Home Search Abstracts View Session E-mail Abstract Author


Session 153 Poster Abstracts
Lypoatrophy/Lipohypertrophy: Predictors and Interventions
Session Day and Time: Monday, 1-4 pm
Room: Hall B


939
Influence of NRTI Choice and Efavirenz vs Lopinavir Treatment Options on Lipoatrophy-associated Adipose Tissue Toxicity: A Longitudinal Study
E Hammond, B McKinnon, C Pace, D Nolan, and Simon Mallal*
Murdoch Univ, and Royal Perth Hosp, Australia

Background: Clinical trials and observational cohort data have demonstrated associations between thymidine NRTI (particularly stavudine) use and lipoatrophy, while more recent data suggest an additional effect of efavirenz (EFV) treatment vs lopinavir (LPV) therapy. This study investigates the tissue-specific effects of HIV treatment choice in longitudinal adipose tissue specimens.

Methods:  Longitudinal adipose biopsies (2 to 4/patient) were obtained from 68 HIV+ patients, including 31 initiating HAART and 43 switching HAART. Samples were assessed for mtDNA depletion (real-time polymerase chain reaction) and tissue histology, with a subset (71 samples) further assessed for cytokine expression (interferon [IFN] -γ, interleukin [IL] -6, tunor necrosis factor [TNF] IL-8, IL-12, IL-18) and macrophage density. Selected HAART regimens included EFV (n = 23), LPV (n = 20), nevirapine (NVP; n = 21), and NRTI-only (n = 29), with "backbone" NRTI in these regimens comprising: non-thymidine NRTI (n = 39), zidovudine (AZT; n = 37), stavudine (d4T; n = 17). We also collected 34 ART-naïve samples. Statistical analysis utilized mixed effects regression models for group comparisons and covariate assessment.

Results:  Compared with ART-naïve samples, adipocyte mtDNA depletion was significantly associated with current use of d4T (median 21%, p <0.0001) or AZT (58%, p = 0.005), but not with non-thymidine NRTI use (82%, p = 0.3). No independent effect of LPV or EFV treatment on adipocyte mtDNA depletion was identified (both p >0.7). Adipose tissue pathology, macrophage infiltration, and cytokine expression were highly correlated and were significantly associated with cumulative exposure to either d4T or AZT (0.003< p <0.05). Non-thymidine NRTI therapy was not independently associated with adipose toxicity (p >0.2) nor was EFV or LPV use (p >0.2). Following NRTI switching, non-thymidine NRTI use was associated with normalization of adipocyte mtDNA within 1 year (p = 0.2 compared with ART-naive), although adipose pathological changes tended to persist (p <0.05 compared with ART-naïve, p >0.2 compared with pre-switch).

Conclusions:  Adipose tissue mtDNA depletion and pathological changes characteristic of clinical lipoatrophy are strongly and specifically associated with thymidine NRTI treatment. No independent effect of EFV or LPV treatment could be identified.