Flow-dependent salivary primidone levels in epileptic children

In 36 epileptic children treated with primidone alone or in combination with additional anticonvulsants, salivary drug levels were compared in resting (I) and in flow-stimulated (II) saliva and were related to the corresponding serum levels. Primidone levels in saliva I and saliva II were highly correlated (r = 0.97) but were significantly (p less than 0.001) lower in saliva II; the mean difference was -38%. Serum primidone levels were highly correlated to salivary primidone levels both in saliva I (r = 0.92) and in salvia II (r = 0.91). A significant negative correlation could be established between the salivary flow rate and the saliva/serum ratio of primidone, especially in saliva I (r = 0.61; p less than 0.001). The mean saliva I/serum ratio was 1.115, reflecting drug accumulation in resting saliva. The reason primidone accumulates remains unclear. When salivary flow was stimulated, the mean saliva/serum ratio decreased to 0.7, indicating the development of a drug concentration slope from blood to saliva. This is explained by the limited permeation of the drug through cellular membranes due to its rather low lipid solubility. From the data it can be concluded that saliva is suitable for monitoring primidone levels provided the conditions of sample collection are standardized.