Home Search Abstracts View Session E-mail Abstract Author


Session 107 Poster Abstracts
Outcomes on ART in Resource-constrained Settings: Randomized Trials and Observational Cohorts
Session Day and Time: Wednesday, 1-2:30 pm
Poster Hall


600    
Lymphocyte T CD4+ Response to ART According to the HIV Type in a Large Cohort Collaboration in West Africa
J Drylewicz1, S Eholie2, M Maiga3, C Amani-Bosse4, P Sow5, D Ekouevi1,6, K Peterson7, E Bissagnene2, F Dabis1, Rodolphe Thiébaut*1, and IeDEA West Africa Collaboration
1INSERM U897 Epidemiology and Biostatistics, Bordeaux, France; 2Service de Maladies Infectieuses et Tropicales (SMIT), CHU de Treichville, Abidjan, Côte d`Ivoire; 3Service d`Hépato-Gastro-Entérologie, Hôpital Gabriel Touré, Bamako, Mali; 4ACONDA - MTCT+ initiative, Abidjan, Côte d`Ivoire; 5ANRS 1215 Cohort, Dakar, Sénégal; 6Programme PAC-CI, CHU de Treichville, Abidjan, Côte d`Ivoire; and 7Fajara Cohort, MRC, Banjul, Gambia

Background:  Data on the response to ART in HIV-2-infected patients are sparse. However, there is some evidence that NNRTI-based ART regimens are not efficient. With the increasing access to ART in West Africa with drug regimens mainly based on the use of NNRTI, the management of HIV-2-infected patients and their response to treatment need to be explored.

Methods:  Collaboration of 11 cohorts of HIV-infected adult patients followed in Senegal (1), Gambia (1), Mali (2), Benin (1), and Côte d’Ivoire (6) with computerized data. CD4 count was modelled using a linear mixed model to describe the CD4 response. Potential bias due to loss to follow-up was checked by comparing these estimates with those from a joint model of longitudinal measurements of CD4 and log time to drop-out.

Results:  Among 17,291 patients (including 16% loss to follow-up the first year), 10,073 patients (6% lost to follow-up) were available for this analysis with at least one CD4 measurement, the type of HIV and the type of ART recorded:  9482 HIV-1, 270 HIV-2, and 321 dually infected. Compared to HIV-2- or dually-infected, HIV-1-only patients were more likely to be younger (37 vs 43 and 40 years old in median) and female (67% vs 54% and 52%). Observed baseline CD4 counts were similar in the three groups (overall median 155, IQR 68; 249 cells/µL). In HIV-1 patients, the most common ART regimen was 2 NRTI and 1 NNRTI (n = 7714) followed by 2 NRTI + 1 PI (n = 931) or boosted PI (n = 326); this distribution of ART regimens was comparable for dually infected patients:  n = 135, 106, and 60, respectively. HIV-2 patients were most often treated with PI-based regimen (n = 109 with single PI, n = 84 with boosted PI), but 45 of them were treated with an NNRTI. The estimated mean CD4 count 12 months after ART initiation was 318/µL in HIV-1, 260 in HIV-2, and 270 in dually infected patients, with a significant lower increase of CD4 count in HIV-2 and dually infected patients (p <0.001). However, the response in HIV-2- and dually infected patients treated by a regimen other than NNRTI-based was not significantly different than the response obtained in HIV-1 only patients (p = 0.12, p = 0.24, respectively).

Conclusions:  The CD4 response was as good in HIV-2- or dually infected patients treated by a PI-based regimen as in HIV-1-infected patients. The poor response of HIV-2-infected patients treated with NNRTI-based regimens endorses the need in West Africa for investigating the HIV type either before starting treatment or in case of failure to first-line ART.