CROI 2019 Abstract eBook

Abstract eBook

Poster Abstracts

1 VA North Texas Health Care Center, Dallas, TX, USA, 2 University of Texas Southwestern, Dallas, TX, USA, 3 University of Texas at Houston, Houston, TX, USA, 4 University of Texas at San Antonio, San Antonio, TX, USA, 5 University of Pennsylvania, Philadelphia, PA, USA Background: While older protease inhibitors (PI) were more frequently associated with central fat accumulation, initiation of currently used ART regimens has been associated with increases in body mass index (BMI), particularly in women and with integrase strand transfer inhibitors (INSTI). The goal of this study was to analyze the differential effect of individual PIs and INSTIs on changes in BMI by sex and race in a large urban HIV clinic. Methods: All patients initiating ART at the Parkland Health and Hospital System in Dallas, TX from 2009 to 2017 were included in the analysis. Exposure to ART was defined as concurrent receipt of at least two nucleoside reverse transcriptase inhibitors (NRTI) and at least one PI, Non-nucleoside reverse transcriptase inhibitor (NNRTI) or INSTI. In regression analysis, we compared yearly change in BMI (kg/m2) between men and women and between Blacks, Hispanics and Non-Hispanic Whites following initiation of PIs (Atazanavir [ATV], Darunavir [DRV] or Lopinavir [LPV]) or INSTI (Raltegravir [RAL], Elvitegravir [EVG] or Dolutegravir [DTG]). We controlled for year of HAART initiation, baseline CD4 count and HIV-1 RNA, and whether patients achieved virologic suppression on HAART. Results: We included 4,048 patients, 69%male, 53% Black, 28% Hispanic, and 16% non-Hispanic Whites. Mean age was 46.3 years (SD 11.9). Mean baseline BMI was 27.0 kg/m2 (6.4). Median follow-up time on HAART was 6.7 years (IQR 2.8 – 11.2). Cumulative exposure to NNRTI, PI, and INSTI-based HAART were 3546, 6184, and 3090 person-years, respectively. The BMI slope per year on NNRTI, PI and INSTI were 0.22, 0.24 and 0.32, respectively. BMI slopes for individual PI- and INSTI-based regimens by sex, race and ethnicity are presented in Table 1. There was no significant interaction between sex and race/ethnicity on BMI gains. Proportion of overweight /obese (BMI, ≥ 25) increased from 51% at HAART initiation to 65% at year 3 (p<0.001). Conclusion: We observed a differential effect of individual INSTI and PI-based HAART regimens on BMI changes by sex. All PIs were associated with greater BMI gain in women than in men, but with no difference by race/ethnicity. LPV-based ART was associated with relatively smaller BMI gains. Among INSTIs, while EVG appeared to be associated with greater BMI gains overall, the effect did not vary or by sex or race/ethnicity. DTG and RAL are associated with greater BMI gains in women, and DTG with greater gains in Blacks & Hispanics.

