CROI 2015 Program and Abstracts

Abstract Listing

Poster Abstracts

704 Occult HBV/HIV Coinfection and Validation of Cost-Effective NAT Pooling PCR Shanmugam Saravanan 1 ; Janardhanan Mohanakrishnan 1 ;Thongadi Ramesh Dinesha 1 ; Jayaseelan Boobalan 1 ; Pachamuthu Balakrishnan 1 ; Kailapuri G Murugavel 1 ; Sunil S Solomon 2 ; Suniti Solomon 1 ; Davey M. Smith 3 1 YRG Centre for AIDS Research and Education, Chennai, India; 2 Johns Hopkins University School of Medicine, Baltimore, MD, US; 3 University of California San Diego, San Diego, CA, US Background: Owing to the cost of HBV viral loads, in most resource-limited settings (RLS) only HBsAg testing is done. This leads to under diagnosis of HBV infection by ignoring occult HBV infection (OHI). Thus, we studied the prevalence of HBV and OHI among HIV infected individuals and evaluated a pooling HBV nucleic acid test (NAT) as a cost-effective alternative to conventional HBV viral load test Methods: ART-naïve, 502 HIV-1 infected individuals visiting YRG CARE were enrolled; median age was 36 years (IQR 31-40) and CD4 count was 381.5 cells/ m L (IQR 232.5-540.25). HBV infection was detected by HBsAg ELISA (Genedia, Korea) followed by anti-HBc, anti-HBs ELISA (DiaSorin, Italy) and HBV DNA PCR was performed among all HBsAg negative individuals to find OHI. Pooling HBV NAT was performed in a matrix of 5x5 pooled samples, i.e. 25 plasma samples, and was evaluated against the gold standard HBV viral load (Abbott HBV RealTime PCR, linearity 1 - 9 log IU/mL) for cost and accuracy Results: The prevalence of HIV/HBV co-infection by HBsAg positivity was 6% (32/502, 95% CI 4.5%-8.86%). To test for OHI, we randomly selected 270 HBsAg negative individuals and performed HBV viral loads, anti-HBc and anti-HBs ELISA and found an overall prevalence of OHI as 10% (27/270, 95% CI 6.81%-1.44%) based on HBV DNA positivity, of which 96.3% (26/27) and 25.9% (7/27) were positive for anti-HBc and anti-HBs, respectively. This shows OHI is associated with anti-HBc alone (OR 9.46, p 0.1846) compared to anti-HBs. Compared to individual HBV viral loads, pooling HBV NAT had a positive predictive value of 100% and negative predictive value of 96.43%, with 100% specificity and 66.67% sensitivity. There would also have been cost (37$ vs. 91$) saving with pooling HBV NAT Conclusions: The prevalence of OHI is high among HIV-infected individuals in India, which would be missed by routine serology tests, thus HBV DNA testing should also be considered. However, conventional HBV viral loads are expensive in RLS, thus pooling HBV NAT should be considered, as it could reduce assay costs by up to 40% 705 Invariant Natural Killer T-Cells in HIV-HBV Coinfection Matteo Basilissi 1 ; CamillaTincati 1 ; Esther Merlini 1 ; Elisabetta Sinigaglia 2 ; Javier Sanchez-Martinez 2 ; Giovanni Carpani 2 ; Antonella d’Arminio Monforte 1 ; Laura Milazzo 1 ; Giulia Marchetti 1 1 University of Milan, Milan, Italy; 2 San Paolo Hospital, Milan, Italy Background: HIV+HBV+ co-infected patients (pts) are at risk of hepatic disease progression, despite effective treatment. The dysregulation of the immune response in HIV may affect the course of HBV disease. Given that invariant Natural Killer T (iNKT) cells mediate viral liver disease and HIV infection, we hypothesized a role for this cell subset in HIV+HBV+ co-infection. Methods: Cross-sectional study: a) HIV+HBV+ (10 pts treated with TDF/FTC; 4 naïve pts); b) HIV+ (10 treated pts, 5 naive pts); c) HBV+ (9 treated pts; 5 naïve pts); d) HIV-HBV- (5 pts). iNKT frequency (CD3+V α 24+CD1d+), CD161 expression and TNF- α /IFN- γ release prior to and after PMA/Ionomycin and α GalactosylCeramide ( α GalCer) were measured by flow cytometry. Kruskal-Wallis and Wilcoxon tests were used for statistics. Results: In HIV+HBV+, a negative correlation was found between AST levels and cytokine release upon α GalCer (TNF- α : p=0.014, r=-0.65; IFN- γ : p=0.06, r=-0.53). No other significant correlations were found between iNKT phenotype/function and liver function enzymes in the other groups. To better understand the uniqueness of iNKT cells in HIV+HBV+ co-infection versus HIV and HBV infections alone, we conducted a comparative analysis of iNKT cell frequency, phenotype and