CROI 2019 Abstract eBook

Abstract eBook

Poster Abstracts

adjusted model, only DBS levels predicted spine BMD decline (p<0.001). All who developed low Z scores had detectable TFV, and most (57%) had high adherence (≥700 fmol/punch). The average mean decline in spine BMD was -1.15% (95% CI: -1.65, -0.64) for estimated daily adherence vs. -0.53% (95% CI: -1.06; 0.00) for 4-6 doses per week vs. -0.46% (95% CI:-1.31, 0.39) for 2-3 doses/wk vs -0.26% (95% CI: -0.81, 0.29) for <2 doses/wk, compared to a 0.72% increase (95% CI: 0.20; 1.25) in those who were not taking PrEP (TFV-DP below limit of detection) (Figure). Conclusion: We found approximately a 1% drop in BMD in highly adherent PrEP users over a median of 24 weeks, with a monotonic relationship between PrEP weekly pill-taking and degree of BMD loss. Future research should explore if dose-limiting strategies such as intermittent PrEP can decrease the risk of bone loss for individuals at higher risk.

1 KEMRI Wellcome Trust Rsr Prog, Kilifi, Kenya, 2 University of Washington, Seattle, WA, USA Background: There are no data on reduction of HIV-1 incidence following programmatic pre-exposure prophylaxis (PrEP) uptake by men who have sex with men (MSM) in sub-Saharan Africa. We assessed HIV-1 incidence and predictors of HIV-1 acquisition in at-risk MSM with access to PrEP in coastal Kenya. Methods: Since June 2017, at-risk MSM followed at monthly visits in an HIV-1 vaccine feasibility cohort study were offered PrEP with adherence and risk reduction counselling, monthly HIV-1 testing and X-pert RNA Qual testing if acute HIV-1 risk criteria were met. Participants who acquired HIV-1 and had documented seroconversion were offered immediate ART. MSM were categorized as taking PrEP if they received a PrEP refill at their previous visit and wanted to continue PrEP during their current visit. Those not receiving a PrEP refill in their previous visit or discontinuing, re-starting or starting PrEP during their current visit were categorized as not on PrEP. Participants who reported missing 6 or less days of PrEP since their last monthly refill were categorized as ≥80% adherent. We used population-averaged multivariable Poisson regression with robust variance estimation to identify predictors of HIV-1 acquisition, analyzing PrEP use defined as above, as well as based on ≥80% reported adherence. Results: Of 178 MSMwho were offered PrEP, 142 (79.8%) started, of whom 31 (17.4%) stopped during follow-up. 89.4% of PrEP users reported ≥80% adherence. During a median follow-up of 14.3 (interquartile range: 8.9–14.6) months, 7 MSM acquired HIV-1, for an incidence rate of 4.1 (95% confidence interval [CI] 2.0–8.7) per 100 person-years. Of the 7 MSM who acquired HIV-1, 4 were not taking PrEP and 3 were, including 2 who reported ≥80% adherence. In multivariable analysis, group sex (adjusted incidence rate ratio [aIRR] 9.9, 95% CI 1.4–68.2) and a recent gonorrhoea infection (aIRR 11.2, 95% CI 1.1–116.7) were independent predictors of HIV-1 acquisition, after adjustment for age, sexual orientation, and alcohol use. The aIRR for any PrEP use was 0.3 (95% CI 0.0–2.2). When ≥80% adherence was tested in the same model, the aIRR was 0.2 (95% CI 0.0–1.9). Conclusion: HIV-1 incidence among at-risk MSM with access to programmatic PrEP was high, and did not differ by PrEP use or reported adherence. A substantial proportion of MSM stopped taking PrEP despite frequent risk reduction counselling. Further research on PrEP adherence and tenofovir drug levels in this cohort is necessary. 946 DOSE-DEPENDENT DECLINE IN BONE MINERAL DENSITY BY LONG-TERM TFV EXPOSURE IN IPREX-OLE Matthew A. Spinelli 1 , David Glidden 1 , Peter L. Anderson 2 , Valdilea Veloso 3 , Vanessa McMahan 4 , Susan P. Buchbinder 1 , Robert M. Grant 1 , Monica Gandhi 1 , for the iPrEx OLE Study Team 1 University of California San Francisco, San Francisco, CA, USA, 2 University of Colorado, Aurora, CO, USA, 3 Instituto Nacional de Infectologia Evandro Chagas, Rio de Janeiro. Brazil, 4 University of Washington, Seattle, WA, USA Background: Oral tenofovir (TFV)-containing PrEP is associated with modest declines in bone mineral density (BMD). TFV-DP in dried blood spots (DBS) is a long-termmarker of TFV exposure well-suited for evaluating the relationship between exposure and toxicities. Characterization of BMD change by estimated weekly pill-taking can assist clinicians in counseling patients, and potentially support use of intermittent PrEP for those at high risk of toxicity. We evaluate the association of dosing frequency and BMD decline for the first time in a large PrEP demonstration project. Methods: Men who have sex with men and transwomen in the optional dual-energy X-ray absorptiometry (DXA) substudy of iPrEx OLE underwent DXA scans and DBS collection at baseline and every 24 weeks. TFV-DP levels were measured in DBS; average weekly dosing adherence was estimated from validated cut-offs. The mean % change in BMD was estimated in each strata of average weekly adherence using a linear mixed effects model. Results: DXA/DBS data were available for 254 individuals over a median of 24 weeks in iPrEx OLE from 6/11-12/13. Overall, the median age was 31 years and 9% identified as transwomen; 15%were Black, 38% Latino. At baseline, 9% had Z-scores <-2 at either spine or hip and 3% developed low Z-scores after starting PrEP. There was a dose-dependent % decline in spine BMD by strata of increasing average weekly adherence (p<0.001 trend); the p-value for trend using the hip BMD outcome was 0.07. When including age, race/ethnicity, gender, body mass index, smoking, stimulant use, and alcohol use in an

