CROI 2015 Program and Abstracts

Abstract Listing

Poster Abstracts

TUESDAY, FEBRUARY 24, 2015 Session P-N1 Poster Session

Poster Hall

2:30 pm– 4:00 pm Natural History and Prognosis of HCV Infection 636 Progression of Liver Disease in LHIV/HCV Coinfected People According to Gender in Icona Cohort: Role of Age as Potential Different Exposure to Estrogens Antonella Cingolani 1 ; Paola Cicconi 2 ; GloriaTaliani 3 ; Alessandro D. Cozzi-Lepri 4 ; Massimo Puoti 5 ; Carmela Pinnetti 6 ; Pier LuigiViale 7 ; Antonella d’Arminio Monforte 2 for Icona Foundation Study Group 1 Catholic University, Roma, Italy; 2 University of Milano, Milano, Italy; 3 University La Sapienza, Roma, Italy; 4 University College London Medical School, Royal Free Campus, London, United Kingdom, London, United Kingdom; 5 Niguarda Hospital, Milano, Italy; 6 National Institute for Infectious Diseases L. Spallanzani, Roma, Italy; 7 Univeristy of Bologna, Bologna, Italy Background: In HCV patients, female gender was reported as protective in liver fibrosis progression. However, post menopausal women show an increased rate of progression, suggesting a causative role of estrogens. Aim of the study was to evaluate the effect of age (proxy of menopausal status), on liver disease progression in a large HIV/HCV population. Methods: All patients (pts) enrolled in the ICONA Foundation study with at least 1 HCVAb+ test available were included. Predictors of liver fibrosis (FIB-4>3.25) at baseline (first HCVab+ test available) were identified by logistic regression (gender; age, risk factor for HIV HBsAg, HCV genotype, CD4, baseline HIV RNA, years from HIV diagnosis, calendar year of enrollment). The effect of aging was examined by comparing Fib4 trends before and after age 50 in men and women, using a two-piecewise random coefficient model. Results: 2833 pts were examined: 739 (26%) women, median age 36 y (IQR 33-40), 6 years from HIV diagnosis (IQR 1-11), median CD4/ml 410 (IQR 221-600). Overall, 15.4% showed liver fibrosis at baseline, with a significantly lower proportion in females when compared to men (11% vs. 17%, p=<.0001). However after adjusting for potential confounders, the protective role of female gender on fibrosis was not confirmed (female AOR 0.8, 95%CI 0.4-1.3). Older age (AOR 2.0, 95%CI 1.4-2.8, +10 years), HBsAg+ (AOR 3.3, 95%CI 1.3-7.9), IVDU (AOR 1.8, 95%CI 1.0-3.4) and CD4 count (AOR 0.9, 95%CI 0.8-0.9 +100 cells/ml) were predictors of fibrosis at baseline. For the analysis of trends, 45498 FIB-4 measurements were analyzed. Median FIB-4 at enrollment was 1.15 ( IQR 0.77-1.88) in women and 1.53 (IQR 0.98-2.58) in men (Wilcoxon test p<.0001). In females, FIB-4 did not significantly change before (estimate 0.13/10 y, p=.47) or after 50 y (estimate 0.13/10 y, p=.38) and no change in trend is observed (difference .006, p=.97). In males, FIB4 significantly increased after 50 y (estimate 0.11/10 years, p=.009); but no change in trend before and after the age cut-off was found (difference in trend 0.04, p=.31). Conclusions: Differently fromwhat reported for HCV monoinfected female population, fibrosis progression in HIV/HCV females seems to be a quite linear phenomenon similar to what observed in the male counterpart. This finding suggests the pathogenetic relevance of potential pathways to inflammatory process other than estrogens deprivation in HIV/ HCV coinfected women 637 A Prognostic Score Estimating the Risk of Liver-Related Death Among HIV/HCV Coinfected Subjects Daniel Grint 1 ; Lars Peters 2 ; Jürgen Rockstroh 3 ; Karine Lacombe 4 ; Andrzej Horban 5 ; Irina Khromova 8 ; Jose Gatell 6 ; Antonella d’Arminio Monforte 7 ; Jens D. Lundgren 2 ; Amanda Mocroft 1 on behalf of EuroSIDA in EuroCoord 1 University College London, London, United Kingdom; 2 University of Copenhagen, Copenhagen, Denmark; 3 University Hospital Bonn, Bonn, Germany; 4 Hospital Saint Antoine, Paris, France; 5 Wojewodzki Szpital Zakazny, Warsaw, Poland; 6 University of Barcelona, Barcelona, Spain; 7 Clinica delle Malattie Infettive e Tropicali, Milan, Italy; 8 Centre for HIV/AIDS & Infectious Diseases Prevention & Control, Kaliningrad, Russian Federation Background: Development of a simple, widely applicable prognostic score for liver-related death (LRD) among HIV/HCV coinfected individuals is essential to predict the future burden of LRD. An accurate prognostic score in this setting will aid comparison of individual risk profiles when prioritising who should be treated with new directly-acting antivirals for HCV. Methods: EuroSIDA patients with HCV coinfection and follow-up after 1/1/2000 were included. Causes of death were classified using CoDe methodology. Fibrosis staging is estimated from liver biopsy, Fibroscan®, APRI and hyaluronic acid. Competing risks Cox proportional hazards modelling with stepwise variable selection was used to identify factors associated with long term risk of LRD. Scaled model coefficients were used to create a prognostic score for LRD. Results: 669 deaths were recorded (159 liver-related) from 3637 coinfected individuals during 14613 PYFU to November 2013. The study population was mostly Caucasian (94%), male (68%), injecting drug users (71%). At baseline the majority had F0/F1 fibrosis (76%), followed by F2/F3 (15%) and F4 (9%). Factors associated with LRD that contribute to the prognostic score are age, CD4 cell count, liver fibrosis stage, HBV coinfection, known duration of HCV infection and cART status ( Figure 1 ). The overall mean score was 5.0 (95% CI 4.9-5.1), but significantly higher among those who died of LRD (8.1 (7.6-8.6)), P <0.0001. A 1-unit increase in the score is associated with 2.6-fold increased risk of LRD (Hazard ratio: 2.6 (2.3-3.0; P <0.0001)). Increasing CD4 cell count from 100-300 cells/mm 3 to >500cells/mm 3 is associated with a 6.8-fold (5.1-9.0) reduced risk of LRD. Prediction of LRD based solely on fibrosis staging achieved an area under the ROC curve (AUROC) of 0.71, whereas the score achieved an AUROC of 0.83. A score cut-off of 6.5 provided the highest combination of sensitivity (71.1%) and specificity (83.1%). The 5-year probability of LRD increased from 1.6% (1.0-2.2) in those at low risk to 3.2% (2.1-4.8), 12.5% (9.4-16.2) and 24.2% (17.8-31.0) in those at medium-low risk, medium-high risk and high risk of LRD respectively, P <0.0001 for separation between strata ( Figure 1 ).

Poster Abstracts

400

CROI 2015