Abstract

The search for determinants responsible for changes in CD4 cells count can help to optimize the efficiency of the nevirapine-based antiretroviral therapy in Côte d'Ivoire. Our study on the response to this treatment was carried out through taxonomy of CD4 cells count trajectories using non-hierarchical-descendant model. One hundred and sixty four patients were grouped according to their baseline CD4 count in three strata (<100/mm3, 100-200/mm3 and >200/mm3). In each category, classes of similar CD4 count trajectories represented by standard paths called meta-trajectories have been formed by the model and we have searched for the determining factors. On the overall, there was a significant variation between classes (p<0.001) of the following variables: average CD4 count, nadir CD4, peak CD4, average CD4 gain. According to the profile of meta-trajectories, CD4 cells gains are obvious during the first six months of treatment, and then changes are variable according to classes regardless of the baseline CD4 count. The other medical follow-up variables showed no significant variation between classes in the categories of initial CD4 count ≥100/mm3. However in patients with baseline CD4<100/mm3, variables like mean weight (p=0.04), mean hemoglobin count (p=0.01) and average gain of hemoglobin (p=0.03) are also explanatory of partition into different classes. Baseline characteristics showed no significant variation between the different classes in the categories of initial CD4 count≥ 100/mm3. However for patients with baseline CD4 count<100/mm3, sex is explanatory of classes partition (p<0.05). Correspondence analysis revealed other determining factors related to the different meta-trajectories (initial CD4 count, initial CD4 cells percentage, adherence, presence, or absence of opportunistic infections prior to the treatment). These factors influencing the CD4 cells response of nevirapine-based regimen in Côte d'Ivoire must be considered to maximize efficiency of the treatment. 

Key words: Nevirapine, CD4 cells count response, meta-trajectories, Non-Hierarchical-Descendant Model, antiretroviral regimen.