Background: When to switch HAART in heavily pre-treated patients with detectable viral load and stable CD4cell count is still controversial

Methods:  An open-label, randomized,48-week prospective trial has been carried out. Patients, Lopinavir/rtv (Lpv/r)-naive, with at least 2 previous HAART failed regimens, viral load between 1000 and 20,000 copies/mL for at least 6 months and CD4 cell count ³200 cells/mL twice consecutively determined, were randomized to maintain current treatment or to switch to an optimized Lpv/r-based regimen. Viral load increase above 30,000 copies/mL, CD4 decrease below 200 cell/mL, and grade III/IV side effects, were considered primary end points. Rate of HAART interruption for any other reason (e.g., poor adherence, quality of life, patient’s request) was also recorded. End-point rate and time distribution was studied with long-rank test and Spearman’s rank multivariable analysis. The-48 week data are here presented.

Results:  Of 201 patients, 102 were randomized in the current treatment arm and 99 in the Lpv/r arm. Mean previous failed regimens were 3; ART exposure time was 6 years. Current treatment and Lpv/r patients’ disposition at baseline wa, respectively:  mean viral load 3.5 and 3.7 (SD 0, 4) copies/mL; CD4 cell count 455 (SD 219) and 388 (SD 227) cells/mL; median resistance mutations 7.74 (0 to 24) and 7.21 (0 to 18). In the current treatment arm, maintenance treatment was protease inhibitor (PI) sparing in 69% of patients. At 48 weeks, the mean viral load change was –0.16 and –1.4 copies/mL and CD4 –37 and +67 cell/mm³  in the current treatment and Lpv/r arms, respectively. Viral load <50 copies/mL was obtained in 49% of Lpv/r patients. Viro-immunological end-point was reached by 11 and 3 of the current vs Lpv/r arm patients. The former patients who suffered viro-immunological failure switched to the Lpv/r-based regimen and all reached viro-immunological success. No grade III/IV adverse events were recorded. Treatment interruption for any other reason was observed in 28 and 23 in the current and Lpv/r, respectively (p = 0.2). Current therapy patients reached a viro-immunologic end-point more often than the Lpv/r arm patients in each protocol analysis, p = 0.03, but in intention-to-treat analysis, p = 0.4. Variables related to therapy interruption were female sex (p = 0.03) and viral load at baseline (p = 0.001); to CD4 decrease <200 cell/mm³; nadir and baseline CD4 cell count (p = 0.002 and 0.02) and to be randomized into the current treatment arm (p = 0.02); to viral load increase to 30,000 copies/mL; and number of previous PI and NRTI (p = 0.01)

Conclusions:  To defer HAART switch in heavily treated HIV patients with low viral load and high CD4 cell count could be an option to save new drugs. Further analyses are needed to characterize the profile of patients who preferentially could switch earlier because of severe risk of heavy viro-immunological impairment.