CROI 2017 Abstract e-Book

Abstract eBook

Poster and Themed Discussion Abstracts

Conclusion: In our experience, DAA treatment strongly improved VACS Index, but did not impact on neurocognitive performance in HIV/HCV co-infected. These results could be explained by a poor contribution of HCV to neurocognitive impairment in HIV co-infected population, even though the elevated frequency of confounding factors in this highly vulnerable population may have masked benefit effect of DAA on neurocognition.

355 LIMITATIONS OF THE INTERNATIONAL HIV DEMENTIA SCALE IN THE CURRENT ERA Benedetta Milanini 1 , Emmanuel Bahemana 2 , Babajide Keshinro Keshinro 3 , Francis Kiweewa 4 , Rither Langat 5 , Winnie Rehema 6 , Isabel E. Allen 1 , Christina Polyak 7 , Julie Ake 7 , Victor Valcour 1 1 Univ of California San Francisco, San Francisco, CA, USA, 2 Walter Reed Prog–Tanzania, Mbeya, Tanzania, United Republic of, 3 Walter Reed Prog–Nigeria, Abuja, Nigeria, 4 Makerere Univ Walter Reed Proj, Kampala, Uganda, 5 KEMRI/Walter Reed Proj, Kericho, Kenya, 6 KEMRI/Walter Reed Proj, Kisumu, Kenya, 7 US Military HIV Rsr Prog, Bethesda, MD, USA Background: The International HIV Dementia Scale (IHDS) was developed as a tool for HIV dementia in both the industrialized and developing world. As initially described, a cut-point of 10 on this 12-point scale had sensitivity of 80%with specificity of 55% in Uganda. Recent publications from Uganda identify very high rates of probable HIV dementia (64%, BMC Psychiatry 2013) using this screening instrument, prompting us to examine performance characteristics for the current era. Methods: 2414 individuals from East Africa underwent testing with the IHDS and a 30-minute cognitive battery that included the World Health Organization (WHO) auditory verbal learning test (AVLT) trial 1, sum of 1-5, and recall; the Trails A test; the grooved pegboard test; and action fluency task. We defined impairment among HIV+ participants as - 1 SD on two tests or - 2 SD on one test when performance was compared to concurrently enrolled controls stratified by age (=35) and education (<6 years, 6-12 years, >12 years). We examined predictive capacity of the IHDS using receiver operator characteristic (ROC) curve. Psychometrists underwent initial certification with re-certification every 6 months. Results: We enrolled participants from Uganda (n=531), Kenya (n=1466) and Tanzania (n=417) with mean (SD) age for HIV+ (n=2009) and HIV-negative (n=405) groups: 39.8 (10.8) and 37.6 (10.5), respectively (p=0.006). Among HIV+, 1651 (67%) were on cART, 979 (51%) had plasma viral loads <50 copies/ml and 702 (36%) met criteria for impairment. The mean (SD) IHDS score was 8.5 (1.7) and 9.0 (1.6) for HIV+ and HIV-negative, respectively (p=0.001). Using the cut-point of 10, 1290 (64%) of HIV+ subjects would be classified as having dementia as well as 215 (53%) of HIV negative controls. The ROC area under the curve (AUC) was maximally 60% offering a sensitivity of 66% and specificity of 66% at a cut point of 9 among HIV+. Conclusion: The IHDS has poor performance characteristics for the identification of impairment in East Africa in the current era. Performance for the most severe form of impairment, HIV Dementia, typically constituting < 5% of patients with access to cART, cannot be assessed from these data. Our data raise concerns for continued use of the IHDS in the era of cART.

Poster and Themed Discussion Abstracts

CROI 2017 140