Work

Emtricitabine and lamivudine concentrations in saliva: a simple suitable test for treatment adherence

Year: 2019

Type: article

Abstract: Sir, ART constitutes the cornerstone of HIV treatment and prevention. Successful ART results in suppressed viral load (VL) levels. Unsuppressed VL is either a result of acquisition of drug resistance mutations or insufficient drug concentrations due to poor adherence. Since clinical and immunological parameters are poor predictors of successful viral suppression, the WHO recommends regular VL monitoring for all individuals on ART.1 However, due to cost and infrastructure issues, the implementation of routine and frequent VL monitoring remains slow in the countries with the highest numbers of individuals on ART. Although measures such as pill count and self-reporting have been reported in several studies to be a good guide to viral suppression, they remain error prone.2 In this context, therapeutic drug monitoring (TDM) is frequently requested. Indeed, low plasma drug concentrations have been shown to predict unsuppressed VL.3 In addition, plasma drug concentration constitutes the definitive marker of short-term drug exposure. Emtricitabine and lamivudine are part of virtually all currently used fixed-dose regimens, thus undetectable emtricitabine/lamivudine levels would be a good proxy for recent poor adherence. Non-invasive saliva sampling could therefore represent an alternative or clinically useful addition to VL measurements and genotype determination among patients suspected of non-adherence. If available as a point-of-care test, it would provide the clinicians with an objective assessment of recent adherence during the clinical visit. To date, very few publications have evaluated saliva in the context of ART TDM.4–7 Nevirapine was found to reach high saliva concentrations, indicating that saliva could be suitable for TDM. In Tanzania, nevirapine saliva concentrations have shown good agreement with self-reported adherence, demonstrating the value of such a TDM approach.8 Concerning the other studied drugs (i.e. lamivudine and zidovudine), the excessive variability in saliva concentrations may preclude any reliable plasma level predictions from saliva drug measurements. Two studies reported a good penetration of emtricitabine into saliva in the context of pre-exposure prophylaxis.9,10 We aimed to evaluate whether saliva would constitute a suitable matrix for ART adherence monitoring, by examining the correlation between emtricitabine and lamivudine levels simultaneously measured in plasma and saliva. A total of 73 patients participating in the Swiss HIV Cohort Study (SHCS) #815 study agreed to donate a non-stimulated saliva sample simultaneously to the blood sample collected as part of the biannual SHCS follow-up visit. Emtricitabine and lamivudine levels were quantified by an LC/MS-MS assay, using stable isotopically labelled internal standards also adapted for saliva concentration determination.11 Overall, 47 and 26 paired plasma and saliva concentrations were collected for emtricitabine and lamivudine determination, respectively. In addition, six saliva samples were collected from three SHCS patients not receiving emtricitabine or lamivudine, and from three HIV-negative individuals, to evaluate the selectivity of the method. Emtricitabine plasma and saliva concentrations ranged from 8 to 2471 ng/mL and 23 to 1017 ng/mL, respectively. The time between sampling and last drug intake (time after dose; TAD) ranged from 0.8 to 37.3 h. For lamivudine, plasma and saliva concentration varied between 45 and 3183 ng/mL and between 15 and 718 ng/mL, respectively, with a TAD of 1.3–27 h. Our results show that emtricitabine and lamivudine saliva concentrations are moderately correlated with plasma concentrations (correlation coefficient 0.63 and 0.58 for emtricitabine and lamivudine, respectively) (Figure 1). Correlation between saliva and plasma concentrations (n = 47 and n = 26 for emtricitabine and lamivudine, respectively). Points represent paired plasma and saliva concentrations. The solid line represents the correlation between these concentrations, with its 95% CI shown in grey. Median saliva/plasma ratios were 0.61 for emtricitabine and 0.35 for lamivudine, with high variability (coefficient of variation 89% and 101% for emtricitabine and lamivudine, respectively). This variability can be explained by multiple factors involving drug properties (i.e. protein binding, saliva pH, salivary flow rates and degree of ionization) or patient characteristics (i.e. renal clearance or adherence issues). Importantly, saliva concentrations from individuals not receiving emtricitabine or lamivudine were below the limit of quantification (LOQ) in saliva (<10 ng/mL), confirming the specificity of the method. Assuming that emtricitabine in saliva follows the same kinetics as in plasma, the t½ of emtricitabine in saliva calculated for each patient using a population pharmacokinetics model of emtricitabine12 gave a mean t½ of 9.4 ± 0.88 h (minimum = 8.4 h, maximum = 12.1 h). It can be extrapolated that after 31 h (i.e. 3.3 × t½) emtricitabine levels in saliva would lie above the assay LOQ of 10 ng/mL in all patients, whereas after 60 h and 72 h (2.5 days and 3 days TAD, respectively) the saliva emtricitabine concentration would fall below the LOQ cut-off of the assay in 91% and 100% of patients, respectively. Thus, knowing the emtricitabine or lamivudine concentrations determined in saliva allows the identification of non-adherent patients who have missed their ART for at least the last 3 days. In conclusion, despite a limited correlation between plasma and saliva concentrations, saliva constitutes a suitable non-invasive surrogate to identify patients who are non-adherent to ART. These first results have been obtained using an MS assay. However, if developed as a qualitative, immunoassay-based, easy-to-use and non-invasive point-of-care test, this assay would be of invaluable clinical usefulness to reduce the number of unnecessary VL measurements and genotype determinations in cases of proven non-adherence, and may thus prevent unnecessary switches to second-line ART. Anagnostopoulos A., Battegay M., Bernasconi E., Böni J., Braun D. L., Bucher H. C., Calmy A., Cavassini M., Ciuffi A., Dollenmaier G., Egger M., Elzi L., Fehr J., Fellay J., Furrer H., Fux C. A., Günthard H. F. (President of the SHCS), Haerry D. (deputy of ‘Positive Council’), Hasse B., Hirsch H. H., Hoffmann M., Hösli I., Huber M., Kahlert C. R. (Chairman of the Mother & Child Substudy), Kaiser L., Keiser O., Klimkait T., Kouyos R. D., Kovari H., Ledergerber B., Martinetti G., Martinez de Tejada B., Marzolini C., Metzner K. J., Müller N., Nicca D., Paioni P., Pantaleo G., Perreau M., Rauch A. (Chairman of the Scientific Board), Rudin C., Scherrer A. U. (Head of Data Centre), Schmid P., Speck R., Stöckle M. (Chairman of the Clinical and Laboratory Committee), Tarr P., Trkola A., Vernazza P., Wandeler G., Weber R., Yerly S. This work is supported by the Swiss National Research Foundation (SNF) (SNF grant no. 324730–165956 to L. A. D., and SNF no. PCEFP3 181355 to N. D. L.). M. S. declares relations with Abbvie, Gilead, MSD, Janssen, Mepha, Sandoz, ViiV. A. C. has received sponsorship and travel grant from Gilead, and unrestricted educational grants from ViiV, MSD, Janssen-Cilag, and BMS. All other authors: none to declare.

Source: Journal of Antimicrobial Chemotherapy

DOI: 10.1093/jac/dkz181