Background: Combinations of PIs could prove effective in patients having failed multiple lines of antiretroviral therapy, provided that negative interactions do not alter their PK profiles. We describe pharmacokinetic (PK) and virological response at 6 weeks to a salvage therapy associating LPV/r and APV, with or without an additional boost by ritonavir (RTV).

Methods: A prospective, randomized, open-label, multicentre trial in patients with CD4+< 500/mm³ and plasma HIV viral load (pVL)> 10,000 copies/mL after at least 2 PIs and 1 NNRTI. For the first 2 weeks, patients were randomized to receive either: LPV/r (group 1), APV (1200 mg/d) + RTV (200 mg/d) (group 2), LPV/r + RTV (200 mg/d) (group 3), APV (1200 mg/d) + RTV (400 mg/d) (group 4). From weeks 2 to 26, all patients received APV and LPV/r with an additional boost of 200 mg/d of RTV for groups 3 and 4. We used the Mann-Whitney test for the comparison of medians between groups and the Wilcoxon signed-rank test for assessing the addition of the second PI between paired observations. The comparison of pVL changes from baseline to week 6 between groups used a Student’s t-test.

Results: 40 patients were randomized, 37 started treatment and were analyzed. At baseline, median CD4+ was 207/mm³ and median pVL was 4.7 log₁₀ copies/mL. Average number of antiretrovirals taken prior to randomization was 7.7. The median number of baseline PI mutations was 7. Median APV and LPV plasma trough concentrations (Cmin) at week 2 and week 6 were as follows:

	Median (range) APV Cmin (ng/mL)		Median (range) LPV Cmin (ng/mL)
Group; RTV mg/day	Gr 2 (n=6) RTV200	Gr 4 (n=8) RTV400	Gr 1 (n=13) RTV200	Gr 3 (n=10) RTV400
W2	1514 (680-3067)	2334 (648-5317)	7126 (3342-14052)	10488 (6414-13784)
W6	873 (569-1816)	721 (323-1104)	4604 (1578-9041)	10439 (4530-12615)

Median APV Cmin was significantly lower after addition of LPV/r (p=0.003), but there was no significant impact of APV on LPV Cmin. Median pVL (log₁₀ copies/mL) changes at week 6 were significantly higher in patients receiving the higher RTV dose: -2.2 (groups 3+4) vs -1.5 (groups 1+2)(p 0.02), but it did not differ whether patients received APV or LPV/r between week 0 and week 2: -1.5 (group 1), -1.4 (group 2), -2.2 (group 3), -2.2 (group4).

Conclusions: In patients having failed multiple lines of treatment, salvage with APV combined to LPV/r and RTV showed significant virological response despite a PK interaction between LPV and APV.