients; one patient with compensated chronic liver disease, one patient with ATT induced FHF and 16 other patients with no underlying chronic liver disease. Reintroduction of ATT was successful in all of these 18 patients (Table 3).

S Agal et al. a warning that hepatic necrosis leading to hepatic failure might develop subsequently.21 Therefore, frequent monitoring (probably weekly as in the present study) of transaminases in such patients might be necessary to detect early ATT-hepatotoxicity. In Group A, where regular LFT monitoring was carried out, ATT-hepatotoxicity was detected in the asymptomatic period in 66.6% patients. Icteric hepatitis did not develop, even in those 33.3% patients who had symptomatic hepatitis because of timely stoppage of hepatotoxic ATT. There was no occurrence of FHF, decompensation or deaths and there was no need for hospitalization/ICU care. In contrast to this, in Group B who presented directly with ATT-hepatotoxicity, all the patients had symptomatic hepatitis and of these patients, 75% had icteric hepatitis. In this group, a signicant proportion of patients required hospitalization (41.6%) and ICU care (29.1%). The mortality rate was 16.6%. Also, the latency period for development of ATT-hepatotoxicity was shorter in Group A than in Group B. Therefore, regular clinical and LFT monitoring was associated with better prognosis in the present study. Because ATT-hepatotoxicity is a result of metabolic idiosyncrasy, alarming symptoms suggestive of hypersensitivity such as fever, rash and itching are usually absent.2224 Non-specic symptoms of anicteric hepatitis, which precede icterus,9 are often reported late. Icteric hepatitis is associated with adverse prognosis and has a mortality rate of 420% in the absence of liver transplantation.8,17,18 Symptoms of hepatitis develop late in acute hepatic necrosis and are often accompanied by established liver failure.25 The factor of the greatest clinical importance for development of severe hepatotoxicity and mortality is probably continuation of ATT, once hepatic dysfunction has occured.23,26 Biochemical abnormalities generally occur well before clinical symptoms/signs of hepatic injury are obvious. Thus, biochemical monitoring can detect ATThepatotoxicity in the presymptomatic period and prevent adverse outcomes. In a study by Nolan et al. in H prophylaxis for latent tuberculous infection (LTBI), 9.1% patients had serious toxicity despite clinical monitoring, but no mortality.18 Byrd et al. have shown the absence of serious toxicity on routine laboratory monitoring in H prophylaxis for LTBI.27 McNeill et al. reduced the rate of serious toxicity from 5% to 0% with the routine application of laboratory monitoring of patients receiving R and Z for LTBI.28 Most hepatic toxicity reactions are evident in the rst month of treatment with a combination of R and H, but can occur up to the third month when Z is combined.3,29 The latency period for development of jaundice and subsequent FHF for the H and R combination can be 610 days.3,24 Therefore, in the present study, LFT monitoring was carried out every week for the rst month and then every fortnight for another 2 months of the treatment. With a proposed scheme of the sequential reintroduction of drugs in therapeutic doses at weekly intervals, reintroduction was successful in 38/39 patients (97.4%). Reintroduction of the ATT in therapeutic doses was based on the rationale that most ATThepatotoxicity is the result of metabolic idiosyncratic

DISCUSSION
In the present study, the incidence of ATThepatotoxicity was 10.5%, which is similar to the pooled data from four prospective Indian studies (11.5%).15 There is a wide variability in reported incidence of ATThepatotoxicity from different countries (136%): 3% in the USA,9 4% in the UK,9 11% in Germany,13 9.9% in Argentina,13 13% in Hong Kong,16 36% in Japan,16 26% in Taiwan16 and 836% in India.15 This higher incidence of ATT-hepatotoxicity in the Asian countries might be the result of ethnic susceptibility,16 inherent peculiarity of drug metabolism and/or the presence of various known risk factors such as HBV infection17,18 or malnutrition.18,19 Also, the lack of uniformity in the denition of ATT-hepatotoxicity, as well as in the use of ATT regimens in various studies, might be responsible for this difference.16 In the present study, 7.8% of the patients who never experienced ATT-hepatotoxicity had transient transaminase elevation. The presence of such asymptomatic transaminase elevation is well documented in the studies from all over the world and there is a prevalence of up to 20%.4,5,9 This phenomenon of transient transaminase elevation, which normalizes on continuing the offending drugs, is referred to as drug tolerance or adaptation. This probably occurs as a result of the effect of immunosuppressive cytokines on T cells that inhibit progression of immune reaction to any drug.20 In 5 20% of patients who initially experience asymptomatic elevation, a progressive increase in transaminases of more than ve times the ULN is seen and this might be

