CLINICAL SCIENCE Body Composition and Metabolic Changes in Antiretroviral-Naive Patients Randomized to Didanosine and Stavudine vs. Abacavir and Lamivudine Judith C. Shlay, MD, MSPH,* Fehmida Visnegarwala, MD,† Glenn Bartsch, PhD,‡ Jack Wang, MS,§ Grace Peng, MS,‡ Wafaa M. El-Sadr, MD, MPH, k Cynthia Gibert, MD,¶ Donald Kotler, MD,§ Carl Grunfeld, MD, PhD,# and Subhasree Raghavan, PhD k for the Terry Beirn Community Programs for Clinical Research on AIDS (CPCRA) Summary: Comparisons of body composition and metabolic changes among antiretroviral-naive patients randomly assigned to didanosine and stavudine- (ddI + d4T) vs. abacavir and lamivudine- (ABC + 3TC) containing regimens were assessed in a nested substudy of an ongoing multicenter randomized trial. At baseline and every 4 months, body cell mass and total body fat were calculated, anthropometric measurements were performed, and fasting metabolic parameters were obtained. The rates of change (unit/mo) estimated using the slopes of regression lines and overall mean changes from baseline were compared by study assignment. Among 96 patients enrolled, 46 received ddI + d4T- and 50 received ABC + 3TC- containing regimens with a median follow-up of 32.4 months. For both study arms, an overall increase in the rates of change was seen for body cell mass. For ddI + d4T, after an initial increase, the rates of change declined for regional fat and total body fat compared with an increase for ABC + 3TC, with the 2 arms being significantly different (P , 0.05). For high-density lipoprotein cholesterol rates of change, ddI + d4T decreased, while ABC + 3TC increased. For both arms, low-density lipoprotein cholesterol decreased, while triglycerides increased. Early and sustained increases in insulin and insulin resistance were seen only for ddI + d4T. In this prospective study, metabolic and body composition changes varied according to whether subjects received ddI + d4T or ABC + 3TC. Key Words: HIV, prospective, randomized controlled trial, lipo- atrophy, nucleoside reverse transcriptase inhibitor, body composition, adverse effects (J Acquir Immune Defic Syndr 2005;38:147–155) INTRODUCTION Morphologic changes (lipoatrophy and lipohypertro- phy), insulin resistance (IR), and dyslipidemia are emerging as important metabolic consequences of antiretroviral therapy in patients with HIV/AIDS. 1–5 Currently, the etiology of these metabolic abnormalities is unclear, with prior studies sug- gesting a direct role for protease inhibitors (PIs). 6–8 Sub- sequently, the use of nucleoside reverse transcriptase inhibitors (NRTIs) has been implicated in the development of the lipo- atrophy component of the HIV lipodystrophy syndrome 9–12 with associations identified with both cumulative duration of NRTI exposure and the current use of thymidine analogues, particularly stavudine (d4T). 13–21 However, previously pub- lished studies have not clearly determined the relative con- tribution of various HIV therapies and duration of these ther- apies with the development of lipoatrophy. 13,22 Few prospective randomized studies have compared body composition and metabolic changes in thymidine analogue–containing regimens to thymidine analogue–sparing regimens. 23–26 These comparisons are clinically important in antiretroviral-naive patients, as recent results have demon- strated a modest improvement in fat mass among patients changing from a thymidine analogue-containing regimen to a thymidine analogue–sparing regimen. 12,27–29 We compared changes in metabolic parameters and body composition among antiretroviral-naive patients in a study conducted by the Community Program for Clinical Research on AIDS (CPCRA). Patients were randomly assigned to receive either didanosine and stavudine (ddI + d4T)- vs. abacavir and lamivudine (ABC + 3TC)-containing regimens to determine differences in a thymidine analogue–containing regimen vs. a thymidine analogue–sparing regimen, respectively. Received for publication April 22, 2004; accepted August 16, 2004. The National Institute of Allergy and Infectious Diseases (NIAID), National Institutes of Health, 5U01AI042170-10 and 5U01AI046362-03, provided financial support for this study as part of the Flexible Initial Retrovirus Suppressive Therapies (FIRST) study (CPCRA 058) and the Metabolic Substudy of FIRST (CPCRA 061). From the *Denver Community Programs for Clinical Research on AIDS, Denver Public Health, University of Colorado Health Sciences Center, Denver, CO; †Houston AIDS Research Team, Baylor College of Medi- cine, Houston, TX; ‡CPCRA Statistical and Data Management Center, University of Minnesota, Minneapolis, MN; §Body Composition Unit of St. Lukes-Roosevelt Hospital, Columbia University College of Physicians and Surgeons, New York, NY; k Harlem AIDS Treatment Group, Harlem Hospital, Columbia University College of Physicians and Surgeons, New York, NY; {Wide-Reaching AIDS Partnership, Veterans Affairs Medical Center, Washington, DC; and #Veterans Affairs Medical Center and University of California, San Francisco, CA. This paper was presented in part at the 15 th International AIDS Conference in Bangkok, Thailand as an oral presentation, July 11–16, 2004. Reprints: Fehmida Visnegarwala, Baylor College of Medicine, Department of Medicine, Section of Infectious Diseases, Room #424, 2015 Thomas Street, Houston, TX 77009 (e-mail: fehmidav@bcm.tmc.edu). Copyright Ó 2005 by Lippincott Williams & Wilkins J Acquir Immune Defic Syndr  Volume 38, Number 2, February 1 2005 147