Prevalence of potential drug-drug interactions involving antiretroviral drugs in a large Kenyan cohort.

backgroundClinically significant drug-drug interactions (CSDIs) involving antiretrovirals are frequent and under-recognized in developed countries, but data are lacking for developing countries. METHODOLOGY AND PRINCIPAL

findingsTo investigate the prevalence of CSDIs between antiretrovirals and coadministered drugs, we surveyed prescriptions dispensed in a large HIV clinic in Kenya. Of 1040 consecutive patients screened, 996 were eligible for inclusion. CSDIs were defined as 'major' (capable of causing severe or permanent damage, contraindicated, avoid or not recommended by the manufacturer, or requiring dose modification) or 'moderate' (manufacturers advise caution, or close monitoring, or capable of causing clinical deterioration). A total of 334 patients (33.5%) were at risk for a CSDI, potentially lowering antiretroviral drug concentrations in 120 (12%) patients. Major interactions most frequently involved rifampicin (12.4%, mostly with efavirenz) and azoles (2.7%) whereas moderate interactions were frequently azoles (13%), steroids (11%), and antimalarials (3%). Multivariable analyses suggested that patients at risk for CSDIs had lower CD4 counts (P = 0.006) and baseline weight (P = 0.023) and WHO Stage 3 or 4 disease (P≤0.007). Risk for CSDIs was not associated with particular regimens, although only 116 (11.6%) patients were receiving WHO second line regimens.

conclusionsOne in three patients receiving antiretrovirals in our programme were at risk of CSDIs. Strategies need to be urgently developed to avoid important drug interactions, to identify early markers of toxicity and to manage unavoidable interactions safely in order to reduce risk of harm, and to maximize the effectiveness of mass antiretroviral deployment in Africa.