[Clinical use of HIV-1 resistance genotyping. Predictive factors of poor virological evolution in salvage treatments].

Objectives: To describe the use of genotype resistance testing (GRT) for virological failure in clinical practice, and the long-term clinical and virological evolution in patients for whom it is requested. To identify the predictive factors of virological failure in patients with antiretroviral (ARV) salvage therapy.

Methods: Observational study in HIV-infected patients for whom GRT was requested for virological failure (VF) in the period of 1 October 1999 to 31 December 2001. Logistic regression analysis was used to determine the predictive factors of virological progression.

Results: Over the period studied, 196 patients required GRT for VF (15%) among those monitored in specific units. GRT was mainly requested for patients who had been extensively pretreated for a mean of 5 years and with a median of 5 ARV combinations. Half the patients presented 3 or more mutations associated with thymidine analogs (TAMs), mutations associated with non-nucleoside analogs (NNRTIs), and 5 or more mutations associated with protease inhibitors (PIs). In 143 (74%) patients, the RTV regimen was changed on the basis of GRT results. In the intent-to-treat analysis, the percentage of patients with plasma VL < 400 cop/mL at 6, 12 and 18 months was 41%, 29% and 17%, respectively. In the on-treatment analysis, the results were 50%, 48% and 46%, respectively. Mean CD4 lymphocyte increase was 59.74 and 94 cells/mm 3. The variables predicting virological failure (plasma VL > 400 cop/mL) at 12 months were plasma VL > 30,000 cop/mL (OR 6, 1.8-19.5) and accumulation of 3 or more TAMs (OR 4.4, 1.3-15) at the start of ARV salvage therapy.

Conclusion: Even though in clinical practice GRT is requested for patients with various treatment failures, when ART salvage treatment was started, plasma VL was undetectable and immunological response persisted in 40% of patients followed-up for 18 months. The factors best predicting virological evolution were VL and the number of baseline TAMs.