Nevirapine (NVP)-based antiretroviral therapy (ART) is one of the limited options in children younger than 3 years old with TB/HIV coinfection. Given the scarce data, we examined the effect of first-line TB therapy on NVP pharmacokinetics (PK) in Ghanaian children.
ART-naïve HIV-infected children aged 3–35 months with and without TB were treated with NVP 200 mg/m2 twice daily plus two NRTIs. The new WHO recommended higher dosages of rifampin and isoniazid were used. After 4 weeks of ART, PK samples were collected at 0, 2, 6, and 12 hours post-dose to measure NVP plasma concentrations, using a validated LC/MS/MS assay. In the co-infected patients, sampling was repeated after 4 weeks off TB therapy. PK parameters were calculated using noncompartmental analysis and were compared between groups using Wilcoxon Rank-sum test and within group using Signed-rank test.
Of the 53 patients, 23 (43%) had TB coinfection, of whom 15 completed PK sampling on (PK1) and off (PK2) anti-TB therapy. Baseline characteristics were similar in the two groups except co-infected children had lower median height-for-age-Z-score. Median NVP concentrations were lowest in the children with TB/HIV coinfection on TB therapy, followed by HIV infection only and highest in the co-infected off TB therapy (Figure). Median NVP Cmax, Cmin and AUC0-12h were not significantly different between children with HIV and those with TB/HIV on or off anti-TB therapy. In multivariate analysis, TB therapy and CYP2B6 516G>T genotype status were joint predictors of NVP PK. Among children with CYP2B6 516GG genotype, NVP exposure was significantly lower in the TB co-infected compared to HIV-infected group; this difference was not seen in children with GT or TT genotypes. The proportion of children with NVP Cmin <3 mg/L was 61% in the co-infected group and 30% in the HIV group (P = 0.03). Among the TB/HIV co-infected children with paired samples, geometric mean ratio (90% CI) PK1/PK2 of NVP Cmax, Cmin and AUC0-12h were 0.68 (0.55–0.85), 0.84 (0.51–1.38) and 0.71 (0.56–0.91). Nine (41%) of 22 children with viral load data at 6 months had HIV RNA >200 copies/mL.
First-line TB therapy reduced NVP plasma exposure in young HIV-infected children, especially those with CYP2B6 516GG genotype. Given that NVP dose optimization with TB therapy may require a genotype-guided approach, evaluation of more compatible alternatives to NVP is needed in young children with TB coinfection.