Can saliva replace plasma for the monitoring of methadone?

The aims of this study were to determine the relationship between saliva and plasma methadone concentrations and the influence of variability in saliva pH. Saliva and plasma samples were taken before the daily dose of methadone in 60 patients undergoing methadone maintenance treatment (MMT). Saliva pH was measured immediately after sampling, and concentrations of (RS)-, (R)-, and (S)-methadone in saliva and plasma were assayed by LC/MS. In addition, unbound (R)- and (S)-methadone concentrations were measured in plasma samples by ultrafiltration. Plasma binding and pH differences between plasma and saliva were then used to estimate methadone saliva/plasma ratios and to compare them with observed values. Saliva pH ranged from 5.1 to 7.6 (mean +/- SD, 6.7 +/- 0.5). Plasma and saliva concentrations correlated weakly [(RS)-, r = 0.14, P = 0.007, n = 44; (R)-, r = 0.10, P = 0.04, n = 43; (S)-, r = 0.22, P = 0.002, n = 43], and the mean saliva-to-plasma methadone concentration ratios were 1.1 (+/-1.3 SD), 1.5 (+/-1.5), and 0.8 (+/-0.8), for (RS)-, (R)-, and (S)-methadone, respectively. Corresponding values based on unbound concentrations of methadone in plasma were 21 (+/-20.6, n = 31), 21 (+/-19, n = 34), and 17 (+/-15, n = 36). The salivary concentration-to-dose ratios showed statistically significant but weak inverse correlations with saliva pH [(RS)-, r = 0.27, P < 0.001; (R)-, r = 0.25, P < 0.001; (S)-, r = 0.29, P < 0.001, respectively]. There were significant correlations between predicted and observed saliva/plasma ratios [(RS)-, r = 0.44, P < 0.001, n = 31; (R)-, r = 0.58, P < 0.001, n = 32; (S)-, r = 0.10, P = 0.04, n = 34], but the mean predicted saliva concentrations were about 5 times lower than the mean observed values. The poor correlations between salivary and plasma methadone concentrations observed in this study are partly related to the effect of variable saliva pH. However, saliva pH explained only 10%-36% of the total variation. As a conclusion, monitoring methadone concentrations in saliva may not be a useful alternative to plasma concentration measurements. Correction for saliva pH measured immediately after collection improves the relationship between saliva and plasma methadone concentration, but most of the variation remains unexplained.