Evaluation of saliva as an alternative matrix for monitoring plasma Zidovudine, Lamivudine and Nevirapine concentrations in Rwanda

Saliva may provide interesting advantages as matrix for compliance measurements, pharmacokinetic studies and therapeutic drug monitoring in resource limited countries. We investigated the feasibility of using saliva for compliance monitoring of zidovudine (ZDV), lamivudine (3TC) and nevirapine (NVP) in 29 HIV-1 infected patients from Rwanda. ZDV, 3TC and NVP drug levels were quantified by an LC/MS-MS method in plasma and stimulated saliva samples and compared using Bland-Altman analysis. Seven patients demonstrated undetectable saliva ZDV levels while five out of these seven also showed no 3TC salivary concentrations. For the other samples, we observed a good agreement between salivary and plasma concentrations of each antiretroviral drug. A significant relation between the difference in saliva and plasma ZDV concentrations and the average ZDV concentration in the two matrices was deduced as follows: y = -380.15 + 1.79 x. The log saliva and plasma concentration difference of both 3TC and NVP was consistent across the range of average log concentration. Overall, we showed large agreement limits suggesting a wide inter patient variability that may result in non-reliable plasma level predictions from saliva drug measurements. Therefore, our results indicate that saliva may serve as a valuable tool only for NVP compliance testing because of its high salivary concentration.