674 WEIGHT GAIN AMONG VIRALLY SUPPRESSED PERSONS WHO SWITCH TO InSTI-BASED ART Frank J. Palella 1 , Nabil Rayeed 2 , Jun Li 3 , Douglas Ward 4 , Jack Fuhrer 5 , Stacey Purinton 2 , Ellen Tedaldi 6 , Richard Novak 7 , Kate Buchacz 3 1 Northwestern University, Chicago, IL, USA, 2 Cerner Corp, Kansas City, MO, USA, 3 CDC, Atlanta, GA, USA, 4 Dupont Circle Physicians Group, Washington, DC, USA, 5 Stony Brook University, Stony Brook, NY, USA, 6 Temple University, Philadelphia, PA, USA, 7 University of Illinois at Chicago, Chicago, IL, USA Background: INSTI-associated weight gain has been described among ART- naïve persons initiating INSTI-containing ART, but not among virally suppressed (VS) persons whose first INSTI exposure is via a switch regimen. We evaluated changes in weight (CW) among such persons in the HIV Outpatient Study (HOPS). Methods: We analyzed medical record data of patients from nine United States HIV clinics who were INSTI-naive and VS for >1 year on non-INSTI-based ART, and switched to INSTI-based ART and remained VS. Participants received INSTI-based ART for >6 months, had >2 weights recorded in the year prior to switch and >1 after. We evaluated CW over time, overall and stratified by demographics, pre-switch body mass index (BMI) and ART use, CD4 at ART start, and INSTI received. We used multivariable random regression mixed model to estimate factors associated with CW. Results: Among 437 patients (median age 51 years, interquartile range 44.5, 57.5), 86 (19.6%) were women, 107 (24.5%) were non-Hispanic Black (NHB). Pre-INSTI regimens often included an NNRTI (193 [44.1%]) or PI (185 [42.0%]) with >1 NRTI (402 [91.5%]). INSTI regimens included raltegravir (236 [54.0%]), elvitegravir (89 [20.4%]), or dolutegravir (112 [25.6%]). Mean CW in the year prior to INSTI was -0.2 kg (95% confidence interval [CI]: -0.6, 0.2). Mean duration of INSTI use was 2.9 years (max=9.7 years). Mean CW on INSTI was 1.2 kg (CI 0.6, 1.9), did not differ by INSTI drug used (p>0.2) and was greater for persons with pre-INSTI BMI < 25 (2.2 kg, CI 1.5, 3.0) than 25-29.9 (0.5 kg, CI -0.5, 1.4) or >30 (0.4 kg, CI -1.7, 2.6), p=0.03; NHB than Non-Hispanic whites, 2.7 kg (CI 1.3, 4.1) vs 1.0 kg (CI 0.2, 1.7), p=0.02; and persons whose pre-INSTI ART did not include an NRTI vs those whose did, 4.5 kg (CI 1.8, 7.3) vs. 0.9 kg (CI 0.3, 1.6), p<0.01. Duration of INSTI use was not associated with CW: mean 1.0 kg (CI 0.5, 1.4) for 6-<12 months (mos), 1.2 kg (CI -0.5, 2.9) for 12-<24 mos, 1.3 kg (CI 0.7, 1.9) for 24-<60 mos, 1.2 kg (CI 0.5, 2.0) for ≥60 mos, p=0.7. In multivariable models NHB race, and no pre-INSTI NRTI use remained associated with greater percent change in weight (p<0.05) while lower pre-INSTI BMI was borderline significant, p=0.08. Conclusion: We observed weight gain among VS persons who switched to INSTI-based ART that was associated with NHB race, no pre-INSTI NRTI use, and lower pre-INSTI BMI .These findings of differential risk for INSTI-related weight gain require further evaluation. 675 DIFFERENTIAL BMI CHANGES FOLLOWING PI- AND InSTI-BASED ART INITIATION BY SEX AND RACE Roger Bedimo 1 , Xilong Li 2 , Beverley Adams-Huet 2 , Jordan E. Lake 3 , Barbara S. Taylor 4 , Deborah Kim 5 , Pablo Tebas 5 , Amneris Luque 2

Poster Abstracts

676 LONG-LASTING ALTERATIONS IN FAT DISTRIBUTION IN PLWH EXPOSED TO THYMIDINE ANALOGUES Marco Gelpi 1 , Shoaib Afzal 2 , Andreas Fuchs 1 , Jens D. Lundgren 1 , Andreas D. Knudsen 1 , Ninna Drivsholm 1 , Anne-Mette Lebech 1 , Birgitte Lindegaard 1 , Jørgen T. Kuhl 1 , Per E. Sigvardsen 1 , Lars Køber 1 , Børge Nordestgaard 3 , Klaus F. Kofoed 1 , Susanne D. Nielsen 1 1 Rigshospitalet, Copenhagen, Denmark, 2 Herlev and Gentofte Hospital, Copenhagen, Denmark, 3 University of Copenhagen, Copenhagen, Denmark Background: Thymidine analogues (TA) and didanosine (ddI) have been associated with redistribution of body fat from subcutaneous (SAT) to visceral (VAT) adipose tissue, which, in turn, is a risk factor for cardiovascular disease (CVD). We explored differences in adipose tissue distribution between people living with HIV (PLWH) with/without prior exposure to TA and/or ddI and uninfected controls and the association with CVD risk factors. Methods: 761 PLWH from the COCOMO study aged > 40 and 2,283 age- and sex-matched uninfected controls from the GCPS study were included. PLWH were stratified according to prior exposure to TA and/or ddI. VAT and SAT were determined by abdominal CT-scan. Hypotheses were tested by linear and

CROI 2019 258