935
Clinical Associations of Extremity Fat Loss: ACTG 5142, a Prospective, Randomized, Phase III Trial of NRTI-, PI-, and NNRTI-sparing Regimens for ART of Naive, HIV-1-infected Subjects
Richard Haubrich*1, S Riddler2, G DiRienzo3, Y Zheng3, W Powderly4, K Garren5, D Butcher6, J Rooney7, J Mellors2, D Havlir8, and the AIDS Clinical Trials Group 5142 Study Team
1Univ of California, San Diego, US; 2Univ of Pittsburgh, PA, US; 3Harvard Sch of Publ Hlth, Statistical and Data Analysis Ctr, Boston, MA, US; 4Univ Coll Dublin, Ireland; 5Abbott Labs, Abbott Park, IL, US; 6Bristol-Myers Squibb, Plainsboro, NJ, US; 7Gilead Sci, Foster City, CA, US; and 8Univ of California, San Francisco, US
Background: Extremity fat loss may be associated
with factors other than ART; these have not been well defined in prospective
studies.
Methods: This open-label, randomized study compared lopinavir
(LPV) + efavirenz (EFV) vs LPV+2NRTI vs EFV+2NRTI. NRTI were selected from zidovudine
(ZDV), stavudine extended-release (d4T XR), or tenofovir (TDF) (each plus lamivudine
[3TC]). DEXA and fasting lipids were performed at baseline, 48 and 96 weeks. Fat
loss was also assessed by patient self-report (yes/no). Extremity fat changes
were analyzed with logistic regression (lipoatrophy was defined as ≥20%
loss of limb fat from baseline by DEXA) and linear models for continuous
metrics. All analyses were intent to treat, without adjustment for multiple
comparisons or regimen changes.
Results: Lipoatrophy at 96 weeks was greater for EFV+2NRTI
(32%), d4T- (43%) and ZDV- (27%) containing regimens, and least for TDF-containing
and NRTI-sparing (LPV+EFV) regimens (8 to 10%). After accounting for regimen
and NRTI, race and ethnicity was not associated with lipoatrophy. Factors
related to fat changes are shown in the table. Self-reported fat loss at 96
weeks did not correlate with objective (DEXA) findings; only 30% of subjects
with lipoatrophy by DEXA as defined by reported fat loss. However, subjects
with self-reported fat loss tended to gain less extremity fat at 96 weeks than
subjects without self-reported fat loss (median 0.25 vs 0.8 kg).
*OR of lipoatrophy, number >1 signifies greater risk
**p value for each factor in models with regimen and NRTI included
*** coefficient is increase (or decrease) in extremity fat per unit increase in
risk factor.
Conclusions: Higher baseline CD4 and lower gain in
body and trunk weight were associated with lipoatrophy and less gain of limb
fat, independent of antiretroviral regimen. Analyzing fat change as a
continuous outcome also identified male sex, non-AIDS, lower baseline extremity
fat, and smaller increases in cholesterol and LDL to be significantly associated
with less gain of limb fat. Several factors other than ART may be important in
the pathogenesis of fat loss.
|
