Background: Universal access to antiretroviral treatment (ART) is still elusive in most developing nations. We report on the effectiveness of the current ART scale-up campaign in Kenya and discuss factors influencing treatment outcomes.
Methods: A multi-center longitudinal and cross-sectional survey of viral load (VL), CD4 T-cells, drug resistance and adherence. VL, CD4 counts and drug resistant mutations were determined using m2000 Abbott HIV-1 assay, FACS and Pol-RT sequencing respectively. Adherence was scored as good, fair or poor based on number of missed doses.
Results: Overall, 35.9% of the 546 patients failed treatment using longitudinal multiple viral load (VL), and 22% to 29% failed using cross-sectional single-VL definitions. More patients (41%) starting first-line D4T+3TC+NVP/EFV failed treatment than those initiating TDF+3TC+NVP/EFV (29%) (P=0.043). Female patients had higher CD4 counts, lower VL, better adherence and significantly less ART failure than males. Using Chi-Square test, the cross-sectional criteria defined failure with 99% to 93% accuracy of the longitudinal VL approach (p<0.001). Patients switched regimen without necessarily failing first-line, and 26% of those switching still failed second-line. Up to 33% of the patients had at least two major drug resistance mutations of NRTI or NNRTI type. Community Peer Support Network (CPSN) activity was significantly associated with improved adherence, reduced VL and reduced treatment failure rates. Patients with multiple active sex partners also had higher VL, but these were not significantly different between independent groups.
Conclusions: D4T-containing primary regimen and weak adherence independently correlate with increased ART failure. A single VL test after 12 months of uninterrupted ART offers an effective alternative to ART failure definition under limited resources. We recommend point-of-care VL testing at least once annually, drug resistance monitoring and peer focused adherence-support to inform treatment and mitigate failure.