Management of ATT-hepatotoxicity reactions and is not dose-dependent. This was in contrast to the guidelines given by the BTS, where sequential reintroduction of gradually increasing doses at intervals of 23 days was suggested.4,9 The period of 2 3 days was considered insufcient for the identication of the offending drug.25 The lack of recurrence of ATThepatotoxicity on the reintroduction might be suggestive of the phenomenon of metabolic adaptation.12 Potent induction of the cytochrome p450 (CYP) enzyme system by R is thought to be responsible for enhanced hepatotoxicity of the combination of H and R, but chronic CYP2E1 over-expression results in an adaptive response in the hepatocytes leading to sustained signal-regulated kinase (ERK) activation mediated by EGFR/c-Raf signaling. This results in desensitization and subsequent down-regulation of an acute cellular response and prevention of cellular death from an additional acute oxidative stress.30 In worldwide studies, different reintroduction regimens have been advocated with variable success, including regimens avoiding Z, regimens avoiding R, low dose escalating regimens or alternative non-hepatotoxic ATT regimens.13,24,31 Also, the role of frequent monitoring of LFT after full reintroduction of standard ATT is suggested, as recurrence of hepatotoxicity might be higher in incidence, earlier in occurrence and more severe.8,32 In the present study, drugs such as E, C and S were used during the reintroduction period as an alternative non-hepatotoxic ATT to prevent the problem of drug resistance and to control the disease.5,9,13 In the present study, two patients who experienced serious ATT-hepatotoxicity-like FHF had successful reintroduction. Reintroduction in the setting of fulminant hepatitis is usually considered unsafe.25 With the help of close monitoring of symptoms and LFT in the present study, it was possible in these patients to reintroduce antitubercular therapy. In conclusion, monitoring symptoms and LFT at regular intervals is important at least in the initial month to pick up early ATT-hepatotoxicity and to prevent ATThepatotoxicity related deaths. As primary drugs like H, R and Z are the most effective drugs, sequential reintroduction of these drugs on weight-based regimen should be carried out in all ATT-hepatotoxicity patients with close clinical and LFT monitoring.

1751
2 Nishioka SA. Antituberculous drugs and hepatotoxicity. Am. J. Gastroenterol. 1996; 91: 14712. 3 Durand F, Bernuau J, Pessayre D et al. Deleterious inuence of pyrazinamide on the outcome of patients with fulminant or subfulminant liver failure during antituberculous treatment including isoniazid. Hepatology 1995; 21: 92932. 4 Joint Tuberculosis Committee of the British Thoracic Society. Chemotherapy and management of tuberculosis in the United Kingdom; recommendations 1998. Thorax 1998; 53: 53648. 5 American Thoracic Society/Centers of Disease Control and Prevention/Infectious Disease Society of America. Treatment of tuberculosis. Am. J. Respir. Crit. Care Med. 2003; 167: 60362. 6 CDC. Update: fatal and severe liver injuries associated with rifampin and pyrazinamide for latent tuberculosis infection and revisions in American Thoracic Society/ CDC recommendations-United States, 2001. MMWR 2001; 50: 7335. 7 Durand F, Jebrak G, Pessayre D, Fournier M, Bernuau J. Hepatotoxicity of antitubercular treatments. Rationale for monitoring liver status. Drug Saf. 1996; 15: 394405. 8 Mitchell I, Wendon J, Fitt S, Williams R. Antituberculous therapy and acute liver failure. Lancet 1995; 345: 5556. 9 Ormerod LP, Skinner C, Wales JM. Hepatotoxicity of antituberculosis drugs. Thorax 1996; 51: 11113. 10 Nolan CM, Goldberg SV, Buskin SE. Hepatotoxicity associated with isoniazid preventive therapy: a 7-year survey from a public health tuberculosis clinic. JAMA 1999; 281: 101418. 11 Wada M, Yoshiyama T, Ogata H, Ito K, Mizutani S, Sugita H. Six-month chemotherapy (2HRZS or E/4HRE) of new cases of pulmonary tuberculosis six year experiences on its effectiveness, toxicity and acceptability. Kekkaku 1999; 74: 35360. 12 Danielides IC, Constantoulakis M, Daikos GK. Hepatitis on high dose isoniazid: reintroduction of the drug in severe tuberculous meningitis. Am. J. Gastroenterol. 1983; 78: 37880. 13 Garg PK, Tandon RK. Antituberculosis treatment induced hepatotoxicity. In: Sharma SK, Mohan A, eds. Tuberculosis, 1st edn. New Delhi: Jaypee Brothers Medical Publishers, 2001; 499506. 14 Sherlock S, Dooley J, eds. Diseases of the Liver and Biliary System, 11th edn. Oxford: Blackwell Science, 2002. 15 Pande JN, Singh SPN, Khilnani GC, Khilnani S, Tandon RK. Risk factors for hepatotoxicity from antituberculosis drugs: a case-control study. Thorax 1996; 51: 1326. 16 Huang Y-S, Chern H-D, Su W-J et al. Polymorphism of the N-acetyltransferase 2 gene as a susceptibility risk factor for antituberculosis drug-induced hepatitis. Hepatology 2002; 35: 8839. 17 Lewis JH. Drug-induced liver disease. Med. Clin. North Am. 2000; 84: 1275311. 18 Singh J, Arora A, Garg PK, Thakur VS, Pande JN, Tandon RK. Antituberculosis treatment-induced hepatotoxicity: role of predictive factors. Postgrad. Med. J. 1995; 71: 35962. 19 Singh J, Garg PK, Tandon RK. Hepatotoxicity due to antituberculosis therapy. Clinical prole and reintroduction of therapy. J. Clin. Gastroenterol. 1996; 22: 21114.

