CROI 2015 Program and Abstracts

Abstract Listing

Poster Abstracts

Prevalence of high risk HPV types and presence of multiple types by cytology outcomes Conclusions: In the largest study of cervical HPV genotypes in HIV-infected women in the Americas, infection by high risk HPV types other than 16 or 18 and infection by more than one high risk type were common. Infection by more than one high risk type and CD4+ lymphocyte count were more strongly associated with abnormal cervical cytology than any individual high risk HPV type, highlighting the need for multi-valent HPV vaccine strategies. 857 Comparison of Three Female Genital Tract Sampling Techniques for HIV RNA Recovery Catherine S. Todd 1 ; Shameem Jaumdally 2 ; Heidi E. Jones 3 ; Hoyam Gamieldien 2 ; Nontokozo Langwenya 4 ; Landon Myer 4 ; Donald R. Hoover 5 ; Jo-Ann Passmore 1 1 FHI360, Bangkok, Thailand; 2 University of Cape Town, Cape Town, South Africa; 3 Hunter College, CUNY School of Public Health, New York, NY, US; 4 University of Cape Town, Cape Town, South Africa; 5 Rutgers New Jersey Medical School, Piscataway, NJ, US Background: Although measuring HIV RNA shedding in women is an important element of HIV prevention research, there is no consensus on optimal female genital tract sampling methods for HIV RNA recovery, including the menstrual cup (MC). We hypothesized that MC specimens provide greater genital tract HIV RNA recovery than endocervical swab (ECS) or swab-enriched cervicovaginal lavage (eCVL). Methods: This study, nested within a clinical trial assessing intrauterine device safety and acceptability for HIV-infected women with CD4>350 cells/mm 3 , collected MC, ECS, and eCVL genital tract samples from consenting women during enrollment (sampling order MC/ECS/eCVL), 3-month (ECS/eCVL/MC), and 6-month (eCVL/MC/ECS) follow-up visits. At all visits, women self-inserted the MC for >45 minutes prior to clinician removal. ECS was collected with flocked swabs; eCVL combined an additional ECS with 5 mL saline lavage. At processing, MC secretions were weighed then diluted at 1 gm/10-fold mL phosphate-buffered saline. Matched aliquot (1 mL) sets of MC, ECS, and eCVL samples from each woman/visit were tested for HIV RNA concentration with PCR. Paired comparisons between methods were analyzed using McNemar’s exact tests for dichotomous outcomes and sign-rank test for medians of continuous measures. Results: In 39 participants, (39 enrollment, 20 3-month, and 6 6-month samples) women had a median age of 30 years (IQR 26-33). Median baseline plasma VL was 3.91 log 10 copies/mL (range: non-detectable - 5.35 log 10 copies/mL). Median MC sample weight was 0.34 gm (range: 0.02-1.52 gm). MC samples with detectable HIV RNA had greater volume (0.37 gm (IQR: 0.29, 0.61), n=40 vs. 0.26 gm (IQR: 0.19, 0.38), n=23; unadjusted Mann Whitney, p=0.02), but did not differ by median insertion time (both groups=95 minutes). MC samples were significantly more likely to have detectable HIV RNA than ECS and eCVL at enrollment (63% vs. 43% and 44%, respectively) and higher median HIV RNA levels than ECS or eCVL at both enrollment and 3 month follow-up (Table). In pooled analysis of women with quantifiable HIV RNA, the unadjusted median difference between MC and eCVL VL was 0.6 (IQR 0.3, 0.8, p<0.001) log 10 copies/mL, 0.8 (IQR 0.3, 1.2, p<0.001) between MC and ECS, and -0.2 (IQR -0.8, 0.3, p=0.31) between ECS and eCVL.

Poster Abstracts

517

CROI 2015