CROI 2024 Abstract eBook

Abstract eBook

Poster Abstracts

1104 Evidence of HIV Infection Prior to Rapid Antibody Test Positivity in PrEP Breakthrough Infections Vivian I Avelino-Silva 1 , Mars Stone 1 , Sonia Bakkour 1 , Clara Di Germanio 1 , Michael Schmidt 2 , Ashtyn Conway 3 , David Wright 4 , Brian Custer 1 , Steven Kleinman 5 , Jairam Lingappa 6 , Patricia Defechereux 7 , Megha Mehrotra 7 , Robert Grant 7 , Michael P. Busch 1 , Philip P. Norris 1 1 Vitalant Research Institute, San Francisco, CA, USA, 2 German Red Cross, Frankfurt, Germany, 3 Bloodworks Northwest Research Institute, Seattle, WA, USA, 4 Westat, Inc, Rockville, MD, USA, 5 University of British Columbia, Vancouver, Canada, 6 University of Washington, Seattle, WA, USA, 7 Gladstone Institute of Virology and Immunology, San Francisco, CA, USA Background: Exposure to antiretrovirals in PrEP users with breakthrough infection can suppress viral replication early after HIV acquisition, leading to delayed or inconsistent seroconversion. Limitations in HIV diagnostic accuracy in this context are higher for assays with lower sensitivity, such as point-of-care rapid antibody tests (RT). Methods: We obtained plasma samples from visits prior to the first positive RT from 251 participants with incident HIV during randomized PrEP trials (iPrEx and Partners PrEP), which used 3rd generation RT (state-of-the-art for HIV diagnosis at the time). Samples were categorized into time intervals and analyzed using higher performance tests (Abbott Alinity s HIV Ag/Ab Combo, Roche Elecsys HIV Duo, and Ortho VITROS HIV Combo for antigen/antibodies detection; Roche cobas MPX for RNA detection) to address positivity comparing participants assigned to active PrEP (N=101) to those assigned to placebo (N=150). Results: A large percentage of samples had positive results in higher performance tests before infection detection by RT. Rates were consistently higher for participants assigned to active PrEP compared to placebo, with statistical significance for the <8 weeks timepoint in 2 of 4 assays. The combination of RNA and antigen/antibody tests further increased overall positivity, although not reaching statistical significance when compared to antigen/antibodies tests alone. We also analyzed signal-to-cutoff/cutoff index results in antigen/antibodies tests over time prior to the first positive RT, by treatment assignment; levels were somewhat higher, and positivity was identified earlier in participants assigned to active PrEP, suggesting longer duration of infection prior to the first positive RT. Conclusion: Compared to point-of-care RT, higher sensitivity lab-based serologic and nucleic acid detection assays can enhance the identification of breakthrough HIV infections in PrEP users. Our findings suggest consideration of adopting higher sensitivity tests in PrEP programs that rely exclusively on point of-care 3rd generation RT for follow-up. For blood banks, our results reinforce that screening with parallel lab-based serologic and RNA tests should be ensured. Study limitations include sample storage beyond the manufacturers' labeled claims and possible laboratory errors. The figure, table, or graphic for this abstract has been removed. 1105 Costs and Clinic Flow of Point-of-Care Urine Tenofovir and HIV Viral Load Testing in South Africa Melody Wang 1 , Pravi Moodley 2 , Mlungisi Khanyile 3 , Elliot Bulo 4 , Makhosazane Zondi 5 , Keshani Naidoo 3 , Yukteshwar Sookrajh 4 , Jienchi Dorward 6 , Nigel Garrett 3 , Paul K. Drain 1 , Monisha Sharma 1 1 University of Washington, Seattle, WA, USA, 2 National Health Laboratory Service (NHLS), Durban, South Africa, 3 Centre for the AIDS Program of Research in South Africa (CAPRISA), Durban, South Africa, 4 Ethekwini Municipality Health Unit, Durban, South Africa, 5 University of KwaZulu-Natal, Durban, South Africa, 6 University of Oxford, Oxford, United Kingdom Background: Point-of-care (POC) urine tenofovir (TFV) tests can monitor real-time ART adherence in clinics and complement POC viral load (VL) tests for HIV treatment management. However, clinic flow and implementation cost of both tests is uncertain. We sought to estimate costs of integrated POC TFV and VL testing for persons initiating ART in South Africa to inform implementation of new diagnostic technologies for person-centered HIV care. Methods: We conducted microcosting within STREAM HIV, a randomized implementation trial evaluating POC HIV TFV and VL testing in a government clinic in KwaZulu-Natal, South Africa. We collected time-and-motion data to assess staff and client time needed for POC TFV testing. We estimated financial and economic costs for capital, clinic consumables, and personnel using a provider perspective. We updated costs for POC VL testing from 2018 to 2022 USD using the gross domestic product price deflator and new prices for test cartridges. We estimated instrument costs assuming five-year lifespans with a 3% annual discount rate. Results: The per-client test costs, including both financial and economic costs, of POC TFV and VL testing were USD $12 and $21, respectively, assuming a clinic

