Longitudinal monitoring of cabotegravir and rilpivirine exposure from hair strand concentrations

21 Background: Virological failure among individuals using long-acting injectable (LAI) treatment for HIV has been associated with low trough concentrations of cabotegravir (CAB) + rilpivirine (RPV) despite on-time injection. A non-invasive method for monitoring months of drug exposure from LAI treatment may help identify persons at-risk. We have developed an approach for the long-term characterization of daily CAB and RPV exposure from LAI treatment in hair strands using mass spectrometry imaging (MSI). Methods: Eighteen participants receiving LAI CAB+RPV for HIV treatment (6 female, 11 male, 1 non-binary) provided five hair strands plucked from the occipital scalp and plasma. CAB and RPV were measured along individual hair strands by infrared matrix assisted laser desorption electrospray ionization (IR-MALDESI) MSI. Profile variability was assessed by relative standard deviation (RSD). Agreement between CAB/RPV in hair and LC–MS/MS plasma concentrations was evaluated by Pearson correlation coefficient (r), also used for comparison to participant BMI. All values were reported as median (range). Results: Participants were sampled during their third (second to seventh) LAI dosing interval, typically (16/18) when returning to clinic for an injection. Sampling occurred 58 (30–68) days after prior injection, reflecting a mix of individuals receiving Q4W (6/18) and Q8W (12/18) treatment. Longitudinal profiles from MSI sample analysis evaluated 86 (46–101) days of growth, capturing at least the most recent and up to two prior LAI dosing intervals. CAB and RPV were detectable in all samples, and patterns of LAI treatment (from initiation to maintenance dosing) could be differentiated in a single sample at Q4W or Q8W frequency (Figure 1). LAI hair profile variability [CAB RSD: 52% (34–84); RPV RSD: 42% (10–89)] was greater than variability of QC standards [RSD: 32% (22–66)] and adherent oral FTC dosing (ENLIGHTEN study, NCT04232540) [RSD: 29% (13–72)]. Positive correlation was observed between hair and plasma CAB/RPV (rCAB = 0.31; rRPV = 0.30). These measurements were each negatively correlated with BMI (CAB: rHair = −0.28, rPlasma = −0.33; RPV: rHair = −0.31, rPlasma = −0.44). Conclusions: Identifying patterns and variability of CAB and RPV exposure has important implications for optimizing care of persons using LAI for treatment. The increased variability for CAB noted in hair uptake may reflect increased variability in plasma concentrations and should be further investigated as a non-invasive measure of systemic pharmacokinetics.