Background:  Virological mechanisms responsible for discordant treatment responses are not well understood. To identify determinants of discordant viro-immunological response to genotype guided antiretroviral therapy. To analyze different evolution of mutational patterns between discordant and failed patients.

Methods:  A cohort of  HIV-infected failing patients who underwent genotype resistance testing, were analyzed. Those who showed increasing CD4 cell for at least 1 year despite ever detectable HIV viral load (discordant viro-immunological response), were compared with viroimunological failed patients. Baseline characteristics, previous HIV clinical and viro-immunologic history, baseline drug resistance mutations, were analyzed. In patients with a subsequent genotype resistance typing, a comparative analysis of mutations was performed. Categorical variables were studied by contingency tables, logistic regression models were constructed for multivariate analysis of binary outcomes.

Results:  From January 2000 to June 2003, 69 discordant viro-immunological response and 168 viroimunologically failed patients were identified and followed-up. The 2 groups were similar in terms of time since HIV infection, antiretroviral exposure history, CDC stage, demographic variables, and baseline viral load. For 100 patients a second genotype resistance testing was available. Determinants of drug resistance at multivariate analysis were:  presence of lopinavir/r in the failing HAART regimen (OR 17.8 95% CI 1.28 to 246.53), introduction of lopinavir/r in the genotype resistance testing guided regimen (OR 6.09 1.10 to 33.55), accumulation of at least 1 new minor (OR 3.19 0.96 to 10.54) or at least a new major (OR 4.98 1.46 to 16.92) PI-drug resistance mutations, CD4 cell at baseline <200/mL (OR 4.41; 1.33 to 14.57), after controlling for baseline viral load, number of previous HAART regimens, presence of PI in the genotype resistance testing guided regimen. A strong association was also found between the appearance of L90M PI mutation in the second genotype resistance testing and discordant viro-immunological response (OR 8.66 1.74 to 42.98) when adjusted for baseline CD4 cell, HIV-RNA viral load, number of antiretroviral regimens experienced, prior PI exposure. Furthermore, carrying M184V in both genotype resistance tests or its appearance in the second one was weakly associated with a higher likelihood of drug resistance (OR 2.27 0.88 to 5.89).

Conclusions:  Patients with drug resistance accumulate a higher number of minor and major PI resistance mutations compared with failing patients. The strongest effect was observed for patients selecting the L90M mutation while on genotype resistance testing guided therapy. Moreover, previous or current lopinavir/r-based regimens had a favourable effect on immunological response even despite virological failure.

Keywords: viroimmunological discordant; genotypic resistance test; lopinavir/r