CROI 2020 Abstract eBook

Abstract eBook

Poster Abstracts

Background: Recent reports describe greater weight gain among antiretroviral therapy (ART)-naïve persons with HIV (PWH) starting integrase strand transfer inhibitor (INSTI)-based antiretroviral therapy (ART) vs. protease inhibitor (PI) or non-nucleoside reverse-transcriptase inhibitor (NNRTI)-based ART. Since many PWH have switched from non-INSTI to INSTI-based regimens, we assessed weight over time among PWH switched to an INSTI regimen (before the introduction of tenofovir alafenamide) in the multisite North American AIDS Cohort Collaboration on Research and Design (NA-ACCORD). Methods: Adult PWH with >2 years of no HIV-1 RNA measurements >1000 copies/mL prior to and following the switch from an NNRTI- or PI- to INSTI-based ART were included. Piecewise linear mixed models with random intercepts and slopes estimated pre- and post-switch weight over time, adjusting for age, sex, race, cohort site, HIV acquisition mode, calendar year, pre-switch ART class (NNRTI vs. PI), and CD4+ T cell count and BMI at the time of switch. We included interaction terms for sex, race, and age (<50 vs. >=50) with regimen and time. Results: A total of 2255 participants switched to an INSTI and had the required follow-up time; of these, 877 met viral suppression criteria and were included. At switch, median age was 50 years, BMI 26 kg/m 2 , and CD4+ count 619 cells/ mm 3 ; 83%were men, and 59%were white. Overall, the annualized weight slope among PI users was +0.80 (95% CI: 0.57 to 1.04) kg/year before switch, which decreased by -0.46 (-0.67 to -0.26) after switch to an INSTI (absolute slope +0.34 kg/year after switch). For NNRTI users, the slope before switch was +0.63 (0.34 to 0.91) kg/year, increasing by +0.50 (0.23 to 0.77) after switch to an INSTI (absolute slope +1.13 kg/year after switch). This difference was primarily driven by an increase in the weight slope among women, non-whites, and older PWH in the NNRTI group (table). Among individual INSTI drugs, the slope change after switch from NNRTI was highest for dolutegravir (DTG) at +0.93 (0.39 to 1.46) kg/ year vs +0.44 (-0.04 to 0.92) kg/year for elvitegravir and +0.23 (-0.13 to 0.58) kg/year for raltegravir. Conclusion: Women, non-whites and older PWH with viral suppression had greater annualized weight gain after switch from NNRTI- to INSTI-based ART, which was greatest for dolutegravir, whereas those switched from a PI had slowing of weight gain. These findings may reflect a heterogenous effect of ART class and agent on body weight regulation. 669 BODY COMPOSITION CHANGES OVER THE MENOPAUSAL TRANSITION IN HIV+ AND HIV– WOMEN Thuy Trang J. Nguyen 1 , Yifei Ma 1 , Rebecca Scherzer 1 , Anjali Sharma 2 , Mark H. Kuniholm 3 , Audrey French 4 , Margaret Fischl 5 , Howard Minkoff 6 , Michael Plankey 7 , Adaora Adimora 8 , Carl Grunfeld 1 , Phyllis Tien 1 1 University of California San Francisco, San Francisco, CA, USA, 2 Albert Einstein College of Medicine, Bronx, NY, USA, 3 State University of New York at Albany, Rensselaer, NY, USA, 4 Rush University Medical Center, Chicago, IL, USA, 5 University of Miami, Miami, FL, USA, 6 Maimonides Medical Center, Brooklyn, NY, USA, 7 Georgetown University, Washington, DC, USA, 8 University of North Carolina at Chapel Hill, Chapel Hill, NC, USA Background: Both overall and central adiposity are reported to increase during the menopausal transition. Whether HIV infection affects adiposity changes during this transition is unknown. We hypothesized that body mass index (BMI) and waist circumference (WC), a marker of visceral adiposity would increase over the menopausal transition and that HIV infection would blunt increases in BMI, and to a lesser extent WC due to the known effects of HIV on subcutaneous fat. Methods: Between 2000 and 2013, 632 HIV+ and 218 HIV– Women’s Interagency HIV Study participants underwent serial measures of BMI, WC, and Anti-Müllerian Hormone (AMH), a biomarker of ovarian reserve. The menopausal transition was categorized as: premenopause(>5years[yrs] before AMH became undetectable), early perimenopause(1-5yrs before), late perimenopause(first visit with undetectable AMH and up to 5yrs after), menopause(>5-10yrs after), and late menopause(>10yrs after). We used multivariable linear mixed regression models which adjusted for demographic,

behavioral, viral hepatitis, and CD4 count to estimate percentage(%) changes in BMI and WC relative to premenopause. Results: Women were mostly African-American (58%); mean age at onset of late perimenopause was ~45yrs in both HIV+ and HIV- women. HIV+ had lower BMI and WC than HIV– (mean:29 vs. 32kg/m 2 ;p<0.0001 and 94 vs 98cm;p=0.004, respectively). Figure shows the % BMI and WC change after adjustment. In HIV– women, we found the expected increase in BMI across the menopausal transition (from 5.2 to 12% higher than in premenopause) whereas in HIV+, the increase was much lower (1.2-1.8% higher) and blunted across the entire menopausal transition (difference in BMI change by HIV status at every stage,p<0.01). By contrast, WC progressively increased over the menopausal transition in HIV+ but the increase was blunted (difference in WC change by HIV status,p<0.01 except early perimenopause,p=0.41 and late menopause,p=0.14). Conclusion: Our findings suggest that HIV infection blunts the expected trajectory of increase in BMI over the menopausal transition, whereas the expected trajectory of increase in WC is preserved but also blunted. Studies are needed to examine whether women with HIV in the menopausal transition are at greater risk for perturbations associated with visceral obesity (e.g. insulin resistance, fatty liver disease) and to determine optimal timing of interventions to reduce visceral obesity.

Poster Abstracts

670 RESISTIN GENE POLYMORPHISM RELATED TO WEIGHT GAIN AND PSYCHIATRIC SYMPTOMS ON InSTI Rumi Minami 1 , Soichiro Takahama 1 , Kazuhiko Koyama 1 , Masahiro Yamamoto 1 1 National Hospital Organization Kyushu Medical Center, Fukuoka, Japan Background: Weight gain and psychiatric symptoms (PSs) have been reported in persons living with HIV (PLWH) receiving ART, especially integrase strand transfer inhibitors (INSTI). Obesity and PSs are correlated with chronic inflammatory states characterized by insulin resistance. Resistin is a pro- inflammatory adipokine and plays a key role in the insulin responsiveness of peripheral tissues and the central nervous system. A single-nucleotide polymorphism (SNP) at −420C>G in the resistin gene is correlated with serum resistin level in Japanese people. We clarified the influence of SNP −420C>G on weight gain and PSs in PLWH receiving INSTI. Methods: Participants were PLWH who started ART with INSTI (n=220) or protease inhibitors (PI; n=62). Body mass index (BMI) was measured before and 6 months after starting ART. PSs was evaluated with the Profile of Mood States or Self-rating Depression Scale. SNP −420C/G was investigated using PCR. Linear regression analysis was used to assess factors associated with BMI change and PSs. We examined several variables in addition to SNP −420C/G that may affect BMI and PSs. Variables that were significant in univariate analyses were incorporated in multivariate models. Results: BMI increased by 0.86 kg/m 2 (p<0.001, paired t-test) in the INSTI group and 0.33 kg/m 2 (p=0.04, paired t-test) in the PI group. In total, 24% of the INSTI group and 14% of the PI group showed PSs. The simple regression

CROI 2020 244