function among study groups according to their treatment status (i.e. therapy-naïve and treatment-experienced subjects). Therapy-naïve (mono-infected HBV or HIV and co-infected) showed similar circulating iNKT cells (p=0.46), CD161 expression (p=0.3), IFN- γ (unstimulated, p=0.6; PMA/Iono p=0.8; α GalCer p=0.07) and TNF- α production (us, p=0.09; PMA/Iono p=0.4; α GalCer, p= 0.9) to HIV-HBV- uninfected controls. Treated subjects showed comparable iNKT frequencies (p=0.8) and CD161 expression (p=0.9) among study groups. Interestingly, HIV-infected individuals showed a peculiar functional capacity of iNKT cells. Indeed, mono-infected HIV+ showed the highest constitutive release of TNF- α (77%, IQR; 69-86; p=0.008) and PMA/Iono-stimulated IFN- γ production (85%, IQR: 77-94; p=0.009). Further, only iNKT cells from co-infected HIV+HBV+ significantly increased TNF- α release after α GalCer (p=0.04). No statistical differences were registered in terms of iNKT function in HBV mono-infection. Conclusions: The inverse correlation between iNKT and liver function in HIV+HBV+ suggests the impairment of circulating iNKT cells in response to increased liver inflammation. Anders C. Boyd 1 ; Karine Lacombe 1 ; Fabien Lavocat 2 ; Sarah Maylin 3 ; Patrick Miailhes 4 ; Caroline Lascoux-Combe 5 ; Constance Delaugerre 3 ; Pierre-Marie Girard 1 ; Fabien Zoulim 2 1 Inserm UMR_S1136, Paris, France; 2 Inserm U1052, Lyon, France; 3 Inserm U941, Paris, France; 4 Hospices Civils de Lyon, Lyon, France; 5 Hôpital Saint-Louis AP-HP, Paris, France Background: Covalently closed circular DNA (ccc-DNA) of hepatitis B virus (HBV) acts as a reservoir for reactivation of viral replication and whose quantification can be used as a marker of persistent intracellular replication. The determinants of intracellular levels of replication have rarely been evaluated in HBV-human immunodeficiency virus (HIV) co-infected patients. Methods: Sixty HIV-HBV co-infected patients with at least one liver biopsy during follow-up in the French HIV-HBV cohort were included. HBV ccc-DNA and total intracellular HBV-DNA were extracted from biopsies and quantified by real-time PCR. Risk-factors of intracellular replication were determined using mixed-effect linear regression models. Results: At the time of biopsy, 35 (61.4%) patients were HBeAg-positive and 23 (46.9%) had detectable serum HBV-DNA (median: 3.10 log10 IU/mL, IQR:2.75-5.38). Among the 22 patients undergoing tenofovir (TDF)-containing antiretroviral therapy, cumulative TDF-duration was at a median 17.8 months (IQR:5.7-31.0). Overall, median HBV ccc-DNA was -1.10 log10 copies/cell (IQR:-1.70, -0.29) and total intracellular HBV-DNA was 0.27 log10 copies/cell (IQR:-0.39, 2.00). In multivariable analysis, patients with HBeAg-positive serology had significantly higher levels of HBV ccc-DNA (+0.76 log10 copies/mL; 95%CI:0.39, 1.13; p <0.001), whereas those with a nadir CD4+ cell count above 250/mm3 had significantly lower HBV ccc-DNA levels (-0.57 log10 copies/mL; 95%CI:-0.95, -0.19; p =0.004). Furthermore, patients with longer than 3 years of cumulative TDF-duration had significantly lower HBV ccc-DNA levels after adjustment (-0.88 log10 copies/cell: 95%CI:-1.40, -0.35; p =0.001). Accordingly, when focusing on patients undergoing TDF with a biopsy at TDF-initiation and sometime during therapy (median duration: 35.3 months, range: 20.2-56.6), most exhibited strong declines in HBV ccc-DNA (median change in log10 copies/cell/year:-0.46, range:-0.67, 0; n =7). HBV ccc-DNA levels did remain detectable at the end of follow-up for all patients, yet at very low levels (median: 0.04 copies/cell, range:0.01, 0.31). The results above were similar when using total intracellular HBV-DNA levels as an end-point. Conclusions: In coinfected patients, severe immunosuppression is associated with intracellular HBV replication. Treatment with TDF is linked to large declines in ccc-DNA, yet replication within the hepatocyte still persists after long periods of treatment. HIV, yet not HBV, appears to be the main driver of iNKT activation in HIV+HBV+ co-infected subjects on treatment. 706 Effect of Immunosuppression and Antivirals on Intracellular HBV Replication in HIV-HBV Coinfection

Poster Abstracts

435

CROI 2015