[Drug-induced hepatotoxicity caused by anti-tuberculosis drugs in tuberculosis patients complicated with chronic hepatitis].

OBJECTIVES

To investigate retrospectively the incidence of drug-induced hepatitis (DIH) caused by antituberculosis drugs including isoniazid (INH), rifampicin (RFP), with and without pyrazinamide (PZA), and to evaluate risk factors for DIH in tuberculosis patients complicated with chronic hepatitis (CH).

MATERIALS

One hundred and seven tuberculosis patients with CH (M/F= 96/11, mean age +/- SE, 60.8 +/- 1.4 yr) admitted to our hospital during 1998-2006, whose laboratory data had been followed before and at least 2 months after starting antituberculosis chemotherapy, were enrolled in this study. Of these, 58 were being treated with anti-tuberculosis chemotherapy consisting of INH, RFP and PZA (HRZ group) and the remaining 49 with INH and RFP (HR group). For a case-control study, patients admitted to the hospital during the same period and without CH were selected to each CH patient (n=107) of the same gender, the same treatment regimens, and the same age. Clinical diagnosis of CH was based on laboratory data and in some cases pathological findings; etiology of CH was C-CH (CH caused by hepatitis C virus) in 68 patients, B-CH (CH caused by hepatitis B virus) in 23, and alcoholic CH in 16.

METHODS

DIH was defined by elevation of serum aspartate aminotransferase (AST) or alanine aminotransferase (ALT) at 1 or 2 months after starting anti-tuberculosis chemotherapy. For patients with serum levels of AST or ALT already abnormally high before starting chemotherapy, an increase of > 1.5 times from the initial serum level was defined to indicate DIH, whereas for patients with AST and ALT within the normal range, and increase of > 3X the normal upper limit was defined to indicate DIH. The incidence of DIH was calculated separately in the groups HRZ and HR for patients with and patients without CH (control). In the HRZ group, the severity of DIH was defined by the maximum serum levels of AST and ALT, and their mean values were compared between CH patients and the control. Risk factors for DIH were examined by comparing patients with and without CH. The clinical course after development of DIH was also followed. [Results] The incidence of DIH in the HRZ group was 13/ 58 (22.4%) for CH patients and 10/36 (27.8%), 2/13 (15.4%) and 1/9 (11.1%) for C-CH, B-CH and alcoholic hepatitis patients, respectively, which was significantly (p < 0.05) higher than that in the control [4/58 (6.9%)]. Confining to the C-CH patients, the incidence of DIH was 10/36 (27.8%) compared with the control 2/36 (5.6%) (p < 0.05). In contrast, the incidence of DIH in the HR group was not significantly different between CH patients and the control, [2/49 (4.1%) vs 2/49 (4.1%)], respectively. The severity of DIH in the HRZ group estimated by the maximum level of serum AST and ALT was not significantly different in CH patients and the control (176.6 +/- 28.1 vs. 311.0 +/- 154.5 IU/L for AST and 187.8 +/- 19.1 vs. 277.8 +/- 72.4 IU/L for ALT). Of the 13 CH patients suffering from DIH caused by antituberculosis chemotherapy containing INH, RFP and PZA, 3 were continued treatment without altering the regimen, and 9 were continued treatment after changing the regimen to INH and RFP, omitting PZA. The one remaining patient was re-treated using INH, RFP and ethambutol (EB), but this again resulted in development of DIH, and he was ultimately treated with INH, EB and levofloxacin, with a successful outcome. Thus, at least 12 out of the 13 CH patients who developed DIH in the HRZ group could be treated by an anti-tuberculosis chemotherapy regimen containing INH and RFP excluding PZA. In C-CH patients who were treated with INH, RFP and PZA, the incidence of DIH was significantly higher when the daily alcohol intake was >20 g [8/18 (44.4%)] compared with those <20 g [0/10 (0%)] (p < 0.05), indicating that alcohol is a risk factor for DIH in C-CH patients treated with INH, RFP and PZA.

CONCLUSIONS

In CH patients, anti-tuberculosis chemotherapy containing INH and RFP without PZA can be used safely. The inclusion of PZA in the regimen does substantially increase the incidence of DIH but nonetheless it can be used with caution, especially bearing in mind that daily alcohol intake of >20 g is a significant risk factor for C-CH patients.