CROI 2017 Abstract e-Book

Abstract eBook

Poster and Themed Discussion Abstracts

formation. Moreover, the risk of developing KS is increased in the context of immune reconstitution inflammatory syndrome. These observations suggest that KS might be fueled by immune activation. To test this hypothesis we studied whether patients previously diagnosed for KS present with a particular immune activation phenotype. Methods: The ACTIVIH study is a cross sectional study in which we analyzed 68 markers of immune, endothelial, and coagulation activation in 120 virologic responders, and have recently identified 5 different profiles of immune activation. We looked in the medical history of these patients for a relationship between past history of KS and immunological profile, using logistic regressions. Results: The frequency of past history of KS (n = 11) was significantly elevated in one patient group (profile 3), as compared with the other groups (p = 0.03, Figure 1). Of note, the frequency of past history of other cancers was not increased in the same group of patients (odds ratio 0.52 [95% CI 0.11-2.46], p = 0.52). Patients who had developed KS, as compared with those who had not, were characterized by a low percentage of naïve (CD45RA+CD27+) CD4+ T cells (29.3±16.0 versus 40.3±16.5, p = 0.04), a tendency to a high percentage of CD4+ T cells expressing the inhibitory receptor PD-1 (32.8±14.5 versus 40.1±9.5, p = 0.06), and a high percentage of senescent (CD57+) NK cells (42.9±17.6 versus 57.6±15.0, p < 0.01). Conclusion: Our data are compatible with the hypothesis that a particular type of chronic immune activation paves the way for KS. The marks of CD4+ T cell activation and NK senescence we unveiled in KS patients are consistent with (i) the previous report that the supernatants of activated T cells transform normal endothelial cells into KS cells, and (ii) the link between KS and immunosenescence. If our data are confirmed by a longitudinal study, the immune activation profiling of HIV patients, might help to identify those at risk to develop KS. Furthermore, deciphering which immune activation pathway(s) fuels the transformation of endothelial cells into KS cells might identify future therapeutic targets.

