Low Serum Concentrations of Rifampicin and Pyrazinamide Associated with Poor Treatment Outcomes in Children with Tuberculosis Related to HIV Status.

OBJECTIVES

To compare the pharmacokinetics of rifampicin (RMP), isoniazid (INH) and pyrazinamide (PZA) between HIV-infected and HIV-uninfected children with tuberculosis (TB) and correlate it with TB treatment outcome.

METHODS

HIV-uninfected (n = 84) and HIV-infected (n = 77) children with TB receiving standard thrice weekly treatment were recruited from 6 hospitals in India. Semi-intensive pharmacokinetic sampling was performed during intensive phase of TB treatment after directly observed administration of drugs. Drug concentrations were measured by high performance liquid chromatography. INH acetylator status was determined, and nutritional assessment was done. Children were followed-up and treatment outcomes noted.

RESULTS

Children with HIV and TB had significantly lower RMP peak concentration (Cmax) (2.6 vs. 5.1 μg/mL; P < 0.001) and exposure [area under the time-concentration curve (AUC0-8); 10.4 vs. 23.4 μg/mL h; P < 0.001] than those with TB. Among HIV-infected children, a significantly higher proportion had stunting (77% vs. 29%; P < 0.001) and underweight (73% vs. 38%; P < 0.001) compared with children with TB. Combining both groups, RMP Cmax (P = 0.001; adjusted odds ratio = 1.437; 95% confidence interval: 1.157-1.784) and PZA Cmax (P = 0.027; adjusted odds ratio = 1.041; 95% confidence interval: 1.005-1.079) significantly influenced treatment outcome.

CONCLUSIONS

HIV infection was associated with lower Cmax of RMP and INH and AUC0-8 of RMP. Over 90% of children in both groups had subtherapeutic RMP Cmax. Cmax of RMP and PZA significantly influenced TB treatment outcome in children with TB. The findings have important clinical implications and suggest the need to increase anti-TB drug doses in children with HIV and TB.