CROI 2015 Program and Abstracts

Abstract Listing

Poster Abstracts

Results: The frequency of Th17 cells ex vivo was significantly increased in HIV-1 + (n=15) as compared to HIV-1 - (n=15; p=0.0001). In preliminary experiments, the median level of IL-23 production in HIV-1 + (n=6) did not reach statistical significance over HIV-1 - (n=6) after in vitro stimulation with E. coli (p=0.39) or P. aeruginosa (p=0.24). However, the correlation between the level of IL-23 production and the frequency of Th17 cells in HIV-1 + was significant after stimulation with E. coli (n=6; p=0.01) and P. aeruginosa (n=6; p=0.02) stimulation. Conclusions: We found increased frequencies of peripheral blood Th17 cells in chronic HIV-1 + patients. This may be due to elevated cytokine production in response to translocated bacteria, as suggested by the relationship between IL-23 levels and Th17 frequencies after bacterial stimulation in HIV-1 + . Th17 cells are preferentially depleted in the intestinal mucosa of HIV-1 + , likely due to infection and killing of these cells. Our data suggest a potential mechanism by which the innate immune response to HIV-1-induced microbial translocation could promote the expansion, activation, viral infection and destruction of Th17 cells in the gut, thus contributing to HIV-1 disease pathogenesis. 330 Effect of Methamphetamine Use on T-Cell Proliferation In Vivo and Ex Vivo Marta Massanella 4 ; Sara Gianella 4 ; Jennifer M. Dan 1 ; Eric Daar 2 ; Michael P. Dube 3 ; Richard H. Haubrich 4 ; Douglas D. Richman 4 ; Davey M. Smith 4 ; Sheldon Morris 4 ; Rachel D. Schrier 4 1 La Jolla Institute of Allergy and Immunology, La Jolla, CA, US; 2 Los Angeles Biomedical Research Institute at Harbor–University of California Los Angeles Medical Center, Torrance, CA, US; 3 University of Southern California Keck School of Medicine, Los Angeles, CA, US; 4 University of California San Diego, La Jolla, CA, US Background: Methamphetamine (METH) is a widely used recreational drug among HIV-infected men who have sex with men (MSM). Its use is associated with worse health outcomes in HIV-infected individuals. Here we investigated the effect of METH-use on immune function and virus shedding in semen. Methods: Paired PBMC and semen samples were collected from our in vivo cohort of chronically HIV-infected MSM on suppressive ART (n=50, of whom 16 self-reported METH-use). PBMC were analyzed for markers of immune activation (CD38, HLA-DR and CD45RA), proliferation (Ki67) and exhaustion (PD-1) by flow cytometry. Seminal levels of cytomegalovirus (CMV) and HIV were quantified by real-time PCR. Ex vivo studies were performed on 19 HIV-infected individuals from a treated, but not uniformly suppressed cohort who tested positive for METH by urine toxicology (UTox+) and 19 HIV-infected UTox negative controls (matched for viral load and CD4 T-cell counts). PBMC proliferative responses to mitogen PHA and to CMV, candida, Mycobacterium tuberculosis, toxoplasma and HIV antigens were assayed in triplicate after 7 days of culture. T-cell proliferation was measured by [ 3 H]-thymidine incorporation; stimulation index was calculated as a ratio of the mean counts per minute (cpm) for each stimulus divided by the mean cpm of unstimulated controls. Mann-Whitney and Fisher exact tests were used for continuous and dichotomous data comparisons, respectively. Results: METH-users had significantly higher CD4 and CD8 T-cell proliferation levels (Ki67 + , p<0.005 for both), CD4 T-cell activation (CD45RA – CD38 + , p=0.005) and CD4-T cell exhaustion (PD-1 + , p=0.0004) compared to non-METH-users. In addition, the proportion of CMV or HIV shedding in METH-users was higher than in non-METH users (75% vs 26% [p=0.002] and 19% vs 3% [p=0.09], respectively). Ex vivo responses of the acute METH UTox+ patients confirmed higher spontaneous PBMC proliferation compared to controls (p<0.05). However, METH UTox+ subjects had significantly reduced proliferation to PHA (p<0.0005) and pathogen antigens (all p<0.04) compared to UTox–. Conclusions: Our findings suggest that METH-use may activate and exhaust the immune system. Furthermore, METH might reduce the immune response to reactivating or invading pathogens leading to a loss of control of CMV and HIV replication, as suggested by an increase in CMV and HIV seminal shedding in the genital tract. Future studies should consider METH-use as a potential modulator of T-cell responses.

THURSDAY, FEBRUARY 26, 2015 Session P-C14 Poster Session 2:30 pm– 4:00 pm Dissecting Pathogenesis Through In Vivo Studies

Poster Hall

331 A Random Forest Approach to Define Immunological Thresholds for CD4 Recovery in HIV-Treated Individuals Josué Pérez-Santiago 1 ; Marta Massanella 1 ; Dan Ouchi 2 ; Elisabet Gómez 2 ; Cecilia Cabrera 2 ; Bonaventura Clotet 3 ; Eugenia Negredo 3 ; Julià Blanco 2

Poster Abstracts

1 University of California San Diego, La Jolla, CA, US; 2 IrsiCaixa Institute for AIDS Research, Badalona, Spain; 3 Lluita Contra la Sida Foundation, Germans Trias i Pujol University Hospital, Badalona, Spain Background: The failure to increase CD4 T-cell counts in some ART-suppressed patients has been related to low CD4 T-cell production, high activation and cell death. While some clinical data indicate that low CD4 threshold (i.e. 200 CD4 cells/ m L) may have strong impact on survival, the lack of an accepted standard definition of immune recovery hamper a proper clinical follow up and has led to discrepant results among different studies. Our aim is to use a supervised machine learning approach to evaluate different clinical definitions of immune recovery. Methods: In a cross-sectional, case-control study, 185 participants on suppressive ART (<50 copies/mL for ≥ 2 years) were evaluated for CD4 T-cell production, immune activation markers, soluble CD14, cell death, clinical and demographical variables. We use random forest classification (R statistical software) to investigate different definitions of immune recovery based on 1) CD4 T-cell counts (using cut-off values ranging from 250 to 600 cells/ m L) and 2) CD4 T-cell increase from CD4 nadir value (Delta CD4, using values ranging from 50 to 500 cell/ m L). Patients below or above the indicated values were considered as discordant, or concordant respectively. Results: Among all CD4 cut-off definitions, 400 cells/ m L cut-off segregated best concordant vs. discordant individuals with a balanced accuracy of 85%, specificity of 87% and sensitivity of 83%. Most important variables included intrinsic and total apoptosis of CD4 T cells, CD4 nadir, and the frequency of Fas + HLA-DR + CD4 T cells. When Delta CD4 definitions were used, the best classifier was obtained with an increase of 300 CD4 cells/ m l with a balanced accuracy of 79%, specificity of 77% and sensitivity of 81%. Surprisingly, the most important variable for delta CD4 classifications was CD8 absolute count, along with CD4 cell death (total, intrinsic apoptosis and necrosis) and CD4 activation, while CD4 nadir was not an essential defining parameter. Variables previously related to discordance (such as thymic production, time on ART or CD8 T-cell activation) did not show significant importance for any classification. Conclusions: The 400 cell/ m l cut-off definition classified better discordant and concordant patients than Delta CD4 classifications. As expected cut-off classification was strongly associated with destruction and activation of CD4 T cells. However, the gain of CD4 T cells is strikingly associated to CD8 absolute counts, corroborating the use of CD4/CD8 ratio as a measure of immune recovery.

261

CROI 2015