Saliva as a specimen for monitoring compliance but not for predicting plasma concentrations in patients with HIV treated with indinavir

The presence of the HIV-protease inhibitor indinavir in saliva was analyzed to investigate whether salivary indinavir concentrations are applicable to monitor compliance and/or predict plasma indinavir levels. Fourteen HIV-infected outpatients treated with indinavir and 24 healthy volunteers who ingested a single dose of indinavir were included. Paired plasma and citric-acid-stimulated saliva samples were analyzed by high-performance liquid chromatography (HPLC). Stimulated salivary indinavir concentrations showed a high correlation (r = 0.85, p < 0.01) with corresponding plasma levels. The median saliva/plasma ratio was 65% (P25 50%; P75 94%). The ratios were independent of the plasma concentration; however, a relation with time after ingestion was seen. The unbound fraction of indinavir in plasma was not significantly correlated with the saliva/plasma ratio after stimulated saliva collection, in contrast with a subset of nonstimulated saliva from healthy volunteers, where we did find a significant correlation. Although stimulated salivary indinavir concentrations are highly correlated with plasma concentrations, it is not possible to predict plasma indinavir levels by the salivary concentrations for purposes of therapeutic drug monitoring, due to large interindividual and intraindividual variation. Nevertheless, monitoring compliance by measuring the presence of indinavir in saliva is possible: ingestion of indinavir can be assessed with a sensitivity of 84.8% in the whole dosing interval or with 98.8% between 1 and 6 hours after the last dose, which is comparable with plasma.