Abstract
OBJECTIVES: Estimate tuberculosis (TB) incidence among patients receiving HAART. Compare the dynamic of the CD4-cell count and viral load before notification of the TB with the dynamic among patients remaining free of TB. DESIGN: Prospective cohort with ascertainment of TB cases from medical records. METHODS: The first 404 adults HIV-1 infected patients enrolled in the Senegalese antiretroviral drug access initiative were eligible. CD4-cell and viral load were assessed at baseline and every 6 months. Patients receiving an antituberculosis treatment at HAART initiation were excluded from analysis. Any TB case notified after the first month of HAART was considered as an incident case. Follow-up was censored at death or at the last visit before March 31, 2008. CD4-cell trajectories until TB notification were compared to non-TB developers within two distinct periods: from HAART initiation to 24 months and after. RESULTS: Over 404 eligible patients, 352 were included in this analysis. Median follow-up reached 73 months and 1821 person-years were accrued. Half of the 42 incident cases were notified before month 19 of HAART yielding to an overall incident rate of 2.3/100 PY [1.7-3.1]. Annual incidence decreased with duration of HAART (trend in incidence: RR=0.26, p<10(-4)). During the first period, CD4-cell count dynamic of most TB patients was identical to the dynamic among patients remaining free of TB. Most cases of the second period occurred in a context of an immunological failure. CONCLUSIONS: This study provides an estimate of TB incidence among patients on HAART in Senegal and supports two underlying mechanisms.