We read with interest the work of Lucena et al. about genetic susceptibility to drug-induced liver injury (DILI). 1 There is an interest in studying the possible relationship between glutathione S-transferase (GST) at the M1 and T1 loci and the risk of hepatotoxicity secondary to drugs. The study of these genetic polymorphisms should be considered a target in the investigation of drugrelated hepatotoxicity mediated by metabolic idiosyncratic mechanisms. [2][3][4] However, in many cases, the drug-metabolism pathways and the exact mechanism and factors contributing to liver toxicity remain poorly understood. From pharmacological studies, there are data showing that GST genotypes may be linked to DILI secondary to hepatocellular damage mechanisms based on the participation of GST enzymes in drug metabolism. 4 Thus, studying drugs in which the DILI mechanism is unknown or secondary to a hypersensitive reaction could be confusing. Most of these investigations have been performed in Asian patients exposed to antituberculosis drugs; it is known that GST enzymes play an important role in isoniazid metabolism. 5-7 Lucena et al. included patients with DILI in whom the main causative therapeutic group of drugs were anti-infectives, especially amoxicillin-clavulanic acid. Among the group who received anti-infective drugs, only five patients were receiving antituberculosis drugs, which were not specified by the authors. Among these five patients, four showed one of these null polymorphisms (GSTM1 or GSTT1) and one had a normal genotype, but the authors did not specify the combination of the null alleles or in which genes.The absence of this data is general for all types of drugs studied; there is no information about what gene was predominantly deleted for each drug. Again, focusing on the antituberculosis drug group, it is not surprising there was no association between the presence of GSTM1 and GSTT1-null mutations and the risk of DILI due to the small number of patients in this subgroup. Our group performed the first study with 95 Caucasian patients originally from Spain to investigate the influence of these null polymorphisms on the risk of antituberculosis drug-induced hepatotoxicity and showed an enhanced risk of DILI in GSTT1-null carriers (odds ratio ϭ 2.6; P ϭ 0.03). 8 The frequency of double GSTM1 and GSTT1 genotype carriers was higher in cases than in controls; however, these differences did not reach significance because the power of the associations related to the small size of the patient subgroup. The study by Lucena et al. corroborated this observation. 1 We also observed a possible association between the risk of severe hepatotoxicity and the presence of the GSTT1-null polymorphism among six patients with severe hepatotoxicity; five (83.3%) presented with the GSTT1-null polymorphism (P ϭ 0.08). Moreover, the maximum peak of alanine aminotransferase was higher among patients with the GSTT1-null polymorphism {342 IU/L [Interquartile Range (IQR) 167-755] versus 216 IU/L [IQR 150-271]}, although without signific...