Effects of switching to PI monotherapy on measures of lipoatrophy: meta‐analysis of six randomized HIV clinical trials

Abstract

BackgroundSwitching from triple combination treatment to protease inhibitor (PI) monotherapy may prevent or reverse adverse events related to long‐term nucleoside analogues. Lipoatrophy is associated with long‐term use of thymidine analogues (zidovudine and stavudine).MethodsA detailed MEDLINE search was conducted to identify randomised clinical trials of triple combination treatment versus PI monotherapy. Summary results from analysis of changes in body composition (DEXA analysis) were collected: the mean change in limb fat and trunk fat to Week 48 or 96, and the percentage of patients with lipoatrophy (20% reduction from baseline in limb fat) or lipohypertrophy (20% rise from baseline in trunk fat).ResultsSix randomised trials of PI monotherapy versus triple therapy with data on body composition changes, measured by DEXA scanning at baseline and Week 48 or 96, were identified: Abbott‐613 (LPV/r vs ZDV/3TC/EFV, induction‐maintenance trial, n=105), Monark (LPV/r vs ZDV/3TC/LPV/r, first‐line trial, n=63), Kalesolo (LPV/r vs LPV/r +2NRTIs, switch trial, n=42), MONOI (DRV/r vs DRV/r + 2NRTIs, switch trial, n=156), MONARCH (DRV/r vs DRV/r + 2NRTIs, switch trial, n=30) and KRETA (LPV/r vs LPV/r + ABC/3TC, switch trial, n=74). In the meta‐analysis, there were greater rises in limb fat in the PI monotherapy arms than the triple therapy arms (mean difference =277g, 95% CI=+36 to+517g, p=0.024). The percentage of patients with lipoatrophy was significantly lower in the PI monotherapy arms (4%) than the triple therapy arms (20%), (p=0.0005). There was no difference between PI monotherapy and triple therapy for mean change in trunk fat (mean difference=−73g, 95% CI = −621 to +475g, p=ns). There was also no significant difference in the risk of lipohypertrophy between the PI monotherapy arms (32%) and the triple therapy arms (27%) (p=ns). In each of the four analyses, there was no evidence for heterogeneity of treatment effects between the trials (Cochran's Q tests, p=ns for each comparison).ConclusionsIn this meta‐analysis, the risk of lipoatrophy was significantly lower for patients taking PI monotherapy, compared to triple therapy. There was no significant difference between the arms for lipohypertrophy. However, several of the trials included zidovudine in the control arm, which carries a higher risk of lipoatrophy than tenofovir and abacavir, which are now more widely used.