Background: One of the more vigorous debates in the field of highly active antiretroviral therapy (HAART) is how to start it and what the optimal drug sequence is.
Methods: A retrospective cohort analysis was performed. The aim was to evaluate which variables could influence the virological response to second-line genotypic-based HAART in patients with virological documented first-line HAART failure. A positive response was defined as a confirmed HIV RNA level < 50 copies/ml.
Results: Two hundred and eight patients were included. Demographic characteristics, risk factors for HIV acquisition, and drugs included in the initial treatment did not significantly influence the considered outcome. According to a multiple logistic model, the presence of thymidine analogue mutations (TAMs) had a negative association with the virological outcome (P = 0.006), whereas the use of a boosted protease inhibitor (PI) in second-line HAART was positively associated with the endpoint (P = 0.001). Patients receiving a genotypic-based second-line HAART containing a boosted PI achieved a viral load < 50 copies/ml in a 74.2% of cases compared with 52.2% of those whose therapy did not contain a boosted PI. This difference was statistically significant (P = 0.002) with an odds ratio (OR) of 2.63 and a 95% confidence interval (CI) ranging from 1.46 to 4.76. This last variable positively influenced the outcome even when the analysis was restricted to patients harbouring a virus presenting TAMs. In this case, second-line HAART was successful in 66.7% of cases with an OR of 3.25 and a 95% CI ranging from 1.28 to 8.25 (P = 0.014).
Conclusions: the wider range of available therapeutic options has made resistance and drug-sequencing considerations a crucial point in selecting first-line HAART. Our data indicate that, by limiting the risk of selecting or accumulating TAMs, it could be possible to save further therapeutic options. In second-line regimens, the higher antiviral effect and genetic barrier of boosted PIs may overcome the limits of the use of NRTI backbones, which retain only a partial effectiveness.