volume of individuals 50 initiating ART clients per month. Key cost drivers for POC TFV and VL tests were the test strip and cartridge consumables, accounting for 52% and 70% of total test costs, respectively. The median time clients spent in the clinic for a visit with a POC TFV test was 49:31 (minutes: seconds). The POC TFV testing took a median time of 9:36 (19% of total clinic visit) including sample collection, sample loading, TFV test processing, and counselling provision based on test results. Overall, 29% of the clinic visit time included direct clinical care and assessment with a provider, with clients spending a median time of 14:15 getting vitals checked, adherence monitored via POC TFV testing, and collecting ART. Waiting in line for ART took most (48%) of the clinic visit time with a median wait time of 23:42. Conclusion: POC TFV testing is low-cost, requires less than 10 minutes, and may be feasible to implement in South African clinics. Findings can inform the policy decisions and budgetary planning for ART monitoring, in South Africa and similar settings that are considering using POC TFV and VL testing to support adherence monitoring in lieu of lab-based VL testing. 1106 A Capillary Blood-Based Self-Collection Method for Monitoring HIV Viral Load During ART Interruption Livio Azzoni 1 , Daniel S. Rosenbloom 2 , Jessicamarie Morris 1 , Jennifer Nguyen 3 , Brian N. Ross 1 , Matthew Fair 1 , Emmanouil Papasavvas 1 , Kenneth Lynn 4 , Emily Hiserodt 5 , Brian Squadroni 6 , Brad R. Evans 7 , Pablo Tebas 4 , Karam Mounzer 5 , Bonnie Howell 2 , Luis J. Montaner 1 1 Wistar Institute, Philadelphia, PA, USA, 2 Merck Research Laboratories, Rahway, NJ, USA, 3 Merck & Co, Inc, Palo Alto, CA, USA, 4 University of Pennsylvania, Philadelphia, PA, USA, 5 Philadelphia FIGHT, Philadelphia, PA, USA, 6 Merck & Co, Inc, Upper Gwynedd, PA, USA, 7 Merck & Co, Inc, Rahway, NJ, USA Background: Multiple barriers limit participation of people with HIV (PWH) in studies in which frequent viral load (VL) monitoring is required such as cure studies involving a treatment interruption (ATI). Convenient, home-based VL testing may increase equitable participation. Here we report a novel VL test based on capillary blood collection, comparing its specificity and sensitivity to conventional clinic-based plasma-based VL (pVL). Methods: We enrolled 21 PWH (5% female, 67% African American, mean age of 51 years) undergoing planned ATIs as part of the BEAT-HIV trial (NCT03588715). At contiguous visits (mean 20.5), we collected A) capillary blood using two 4-well Tasso-M50 devices, and B) matched plasma samples. Between visits, participants self-collected capillary blood using two devices that were mailed back (home collection). We employed a robotic automation process for magnetic bead RNA extraction and RT-qPCR readout with dual LTR/GAG FAM-labeled primers that amplify on the same fluorescence channel. Samples and standards underwent qPCR in duplicate. Results: The devices were well-accepted, with a collection failure rate <10%. We analyzed a total of 5392 M50 PCR reads (3058 clinic, 2334 home collection). Among the 14 participants where M50 background levels could be determined from ≥ 4 weeks of suppressive treatment, M50 VL 2x above median background was predictive of matched pVL ≥ 200 c/mL (Figure 1. Sensitivity 66%, specificity 81%, PPV 71%, NPV 76%, ROC curve AUC 85%). When M50 VL was < median background, pVL was reliably low (NPV 84% for pVL ≥ 200 c/mL, NPV 90% for pVL ≥ 1000 c/mL). 13 of 14 (93%) participants experienced an increase in M50 VL at the collection immediately following pVL rebound ≥ 200 c/mL (median increase 9-fold, range 1.8- to >1000-fold); the remaining participant experienced an increased M50 VL at the subsequent collection (to 1.6-fold above background). An M50 VL ≥ 2x background was always followed by pVL rebound ≥ 200 c/mL within 4 weeks. Conclusion: Our new M50 assay showed good NPV for pVL > 200 c/ml in a cohort of PWH undergoing ATI. This suggests that, upon clinical validation, the assay could be used for home monitoring of VL during ATIs. This approach could enhance equitable participation in HIV research by minimizing participants' visit burden. Other potential applications, such as monitoring new infections in prophylaxis studies or assessing changes in residual cell-associated HIV in cure studies, are warranted.

Poster Abstracts

CROI 2024 358