CROI 2020 Abstract eBook

Abstract eBook

Poster Abstracts

Conclusion: We report DRM prevalence among all PWH with RT or IN sequences in NHSS and without evidence of recent viral suppression to describe possible sources of HIV exposure for people on PrEP. We document an established reservoir of M184V and low prevalence of K65R and DRMs predicted to affect CAB. While M184V reduces susceptibility to FTC, its impact on PrEP efficacy remains unclear as it also reduces viral fitness and increases susceptibility to TDF. Differences in DRM prevalence by sex, race/ethnicity, and transmission risk suggest differential risk for exposure to HIV strains that might impact PrEP. Continued DRMmonitoring is essential for identifying potential threats to PrEP effectiveness in an era of expanding use.

513 COUNTRY-LEVEL DRIVERS OF NNRTI RESISTANCE IN SOUTHERN AFRICA Julien Riou 1 , Carole Dupont 1 , Silvia Bertagnolio 2 , Ravindra K. Gupta 3 , Leigh F. Johnson 4 , Roger Kouyos 5 , Matthias Egger 1 , Christian L. Althaus 1 1 University of Bern, Bern, Switzerland, 2 WHO, Geneva, Switzerland, 3 University College London, London, UK, 4 Centre for Infectious Disease Epidemiology and Research, Cape Town, South Africa, 5 University of Zurich, Zurich, Switzerland Background: The rise in the prevalence of drug resistance to non-nucleoside reverse-transcriptase inhibitors (NNRTIs) in HIV-infected individuals initiating antiretroviral therapy (ART) is a major problem in countries of southern Africa. Understanding the dynamics and drivers of NNRTI resistance at the country level is of critical importance for planning future ART programs. Methods: We first collected survey data on pretreatment drug resistance (PDR) to NNRTIs in nine countries of southern Africa from 2000 to 2018, including 66 studies and 14,639 individuals. We then fitted a mechanistic transmission model to key indicators of the local HIV epidemics (HIV prevalence, ART coverage and AIDS mortality) and the levels of PDR using a hierarchical Bayesian framework. For each country, we estimated the rate at which treatment failure with NNRTI resistance (TFNR) occurs during ART. We further explored the association between TFNR and socio-economic covariates. Results: The model reliably described the local dynamics of HIV and the rise of NNRTI PDR, with the exception of Malawi and Zambia where data quality was insufficient. Predicted levels of NNRTI PDR in 2018 ranged between 4.7% (95% credible interval: 2.2, 9.8) in Mozambique and 32.8% (26.4, 38.7) in Namibia. The main driver of NNRTI resistance was the conjunction of ART coverage and the rate of TFNR. Estimates of the rate of TFNR were lowest in Botswana (0.002 per year; 0, 0.006) and highest in the Republic of South Africa (0.14 per year; 0.11, 0.17). The regional average of this rate was 0.07 per year (0.04, 0.25) corresponding to a probability of 8% (4, 22) that patients initiating ART show treatment failure due to the acquisition of NNRTI resistance after one year. TFNR was associated with external health expenditure (Pearson correlation: -0.43; -0.59, -0.19) and out-of-pocket health expenditure (0.39; 0.01, 0.75). Conclusion: Even with the introduction of dolutegravir, NNRTIs will remain a central component of first-line regimen in southern Africa. Between-country comparison shows that NNRTI PDR can be controlled despite high levels of ART coverage, as in Botswana, likely because of better patient management and lower exposure to ART before treatment initiation. Our results suggest that the ability to control PDR is associated with features of the healthcare financing system at the national level. Additional data on NNRT PDR and ART management is urgently needed in some countries of southern Africa.

Poster Abstracts

514 ADVANCE TRIAL: HIGHER RISK OF TREATMENT-EMERGENT RESISTANCE ON FIRST-LINE TDF/FTC/EFV Willem D. Venter 1 , Michelle A. Moorhouse 1 , Simiso Sokhela 1 , Bryony Simmons 2 , Andrew Hill 3 1 Wits Reproductive Health and HIV Institute, Johannesburg, South Africa, 2 Imperial College London, London, UK, 3 University of Liverpool, Liverpool, UK Background: In low- and middle-income countries, most treatment-naïve people living with HIV (PLWH) take first-line treatment with no baseline resistance testing. In the SINGLE trial, there was a significantly higher risk of treatment-emergent drug resistance in the TDF/FTC/EFV arm (1.7%) compared with the ABC/3TC/DTG arm (0.0%). In South Africa, over 10% of treatment-naïve patients have transmitted NNRTI drug resistance. Methods: We conducted a 96-week, open-label randomised trial in South Africa, comparing TAF/FTC/DTG, TDF/FTC/DTG and TDF/FTC/EFV. Inclusion criteria included age ≥12 years, no prior ART >30 days, and HIV-1 RNA >500 copies/mL. There was no screening for baseline drug resistance, consistent with South African treatment guidelines. Patients with at least one HIV RNA result above 1000 copies/mL after 24 weeks of randomised treatment were genotyped prospectively, together with a test of their stored baseline sample. For this analysis, virological failure was classified as HIV RNA >1000 copies/mL at Week 12 or 24, >200 copies/mL at Week 24 or >50 copies/mL at Week 48 or later, while taking randomised treatment. The number of genotyped patients with treatment-emergent WHO major NRTI or NNRTI mutations was compared between the arms (Fishers exact test). Results: We randomised 1053 PLWH between February 2017 and May 2018: 99% black, 59% female, mean age 32 years, with mean CD4 336 cells/uL. At week 48, the percentage of participants with HIV RNA <50 copies/mL was 83.8% for TAF/FTC/DTG, 84.9% for TDF/FTC/DTG and 78.6% for TDF/FTC/EFV. As shown in Table 1, NRTI or NNRTI emergent resistance at virological failure was more common for first-line treatment with TDF/FTC/EFV (10/14; 71%) compared with the combined TAF/FTC/DTG and TDF/FTC/DTG arms (2/23; 9%), p<0.01. The most common treatment-emergent NRTI mutations were M184V (n=5), K65R (n=2); most common NNRTI mutations were K103N (n=3), and P225H (n=2). Of the 10 patients developing NRTI or NNRTI RAMS at VF in the TDF/FTC/EFV arm, 8 (80%) already had at least one NRTI or NNRTI RAM at baseline. No treatment- emergent integrase mutations were observed. Conclusion: In the ADVANCE study, there were similar rates of virological failure between the arms. However the patients in the TDF/FTC/EFV armwere significantly more likely to develop NRTI or NNRTI mutations at VF (71%) compared to the DTG arms (9%). Most patients with treatment-emergent resistance already had NRTI or NNRTI mutations at baseline.

CROI 2020 183