Background:  Response to Kaletra (LVP/r) based salvage regimens is influenced by both the number of mutations at the protease and LPV plasma levels. Nevertheless, the interaction between these 2 parameters has not been clearly defined for LPV. Moreover, cholesterol and triglyceride levels may be related to LPV plasma levels.

Methods:  All HIV⁺ patients with prior exposure to ARV from all 3 drug families and failing their current treatment started a salvage regimen based on LPV/r at standard doses. Genetic sequences were obtained at baseline. Drug resistance interpretation was performed according to IAS-USA guidelines. LPV trough concentrations (Ctrough) were measured at 3 months. Viral load, CD4⁺ counts, fasting cholesterol and triglyceride levels were measured at baseline and at 3, 6, and 12 months. Virological response was defined as >1 log reduction in plasma HIV-RNA and/or to <50 copopies/mL. The genotypic inhibitory quotient was calculated as LPV Ctrough/ number of protease mutations.

Results:  We analyzed 126 patients:  virological response was observed in 85/126 (68%), 74/120 (62%), and 65/113 (58%) at 12, 24, and 48 weeks, respectively. At baseline, the median (IQR) number of resistance mutations was 4 (2 to 7); 48 weeks responders had 2 (2 to 5) vs 7 (3 to 8) in non-responders (p = 0.02). The median (IQR) LPV Ctrough was 6,2 (2,2 to 8,7) mg/mL: 6.4 (4 to 9,2) in responders vs 4.9 (0,36 to 7,3) in non-responders (p = 0.03). In the multivariate analysis, both LPV Ctrough >5 mg/mL and £6 resistance mutations were independent predictors of 48 weeks responders (p = 0.02; OR 6; 95%CI: 1.3 to 28.8; and p = 0.002; OR 21.5; 95%CI: 4,3 to 105, respectively). Logistic regression analysis showed that a genotypic inhibitory quotient >1.2 was a predictor of 48 weeks response (p = 0.0001; OR 13, 95%CI 3.4 to 48.7). genotypic inhibitory quotient was closely related to viral load drop at 48 weeks (r = -0.52; p <0.0001); CD4 count increased +114 cells/mL (p <0.0001). Patients with both higher LPV Ctrough and genotypic inhibitory quotient showed greater CD4 increases at 48 weeks (r = 0.23, p = 0.025; and r = 0.315, p = 0.019, respectively). In patients with neither diet nor lipid lowering drugs, triglyceride and cholesterol percentage of increase at 48 weeks was correlated with LPV and RTV plasma levels, respectively (r = 0.28, p = 0.04; and r = 0.314, p = 0.018).

Conclusions:  Baseline HIV protease genotyping and LPV Ctrough independently predict response to LPV/r-based salvage therapy at 48 weeks. Nevertheless, genotypic inhibitory quotient offers the best predictor of virological response to Kaletra as well as a good estimator of viral load decrease. LPV Ctrough was also predictive of both CD4⁺ T-cell recovery and triglyceride elevations. In contrast, cholesterol elevations seem to be driven by RTV Ctrough instead of LPV Ctrough

Keywords: GIQ; Lopinavir