To the Editor, Previous reports have shown that polymorphisms in NAT2 and CYP450 2E1 genes contribute to antituberculosis drug (ATD)-induced hepatotoxicity in adults in different populations.1–3 In the present study, the association between risk of hepatotoxicity and polymorphisms in these genes was examined in children. The study involved 111 pediatric patients with tuberculosis (TB). Most of the patients were treated with isoniazid (INH; 5 mg/kg bodyweight per day), rifampicin (10 mg/kg bodyweight per day) and pyrazinamide (20–35 mg/kg bodyweight per day). A few patients with TB meningitis also received ethambutol (20 mg/kg bodyweight per day). After 2 months, pyrazinamide and ethambutol were discontinued and INH and rifampicin were continued for the next 4 months. Of the 111 patients, nine showed symptoms of hepatotoxicity 1–6 weeks after commencing ATD treatment. One girl with hepatotoxicity had a positive test for hepatitis B virus surface antigen, so this patient was excluded from the study. The remaining 102 patients were considered non-toxic controls. However, subsequently during the course of treatment one non-toxic control was found to be infected with HIV, so this patient was also excluded from the study. Therefore, eight of the total 109 (7%) pediatric patients with TB developed hepatotoxicity in this study (Table 1). Although children of both sexes had TB, all patients who developed hepatotoxicity were boys. The absence of any girls who developed hepatotoxicity in this study may indicate sex specificity for the development of ATD-induced hepatotoxicity in children, or it may be an effect of the small sample size. A report on Japanese pediatric patients also observed that more boys had hepatotoxicity than girls;4 however, this finding should be confirmed with a larger sample size. Total bilirubin contents and activities of serum alanine aminotransferase, aspartate aminotransferase and alkaline phosphatase (ALP) were significantly higher in hepatotoxic patients than in non-toxic controls (P < 0.0001). As cases and controls were similar in mean ages and treated with the same regimen of ATD, it is highly probable that hepatotoxicity was manifested in some patients due to genetic susceptibility. N-acetyl transferase 2 (NAT2) is expressed predominantly in the liver and is involved in the generation of precursors of hepatotoxins (such as acetyl hydrazine) in INH hepatotoxicity. Polymorphisms at KpnI, TaqI and BamHI sites on the NAT2 exon could modulate the activity of the NAT2 enzyme and hence increase the risk of hepatotoxicity. In slow acetylators, acetyl hydrazine might accumulate in the liver as a source of hepatotoxin, because conversion of monoacetyl to diacetyl hydrazine, which is probably non-toxic, is lower in these individuals. A few reports, based on NAT2 genotyping, have also shown that slow acetylators are susceptible to ATD-induced hepatotoxicity in different adult populations.1,2 However, in the present study, genotyping data revealed that the proportions of NAT2 slow acetylators were similar in the hepatotoxic cases and in the non-toxic controls (63% and 56%, respectively). As the sample size was low, the effect of acetylation status on the risk of hepatotoxicity may not have been able to be observed in this study. In an Indian population, the majority of the controls were slow acetylators (56%), whereas the majority of controls were fast/rapid acetylators (90% and 78%, respectively) in Japanese and Taiwanese populations.1,2 So, apart from low sample size, differences in age (adult versus children) and proportions of slow acetylators might have contributed to the differences in results obtained in this and other studies. Several phenotypic studies also demonstrated that the acetylation status of patients might not have a significant role in the development of ATD-induced hepatotoxicity.5 Acetyl hydrazine, one of the INH metabolites, could be oxidized by the CYP2E1 enzyme into ultimate hepatotoxins, such as acetyldiazene, ketene and acetylonium ion. Several studies have reported that the variant 'c2', C and I alleles (at PstI-RsaI, DraI and 96 bp indel polymorphic sites, respectively) on CYP2E1 are associated with enhanced enzyme activity. In this study, frequencies of variant C/C and I/I genotypes in hepatotoxic cases were significantly higher than non-toxic controls (OR = 11.0, 95% CI = 1.02–110; same value for both genotypes, Table 2). So, it can be interpreted that variant C/C and I/I genotypes could increase the production of hepatotoxins and hence the risk of hepatotoxicity. The frequency of C allele was over-represented in cases compared to controls (P = 0.046); however, those of I alleles were not significantly different in cases and controls (P = 0.2). Because alleles at PstI and RsaI sites are in strong linkage disequilibria, polymorphisms at these sites could be expressed as c1 and c2 alleles. The variant c2/c2 genotype was absent in cases and controls. Although 13% of cases and 2% of controls had the heterozygous c1/c2 genotype, they were not significantly different. In Taiwanese adult hepatotoxic patients, the c1/c1 genotype was over-represented;3 however, this was not found in the present pediatric population. The differences in the frequencies of wild-type c1 and variant c2 alleles in different populations (0.76 and 0.24 in Taiwanese and 0.99 and 0.01 in Indian populations) might have contributed to this difference. Although polymorphisms at PstI and RsaI sites did not increase the risk of hepatotoxicity, polymorphisms at DraI and indel sites on the same gene did increase the risk of hepatotoxicity. The c1-I-C and c1-D-D haplotypes, estimated from genotypes at PstI-RsaI, indel and DraI sites, were observed to be major haplotypes present in both cases and controls. However, the c1-I-C haplotype increased the risk of hepatotoxicity (OR = 4.6, 95% CI = 1.3–16.3, data not shown). This haplotype contains the variant I and C alleles and the wild-type c1 allele, which also corroborate the result that showed c1/c1 genotype increased the risk of hepatotoxicity.3 Currently haplotype rather than genotype is considered as the functional unit in determining susceptibility to a disease. In this context, c1-I-C haplotype on CYP2E1 could be a marker for ATD-induced hepatotoxicity. So, pathways of hepatotoxicity development in children (this study) and adults (in other studies) may have a commonality such as involvement of the CYP2E1 gene. In conclusion, this study found that variant genotypes (C/C and I/I) and haplotype (c1-I-C) on CYP2E1 increased the risk of hepatotoxicity in pediatric patients. However, a study with larger sample sizes will provide a better basis on which to predict ATD-induced hepatotoxicity in different populations. The authors are grateful for the cooperation received from the patients. Logistic support from Professor PP Majumder, Indian Statistical Institute, and Dr Meena Basak, Principal of Dr BC Roy Memorial Hospital for Children, is also highly appreciated. This work was supported by a grant from Indian Council of Medical Research, Government of India, 2002–2004.
No takes yet. Share an insight, caveat, or question.
Roy et al. (2006) studied this question.
Synapse has enriched 5 related papers. Consider them for comparative context: