Background:  In 1998, despite concerns expressed by several decision-makers and financing agencies, Senegal was one of the first countries in Sub-Saharan Africa to launch a governmental program for ART access. To address the various scientific and practical issues related to such a program, an observational cohort was initiated. With a 10-year follow-up, this cohort contributes to a better knowledge of the short- and long-term evolution with ART.

Methods:  Between August 1998 and April 2002, 404 volunteers were included. The initial ART regimen was a triple drug combination. After comprehensive clinical and biological assessments at inclusion, patients were examined at least every 2 months and had a biological evaluation every 6 months. Several endpoints were examined, including time to first undetectable viral load, time to first virological failure (viral load >1000 copies/mL), time to death, CD4 increase over time and adherence (assessed in a subset, n = 158).

Results:  The total follow-up accrued to 2157 person-years. At baseline, more than half of the patients were CDC C stage and the median CD4 cell count was 128 cells/mm³ showing advanced stages of disease. The median CD4 cell count increased to 286 cells/mm³ after 1 year and reached 455 cells/mm³ after 6 years. Most of the 111 deaths recorded occurred early after HAART initiation; the mortality rate decreased sharply from 12.8 (9.7 to 17) per 100 person-years over the first year to 1.7 (0.95 to 3) after 5 years of follow-up. However, a significant increase in the mortality rate after 7 years on HAART was noted. Most of the patients reached undetectability (91%), but among them 130 (40%) virological failures were observed. Virological failure was associated with an initial PI-containing regimen (HR = 1.96; 1.38 to 2.77) and a younger age (HR = 1.27; 1.04 to 1.56). The estimated proportion of patients with virological failure who developed resistances to drugs ranged between 50% and 75%. The average adherence was greater than 90% and was initially negatively associated with an PI-containing regimen (OR = 0.32; 0.22 to 0.46). The prevalence of lipodystrophy and metabolic syndrome were rather low (31% and 14.8%).

Conclusions:  Through its long duration of follow-up, this cohort helps to identify several critical aspects of the evolution after HAART initiation in Africa:  an initial high mortality rate, and concerns about an increase in late mortality rate.