Poster and Themed Discussion Abstracts

606 TREATMENT AND OUTCOMES OF NON–SMALL-CELL LUNG CANCER IN LATER ART-ERA HIV INFECTION Keith M. Sigel 1 , Kristina Crothers 2 , Sheldon T. Brown 3 , Cynthia L. Gibert 4 , Matthew B. Goetz 5 , Roger Bedimo 6 , Roxanne Wadia 7 , Maria Rodriguez-Barradas 8 , Fatma Shebl 9 , Lesley Park 10 1 Icahn Sch of Med at Mt Sinai, New York, NY, USA, 2 Univ of Washington, Seattle, WA, USA, 3 James J. Peters VA Med Cntr, Bronx, NY, USA, 4 Washington, DC, VA Med Cntr, Washington, DC, USA, 5 VA Greater Los Angeles Hlth Care System, Los Angeles, CA, USA, 6 VA North Texas Hlth Care Cntr, Dallas, TX, USA, 7 VA Connecticut Hlthcare System, West Haven, CT, USA, 8 Baylor Coll of Med, Houston, TX, USA, 9 Yale Univ, New Haven, CT, USA, 10 Stanford Med, Stanford, CA, USA Background: Studies evaluating lung cancer outcomes in HIV+ persons have found worse survival which has been attributed to treatment disparities or more aggressive cancer behavior. We used population-based data to compare treatment trends and outcomes among HIV-infected (HIV+) and uninfected Veterans diagnosed with non-small cell lung cancer (NSCLC). Methods: We linked the Veterans Aging Cohort Study with national Veterans Affairs cancer registry data to identify 1,456 (581 HIV+) cases of incident NSCLC diagnosed during 2002-2015. We collected data on the first course of lung cancer treatment including surgical approach, use of radiotherapy and chemotherapy. We used Kaplan-Meier methods to compare median overall survival (OS) by HIV status and Cox regression to evaluate predictors of OS in HIV+ NSCLC patients. We also restricted these analyses to lung cancer patients diagnosed in the later antiretroviral (ART)-era (2009-2015; n=344 HIV+). Results: HIV+ NSCLC patients were younger than uninfected NSCLC patients (Table 1; p=0.01). Stage and histologic subtype did not differ by HIV status. HIV+ patients were equally likely to receive surgery (and did not differ in surgical approach) and radiotherapy as uninfected patients, but were less likely to be treated with chemotherapy (35% vs 45%; p<0.001), a difference that persisted when limiting to the later time period. Median OS was significantly worse for HIV+ patients than for uninfected patients (10 months vs 13 months; p<0.001) but in the later ART-era (2009-2015) there was no difference in OS (p=0.2). In the overall cohort, there was no difference in survival between stage I HIV+ and uninfected patients; and for early stage patients treated with surgical resection there was also no difference in cancer recurrence rates. Among HIV+ NSCLC patients who received stage-appropriate treatment older age, cancer stage, poorer viral suppression (>500 copies/ml) at cancer diagnosis, low CD4 (<200 cells/mm 3 ) and low CD4/CD8 ratio (<0.7) were all independently associated with worse survival (all p<0.05). Except for poorer viral suppression, these factors were all associated with worse OS when restricting the cohort to cancer cases diagnosed in the later ART-era. Conclusion: In the late ART-era, we found that HIV+ NSCLC patients experienced cancer treatment disparities, but had outcomes that were similar compared to uninfected patients, and that several HIV biomarkers were independently predictive of outcomes in NSCLC patients. 607 ETIOLOGY OF HEPATOCELLULAR CARCINOMA IN WEST AFRICA: THE HIV CONTRIBUTION Antoine Jaquet 1 , Boris Tchounga 2 , Aristophane tanon 2 , Aklesso Bagny 3 , Didier K. Ekouevi 4 , Hamar A. Traore 5 , Annie J. Sasco 1 , Moussa Maiga 6 , François Dabis 1 , for the IeDEAWest Africa Collaboration 1 Univ de Bordeaux, Bordeaux, France, 2 CHU de Treichville, Abidjan, Cote d’Ivoire, 3 CHU Campus, Lomé, Togo, 4 Univ de Lomé, Lomé, Togo, 5 CHU du Point G, Bamako, Mali, 6 Hôpital Gabriel Touré, Bamako, Mali Background: Hepatocellular carcinoma (HCC) is a leading cause of cancer in West Africa where Hepatitis B Virus (HBV) is endemic. However, limited information is available on other risk factors such as alcohol use, hepatitis C virus (HCV) and HIV infections. Methods: A case-control study was conducted in referral hospitals of Abidjan (Cote d’Ivoire), Bamako (Mali) and Lome (Togo). Cases were matched with two controls on age, gender and participating site. All participants underwent a standardized abdominal ultrasound and a serum α-fetoprotein measurement. The diagnosis of HCC was based on the combination of one or more space-occupying ultrasound lesions suggestive of an HCC and a serum α-fetoprotein level ≥400 ng/ml. HIV, HBV and HCV serology tests were systematically performed. Alcohol use was assessed using the AUDIT questionnaire (a score >7 defined hazardous drinking). A conditional logistic regression model estimated the associations by the Odds Ratio (OR) with its 95% confidence interval (CI). Results: A total of 160 HCC cases (Abidjan n=44, Lome n=40, Bamako n=76) and 320 controls were recruited. They were predominantly male (80.0%) and had a median age of 47 years (IQR 38 – 57). The prevalence figures of HBV, HCV and HIV infection were 9.7%, 4.7% and 2.8%, respectively in the control group; hazardous drinking was reported by 5% of them. In multivariate analysis, HBV, HCV and hazardous drinking were independently associated with HCC (table). Combining the effect of HBV and HIV infection in an additional model; HBV/HIV co-infected patients had an OR of 82.0 (CI 6.2-1 075.0) and HBV mono-infected had an OR of 59.9 (CI 19.6-182.6) (Ref: HBV/HIV-negative). Combining the effect of

CROI 2017 257