CROI 2017 Abstract e-Book

Abstract eBook

Oral Abstracts

Oral Abstracts

132 BONE DENSITY AND TRABECULAR BONE SCORE IMPROVE FRACTURE PREDICTION IN HIV+WOMEN Jingyan Yang 1 , Anjali Sharma 2 , Josh Aschheim 3 , Kathryn Anastos 2 , Mardge H. Cohen 4 , Elizabeth T. Golub 5 , Deborah Gustafson 6 , Daniel Merenstein 7 , Phyllis Tien 8 , Michael T. Yin 1 1 Columbia Univ Med Cntr, New York, NY, USA, 2 Albert Einstein Coll of Med, Bronx, NY, USA, 3 Arcadia Hlth, New York, USA, 4 John H.Stroger Jr. Hosp of Cook County, Chicago, IL, USA, 5 The Johns Hopkins Univ, Baltimore, MD, USA, 6 SUNY Downstate Med Cntr, Brooklyn, NY, USA, 7 Georgetown Univ, Washington, DC, USA, 8 Univ of California San Francisco, San Francisco, CA, USA Background: The FRAX algorithm predicts the 10-year risk of a major osteoporotic (MO) fracture at the spine, hip or forearm combined or at just the hip using clinical risk factors (CRF) alone or with addition of dual-energy xray absorptiometry (DXA) measures: femoral neck bone mineral density T-scores (FNT) and lumbar spine trabecular bone score (TBS), a measure of gray scale homogeneity that correlates with trabecular microarchitecture. FRAX based upon CRF alone underestimates fracture risk in HIV-infected adults. Our objective was to determine whether addition of FNT and TBS improves accuracy of fracture probability assessment in HIV-infected women enrolled in the Women’s Interagency HIV Study. Methods: We included 1148 women (900 HIV-infected and 248 uninfected) with age > 40 years, complete CRF data for FRAX calculation, and 10-year observational data for incident fragility fractures. 220 (19%) of the women had baseline DXA measurements of FNT and TBS. Accuracy of the FRAX calculation was compared by the observed/estimated (O/E) ratios of fracture in 3 models: CRF only; CRF and FNT; or CRF, FNT and TBS. Accuracy is perfect if O/E=1; O/E >1 indicates underestimation. Results: Mean age of the cohort was 47±6 years. HIV-infected women were more likely to be African American, non-smokers, and thinner than uninfected women, but less likely to report glucocorticoids or alcohol use. During the 10-year follow-up, observed fracture rates did not differ significantly in HIV-infected and uninfected women for MO (8.1% vs 5.2%, p=0.13) or hip fractures (2.7% vs 1.2%, p=0.24). FRAX using CRF was less accurate in HIV-infected than uninfected women for predicting MO (O/E=4.33 vs. 2.92, p<0.001) and hip fractures (O/E=17.96 vs 7.61, p<0.001). Among HIV-infected women, accuracy improved greatly when FNT was included in the FRAX calculation for MO (O/E=4.33 vs 3.50, p<0.001) and hip fractures (O/E=17.96 vs 2.74, p<0.001), and further improved when both FNT and TBS were included in the FRAX calculation (both p<0.05, Table). After addition of DXA measures, accuracy of FRAX no longer differed between HIV-infected and uninfected women (Table). Conclusion: Accuracy of FRAX is improved with addition of FNT and TBS to clinical risk factors in HIV-infected women. These observational data support existing guideline recommendations for DXA screening in HIV-infected adults over age 50; and incorporation of DXA data into fracture prediction with FRAX.

133 FRAILTY PROGRESSION AND RECOVERY AMONG PERSONS AGING WITH HIV AND SUBSTANCE USE Damani A. Piggott , Karen Bandeen-Roche, Shruti H. Mehta, Todd Brown, Huanle Yang, Jeremy D. Walston, Sean X. Leng , Gregory D. Kirk The Johns Hopkins Univ, Baltimore, MD, USA

Background: Frailty is a critical aging-related syndrome marked by diminished physiologic reserve and heightened vulnerability to stress, predictive of major adverse clinical outcomes in older adults. We have previously demonstrated that frailty burden is heightened with HIV infection and strongly associated with increased hospitalization and death among persons aging with HIV. Though frailty is considered dynamic, little data exist on the factors determining transitions between frailty states among HIV-infected or uninfected adults. Methods: Frailty was assessed semiannually among HIV-infected and uninfected persons with a history of injection drug use in the AIDS Linked to the IntraVenous Experience (ALIVE) cohort from 2005 through 2013 based on the 5 Fried physical frailty phenotype domains – weight loss, low physical activity, exhaustion, decreased grip strength, and slow

54

CROI 2017