Association between pharmacological tenofovir adherence measures and subsequent 24-week viral load outcomes for people living with HIV in South Africa

BackgroundObjective measures of antiretroviral therapy (ART) adherence, such as tenofovir (TFV) concentrations, may allow for targeted interventions to prevent future HIV viraemia. We evaluated three TFV adherence measures and their association with current and 24-week viral load (VL) outcomes. MethodsIn a sub-study of the South African-based POwER trial, we measured urine TFV using a point-of-care (POC) antibody-based assay and two metrics assessed via liquid-chromatography-mass-spectrometry: quantitative urine TFV, and TFV-diphosphate (TFV-DP) concentrations in dried blood spots (DBS). We assessed the association between current and subsequent 24-week VL outcomes and these three measures using logistic regression models. We also compared the baseline characteristics of individuals with detectable vs. undetectable POC urine TFV results at enrolment. ResultsOf 124 participants, 54.8% female, median age 39 years, 101 (81.5%) had detectable POC urine TFV and 23 (18.5%) did not. Higher TFV-DP concentrations in DBS were negatively associated with viraemia after 24-weeks (OR 0.83, 95% CI 0.725-0.928, p=0.003), whilst a detectable POC urine TFV (OR 0.62, 95% CI 0.22-1.82, p=0.380) and quantitative urine TFV (OR 0.98, 95% CI 0.95-1.00, p=0.153) showed no significant association. Compared to those with detectable POC urine TFV, those with undetectable POC urine TFV were more likely to be concurrently viraemic at enrolment (78.3% vs 25.7%, p<0.001) and have a CD4 count <200 cells/uL (34.8% vs 12.9%, p=0.001). ConclusionDBS TFV-DP was associated with viraemia at 24-weeks and could help predict future viraemia. Undetectable POC urine TFV was associated with recent ART initiation, current viraemia, and lower CD4 count. Trial registrationPan African Clinical Trials Registry (PACTR202001785886049) and the South African Clinical Trials Registry (DOH-27-072020-6890).