938
Switching from Combivir to Truvada Preserves Limb Fat: Results of a DEXA Sub-Study of a 48-Week Randomized Study
Graeme Moyle*1, M Fisher2, and SWEET Study Group
1Chelsea and Westminster Hosp, London, UK and 2Brighton and Sussex Univ Hosp, Brighton, UK
Background: Lipoatrophy is a stigmatizing toxicity
of thymidine analogs and obstacle to medication adherence and treatment
success. We investigated the effects on limb fat of continued combivir (CBV) vs
replacement with truvada (TVD) in patients with no clinical lipoatrophy.
Methods: A 48-week study in subjects stable on twice-daily
CBV+once-daily efavirenz (EFV), with HIV RNA <50 copies/mL for ≥6
months, randomized 1:1 to continue CBV+EFV or switch to once-daily TVD+EFV. Limb
fat and bone were assessed by dual-energy X-ray absorptiometry (DEXA). Analyses
were performed to assess factors associated with limb fat changes.
Results: Of 250 subjects we randomized, 234 received
at least 1 dose of study medication: 117 continued CBV and 117 switched to
TVD. The DEXA sub-study randomized 100 subjects, 74 (36 CBV arm, 38 TVD arm) had
data available at baseline and week 48; 36-month median prior use of CBV (86% no
other prior thymidine analogue). Baseline characteristics were well matched. Baseline
mean limb fat was 6208 g (TVD) and 6148 g (CBV). At 48 weeks <50 copies/mL
viral suppression was maintained in 88% TVD vs 85% CBV (intent-to-treat M/S=F:
95%CI –6% to +11%, p = 0.70). Mean limb fat increased +261 g (TVD), but
decreased –187 g (CBV). Mean treatment difference of 448 g, 95%CI 57 to 839; p
= 0.02. In those with the highest tertile of limb fat at baseline (5940 to 22,433
g) changes were +147 g (TVD) and –280 g (CBV). For those in the lowest tertile
of limb fat (1199 to 3899 g) changes were +383 g and –198 g. In a post hoc
analysis, limb fat gain was significantly greater in those patients who had
been exposed to CBV for <3 years, +283 g (TVD) vs –451 g (CBV), treatment
difference of 734 g (p = 0.01) relative to ≥3 years, +240 g (TVD)
vs +140 g (CBV). A post hoc analysis of bone mineral density (BMD)
showed no statistical difference in mean change from baseline for hip: –0.012 (TVD)
p = 0.143 vs 0.008 (CBV) p = 0.633; or for lumbar spine: –0.002 (TVD)
p = 0.960 vs 0.011 (CBV) p = 0.450. There was no statistical
difference from baseline in renal function.
Conclusions: Switching from CBV to TVD in those
receiving EFV maintains virological control and results in increased limb fat
vs loss on CBV, with no change in hip or lumbar spine BMD. Earlier switch from
CBV to TVD may protect more against limb fat loss and allow limb fat recovery.
|