Poster Abstracts

947 LOW URINE TFV BY A NOVEL IMMUNOASSAY IS ASSOCIATED WITH HIV SEROCONVERSION ON PrEP Matthew A. Spinelli 1 , David Glidden 1 , Warren Rodrigues 2 , Guohong Wang 2 , Michael Vincent 2 , Catherine A. Koss 1 , Hideaki Okochi 1 , Karen Kuncze 1 , Megha Mehrotra 1 , Patricia Defechereux 1 , Susan P. Buchbinder 1 , Robert M. Grant 1 , Monica Gandhi 1 , for the iPrEx OLE Study Team 1 University of California San Francisco, San Francisco, CA, USA, 2 Abbott Labs, Abbott Park, IL, USA Background: Pharmacologic adherence measures can be used to interpret and support PrEP adherence. Current methods to analyze tenofovir (TFV) levels use liquid chromatography-tandemmass spectrometry (LC-MS/MS), which limits clinical use given cost and run time. We developed a novel antibody-based assay (in development for point-of-care (POC) testing) to quantify TFV in urine. We then tested the association of urine TFV with HIV acquisition in iPrEx OLE, a PrEP demonstration trial that enrolled men and transwomen. Methods: Spearman’s correlations between urine TFV levels via the immunoassay and hair TFV/emtricitabine(FTC) and dried blood spot (DBS) TFV-DP/FTC-TP levels via LC-MS/MS in iPrEx OLE were calculated. We calculated the sensitivity/specificity of an undetectable urine TFV for very low DBS TFV-DP levels (estimated <2 doses/week). We then compared levels of urine TFV by the immunoassay at visits where seroconversion was diagnosed, prior to HIV seroconversion, and in those who remained HIV-negative using Kruskal-Wallis’ test. We evaluated the association of an undetectable urine TFV with HIV seroconversion using generalized estimating equations. Results: Among 125 participants, the median age was 33, 14%were Black, 44% Latino. Urine TFV levels correlated with hair TFV (Ρ 0.4, p<0.001), hair FTC (Ρ 0.5, p<0.001), DBS TFV-DP (Ρ 0.5, p<0.001) and DBS FTC-TP levels (Ρ 0.7, p<0.001). When comparing an undetectable urine TFV to very low adherence by DBS (estimated <2 tablets/week) it was 70% sensitive, but 94% specific (100% sensitive/81% specific compared to undetectable DBS levels). The median urinary TFV level by the immunoassay was 15,000 ng/ml (IQR: 1,000-45,000) in those who remained HIV-negative; 5,500 (IQR: 1,000-23-000) in 11 individuals who eventually seroconverted (median 36 wks prior); and undetectable (<1000 ng/ml) in all 9 individuals at the time of seroconversion (p<0.001) (Figure).

CROI 2019 370