Abstract Body


We previously developed a point-of-care (POC) assay to measure urine tenofovir (TFV) levels among patients on tenofovir disoproxil fumarate (TDF), with a cut-off of 1500 ng/ml indicating adequate adherence. Since tenofovir alafenamide (TAF) results in lower plasma concentrations than TDF, the clinical utility of the existing assay for patients on TAF is unknown. We thus explored whether urine TFV levels above the 1500 ng/ml cut-off correlate with future virologic suppression (VS) (< 200 copies/mL) among people living with HIV (PLWH) on TAF.


We collected urine samples from patients with HIV on TAF-based regimens at two San Francisco HIV primary care clinics from 06/2019-12/2021. We used our electronic medical record to describe sociodemographic/clinical characteristics, then measured continuous urine TFV levels via liquid chromatography/tandem mass spectrometry. Finally, we used generalized estimating equations to fit a model comparing urine TFV levels (above and below 1500 ng/ml) for samples with and without VS on the first viral load measured after urine collection.


Our analysis included 83 samples (68 suppressed and 15 unsuppressed) from 67 unique PLWH. Samples from person-visits with/without VS were similar in age (median 55 vs 45 years), sex (88% vs 100% male), gender (75% vs 67% cis-male), race/ethnicity (46% vs 33% Black; 37% vs 33% White), glomerular filtration rate (GFR, 84% vs 87% >60 ml/min), creatinine (1.02 vs 0.85 mg/dL), and weight (81.6 vs 73.7 kg); all with p >0.05. The median (interquartile range) urine TFV levels by LC-MS/MS were respectively 3190 (1460, 5720) ng/ml in samples from visits with VS and 690 (70, 2100) ng/ml from visits without suppression. In adjusted modeling accounting for age, race/ethnicity, and GFR (Table 1), urine TFV levels >1500 ng/ml were strongly associated with future VS (OR 5.66; 95% CI 1.59-20.14; p=0.007). The sensitivity/specificity/PPV/NPV at the 1500 ng/ml cut-off were respectively 75%, 67%, 91%, and 63%.


Urine TFV levels above 1500 ng/ml were strongly predictive of future VS among patients on TAF, suggesting that individuals taking TAF who flag below 1500 ng/ml on POC testing would benefit from enhanced adherence counseling. Our results further imply: (1) that the existing POC assay originally developed for TDF may have real-world applicability to predict VS among people with HIV on TAF, and (2) that a single POC assay may support adherence monitoring for patients on both TAF and TDF worldwide.

Table 1: Adjusted Generalized Estimating Equations Model for Future Virologic Suppression by Urine Tenofovir (TFV) Levels (above 1500 ng/ml) among People with HIV on TAF-based Antiretroviral Therapy