CROI 2017 Abstract e-Book

Abstract eBook

Poster and Themed Discussion Abstracts

the 2 year follow-up, HAND stage frequencies (%) for HIV+ individuals on CART were 3% for ANI, 9% for MND, and 2% for HAD (p<0.001 compared to the matched pre-CART HAND stages of the same 233 HIV+ individuals). HAND stage improved in 44% of HIV+ individuals and worsened in 8% of HIV+ individuals on CART. Conclusion: Individuals with HIV subtypes D and A showed significant improvement in HAND stage after 2 years of CART treatment compared to their pre-CART status, with a marked decrease in the frequency of HAD. These are similar to results in the US. CART should be initiated in early HIV infection to treat and potentially prevent neurocognitive complications in Sub-Saharan Africa. 360 ADDITIONAL SCREENING TESTS FOR HIV-ASSOCIATED NEUROCOGNITIVE DISORDERS Mattia Trunfio 1 , Chiara Montrucchio 1 , Daniela Vai 2 , Daniele Imperiale 2 , Alessandro Livelli 2 , Giancarlo Orofino 3 , Alice Trentalange 1 , Stefano Bonora 1 , Giovanni Di Perri 1 , Andrea Calcagno 1 1 Univ of Torino, Torino, Italy, 2 Maria Vittoria Hosp, Torino, Italy, 3 Amedeo di Savoia Hosp, Torino, Italy Background: The diagnosis of HIV-associated neurocognitive disorders (HAND) relies on complete neurocognitive tests (NCT) that are time-consuming. Appropriate screening tests are lacking since both the three questions (3Qs, EACS Guidelines) and the International HIV Dementia Score (IHDS) are associated with poor sensitivity and intermediate/good specificity in different population groups. Methods: Consecutive HIV-positive patients screened for neurocognitive impairment were prospectively enrolled. Clinical and biochemical data were recorded. 3Qs, IHDS, Clock Drawing Test (CDT) and Frontal Assessment Battery (age and education adjusted, aFAB) were administered: 3Qs≥1, IHDS≤10, CDT>2 and aFAB≤14 were deemed abnormal. Patients showing abnormal IHDS or with open cognitive symptoms underwent full NCT including eight NC domains afterwards: HAND was diagnosed according to the Frascati’s criteria. Data are expressed as medians (interquartile ranges) and tested through non-parametric tests. Results: 669 patients were enrolled (87% on cART, 76%male, median age 50 years): screening tests were abnormal in 171 (26.6%), 266 (42.3%), 70 (15.4%) and 20 (7.9%) patients, according to 3Qs, IHDS, CDT and aFAB, respectively. 279 patients underwent full NCT (84% on cART, 72%male, median age 51 years). Plasma HIV-RNA was <50 copies/mL in 186 subjects (81.6%), median and nadir CD4+ T-cell/uL count were 516 (304-777) and 153 cells/uL (57-285). HAND was diagnosed in 141 patients (50.5%): 100, 32 and 9 subjects with ANI, MND or HAD. A significant correlation was observed between IHDS (p<0.0001) and aFAB (p<0.0001) with HAND diagnosis. Sensitivity, specificity and correct classification rate (CCR) of screening test for HAND diagnosis were as follows: 3Qs (36.5%, 56.4%, 46.4%), IHDS (72.8%, 55.7%, 64%), aFAB (24.5%, 100%, 61.9%), CDT (30.1%, 71.6%, 52.2%). Considering only symptomatic neurocognitive disorders, aFAB presented the highest CCR (84.9%). The concomitant use of aFAB or CDT with IHDS did not substantially improve screening tests’ accuracy. Conclusion: All screening tests showed incomplete accuracy in predicting HAND. While IHDS presented the highest sensitivity, aFAB was associated with a very high specificity and the highest correct classification rate for symptomatic HAND. The 10-minute, easy-to-administered, FAB test warrants further studies in HIV-positive patients. 361 COGNITIVE COMPLAINTS AND DEVELOPMENT OF FALLS AMONG HIV+ AND HIV- WOMEN Anjali Sharma 1 , Donald R. Hoover 2 , Qiuhu Shi 3 , Michael Plankey 4 , Phyllis Tien 5 , Kathleen M. Weber 6 , Michelle Floris-Moore 7 , Hector Bolivar 8 , Marica M. Holstad 9 , Michael T. Yin 10 1 Albert Einstein Coll of Med, Bronx, NY, USA, 2 Rutgers, the State Univ of New Jersey, Piscataway, NJ, USA, 3 New York Med Coll, Valhalla, NY, USA, 4 Georgetown Univ, Washington DC, USA, 5 Univ of California San Francisco, San Francisco, CA, USA, 6 Cook County Hlth & Hosps System, Chicago, IL, USA, 7 Univ of North Carolina at Chapel Hill, Chapel Hill, NC, USA, 8 Univ of Miami, Miami, FL, USA, 9 Emory Univ, Atlanta, GA, USA, 10 Columbia Univ, New York, NY, USA Background: We previously reported greater fracture rates in aging HIV+ vs. HIV- women in the Women’s Interagency HIV Study (WIHS). Because fracture may be due to reduced bone strength combined with greater risk of falls, we compared rates of falls and evaluated relationships between subjective cognitive complaints and fall in HIV+ and HIV- women Methods: We analyzed 1876 (1289 HIV+, 587 HIV-) women with 18 months of data on self-reported falls during the 6 months prior to each study visit. The primary exposure of interest was subjective cognitive complaints (self-reported major problems with memory or concentration, confusion, or inability to perform routine mental tasks). Hierarchical models evaluated associations between subjective cognitive complaints (and HIV status) and having any fall (vs. none), after adjusting for: (1) demographics, (2) co-morbid conditions, (3) substance use/CNS active medications, and (4) HIV-specific factors. Logistic regression models for prediction of falls were fit, with covariates associated with any fall in univariate analysis (p<0.1) included in multivariable models. Associations with falls did not vary across the three visits, thus visits were pooled together and generalized estimating equations with logit link adjusted for within-person correlation due to use of repeated measures. Results: HIV+ women were older than HIV- women (median 49 vs. 47yr, p<0.0001), and more likely to report neuropathy (21% vs. 14%, p=0.0003); 11% of all women reported cognitive complaints. On average, at least one fall was reported in 17.5% of HIV+ and 17.6% of HIV-women. HIV remained unassociated with incident falls in multivariate analyses. Subjective cognitive complaints were associated with increased odds of having any fall (Table models 1-2), however this association was reduced by 45% after fully adjusting for covariates associated with falls, in particular medical comorbidities (Table models 3-4). Similar patterns were seen in HIV+ women. Conclusion: HIV+ women did not have more frequent falls compared with HIV- women. Subjective cognitive complaints were associated with greater odds of having a fall; this risk was reduced by 45%with full adjustment of covariates. Among HIV+ women, the association between cognitive complaints and falls appears to be mediated by comorbid medical illness. Additional studies are needed to understand which comorbid illnesses are most influential and whether management of those conditions can prevent falls among aging HIV+ women.

Poster and Themed Discussion Abstracts

CROI 2017 143