Abstract
AIMS: To screen and validate rare variants in the protoporphyrin IX (PPIX)-metabolizing pathway associated with anti-tuberculosis drug-induced hepatitis (AT-DIH). METHODS: A two-stage matched case-control study was conducted. Firstly, a 1:1 matching design (50 cases and 50 controls) was adopted to screen for rare variants using whole exome sequencing (WES). Secondly, a 1:4 matching design (132 cases and 528 controls) was employed to validate these rare variants via TaqMan genotyping. The association between these variants and AT-DIH risk was evaluated using odds ratios (ORs) with 95% confidence intervals (CIs), with the false discovery rate (FDR) applied for multiple comparison correction. RESULTS: The optimal sequence kernel association test (SKAT-O) and set-based tests identified 19 and 1 genes significantly associated with AT-DIH, respectively. The nuclear receptor coactivator 3 (NCOA3) gene was detected by both methods (SKAT-O: p = 0.002; set-based tests: p = 0.038), and seven algorithms identified four NCOA3 rare variants as deleterious. Further validation showed that the CC genotype of rs2230782 increased the risk of AT-DIH (OR = 15.074, 95%CI: 1.442-157.525, FDR p = 0.046), and this association remained significant in the two-stage combined analysis (3 cases and 1 control with CC genotype, OR = 14.832, 95%CI: 1.422-154.695, FDR p = 0.048). CONCLUSIONS: NCOA3 rs2230782 is associated with AT-DIH in Chinese anti-tuberculosis patients.