CROI 2017 Abstract e-Book

Abstract eBook

Poster and Themed Discussion Abstracts

Poster and Themed Discussion Abstracts

976 COMBINING MEASURES OF ADHERENCE CAN ASSESS PATTERNS OF PREP DRUG-TAKING AndrewM. Abaasa 1 , Craig Hendrix 2 , Monica Gandhi 3 , Peter Anderson 4 , Anatoli Kamali 5 , Eduard Sanders 6 , Gaudensia Mutua 7 , Frances Priddy 5 , Haberer Jessica 8 1 MRC/UVRI Uganda Rsr Unit on AIDS, Entebbe, Uganda, 2 The Johns Hopkins Univ, Baltimore, MD, USA, 3 Univ of California San Francisco, San Francisco, CA, USA, 4 Univ of Colorado, Aurora, CO, USA, 5 Intl AIDS Vaccine Initiative, New York, NY, USA, 6 Kenya Med Rsr Inst, Kilifi, Kenya, 7 Univ of Nairobi, Nairobi, Kenya, 8 Cntr for Global Hlth, Massachusetts General Hosp, Boston, MA, USA Background: Measuring patterns of drug-taking with PrEP remains a challenge. In phase 1/2 PrEP trials in East Africa, we use non-pharmacokinetic (non-PK) and pharmacokinetic (PK) adherence metrics in combination to assess patterns of drug-taking. Methods: From Oct 2009 to Mar 2010, the International AIDS Vaccine Initiative (IAVI) randomized participants to daily tenofovir, disoproxil fumarate(TDF)/emtricitabine(FTC) or placebo or intermittent TDF/FTC or placebo in Uganda (Ug) and Kenya (Ke). Eligible participants were HIV-negative aged 18-49 years, followed at Weeks 1, 2, 4 and then monthly for 4 months. Adherence was measured by 5 metrics: Electronic Monitoring (EM), Self-Report (SR), and concentrations of drug levels in plasma & cryopreserved peripheral blood mononuclear cells (PBMC) (short-term) and hair (long-term). The EM and SR data were averaged over the duration of exposure represented by the 3 PK measures. We categorized adherence as low, moderate and high (table) by each metric and assessed patterns of drug-taking via the 3 PK measures. A discriminant (C-statistic) analysis for single or combined methods for adherence measuring was assessed by logistic regression models with outcome as high versus moderate or low and displayed by receiver operating characteristic (ROC) curve area. Results: The analysis involved 48 participants with a mean age of 29.2 (SD=6.9). Discrimination for the EMmeasure was poor for SR (AROC 0.53) and best for the hair (AROC 0.87). A short-term PK metric (PBMC) did not improve discriminant ability. Using 3 measures, the highest percent of variability in EM adherence was explained with SR plus hair and PBMC in Ug (24.6%) and with SR plus plasma and hair in Ke (63.5%). The combination of the 3 PK metrics in Ug revealed high adherence in the short-term and low adherence in the long- term. In Ke, combining the 3 metrics showed high adherence just prior to study visits, but low adherence prior to that. Conclusion: SR adherence in PrEP is of low utility. Combining short-term (plasma) and long-term (hair) PK metrics of adherence can reveal patterns of drug-taking during PrEP. 977 SUBSTANCE-USING MSM ON HIV PREEXPOSURE PROPHYLAXIS HAVE BETTER ADHERENCE Martin Hoenigl 1 , Sonia Jain 2 , David J Moore 1 , Deborah Collins 3 , Xiaoying Sun 2 , Peter L. Anderson 4 , Katya Calvo 5 , Michael Dube 6 , Sheldon Morris 1 , for the California Collaborative Treatment Group (CCTG) 595Team 1 Univ of California San Diego, San Diego, CA, USA, 2 Univ of California San Diego, La Jolla, CA, USA, 3 Long Beach HHS, Long Beach, CA, USA, 4 Univ of Colorado Denver, Aurora, CO, USA, 5 Los Angeles Biomed Rsr Inst at Harbor–UCLA Med Cntr, Torrance, CA, USA, 6 Univ of Southern California, Los Angeles, CA, USA Background: The effectiveness of tenofovir/emtricitibine (TDF/FTC) for HIV preexposure prophylaxis (PrEP) strongly depends on maintaining adherence. We hypothesized that among men who have sex with men and transgender women enrolled in a randomized controlled PrEP trial, substance users would have lower levels of PrEP adherence. Methods: CCTG 595 was a randomized controlled trial of individualized texting versus standard care for adherence to daily TDF/FTC. We examined alcohol and substance use over 48 weeks for association with dried blood spot (DBS) intracellular tenofovir diphosphate (TFV-DP) levels at weeks 12 and 48 (i.e., composite outcome, see Table; cutoff ≥719 fmol/ punch that approximates ≥4 doses in the past week). Substance use was assessed (for the past 3 months) at all study visits using a SCID screening questionnaire for “No use”, “Some use” (1-4 times) and “Heavy use” (5 or more times) of any substance of abuse combined (marijuana and alcohol excluded) and also for each substance separate. Problematic use was assessed using the DAST-10 and AUDIT. We also assessed whether alcohol and substance use impacted study completion and incident sexually transmitted infections (STIs). Fisher’s exact test and logistic regression were used. Results: Of 394 subjects at baseline, any substance use was reported by 73% and alcohol use by 83% of participants. Overall, 71% of the 394 participants had TFV-DP levels ≥719 fmol/punch at week 48. Ongoing “Heavy” substance use (any) and “Some” or “Heavy” alcohol use were significantly associated with better adherence in logistic regression (Table), while problematic use had no significant impact on adherence. No particular substance contributed more to this association and notably METH users did not have worse adherence than non-METH users. In general, intensity of alcohol or substance use at baseline was not associated with study completion. Any substance use, but not alcohol use, was strongly associated with incident STI on study (Odds ratio of 2.5 and 2.6 for “Some” and “Heavy” use compared to “No” use; p<0.001).

CROI 2017 423