ACKNOWLEDGMENTS
We thank all the trial participants for their consent and participation in the study. We thank authorities of Jagjivanram Western Railway Hospital and Bombay Hospital, Mumbai for their kind permission to conduct the study and publish the results. We also thank the staff of the Gastroenterology Department and Laboratory of both the hospitals.

REFERENCES
1 Maher D, Chaulet P, Spinaci S, Harries A. Treatment of tuberculosis: guidelines for national programs. WHO Global tuberculosis programme. WHO/TB/97,220.

1752
20 Marino G, Zimmerman HJ, Lewis JH. Management of drug-induced liver disease. Curr. Gastroenterol. Rep. 2001; 3: 3848. 21 Yamamoto T, Suou T, Hirayama C. Elevated serum aminotransferase induced by isonizid in relation to isoniazid acetylator phenotype. Hepatology 1986; 6: 2958. 22 Steele MA, Burk RF, DesPrez RM. Toxic hepatitis with isoniazid and rifampin: a meta-analysis. Chest 1991; 99: 46571. 23 Lee WM. Drug induced hepatotoxicity. N. Engl. J. Med. 1995; 333: 111827. 24 Pessayre D, Bentata M, Degott C et al. IsoniazidRifampin fulminant hepatitis. A possible consequence of the enhancement of isoniazid hepatotoxicity by enzyme induction. Gastroenterology 1977; 72: 2849. 25 Delvin J, Hutchison DCS, Fitt S, Wendon J, Williams R. Hepatotoxicity of antituberculosis drugs. Thorax 1996; 51: 116870. 26 Moitinho E, Salmeron JM, Mas A, Bruguera M, Rodes J. Severe hepatotoxicity of tuberculostatic agents: increase in the incidence. Gastroenterol. Hepatol. 1996; 19: 44851.

S Agal et al.
27 Byrd RB, Horn BR, Solomon DA, Griggs GA. Toxic effects of isoniazid in tuberculosis chemoprophylaxis: role of biochemical monitoring in 1000 patients. JAMA 1979; 241: 123941. 28 McNeill L, Allen M, Estrada C, Cook P. Pyrazinamide and rifampin vs. isoniazid for the treatment of latent tuberculosis: improved completion rates but more hepatotoxicity. Chest 2003; 123: 1026. 29 Kimmoun E, Samuel D. Antituberculous drugs in patients with chronic liver disease. J. Gestroenterol. Hepatol. 2002; 17: S40812. 30 Schattenberg JM, Wang Y, Rigoli RM, Koop DR, Czaja MJ. CYP2E1 overexpression alters hepatocytes death from Menadione and fatty acids by activation of ERK1/2 signaling. Hepatology 2004; 39: 44455. 31 Girling DJ. The hepatic toxicity of antituberculosis regimens containing isoniazid, rifampicin and pyrazinamide. Tubercle 1978; 59: 1332. 32 Tahaoglu K, Atac G, Sevim T et al. The management of anti-tuberculosis drug-induced hepatotoxicity. Int. J. Tuberc. Lung Dis. 2001; 5: 659.