More than 100 drugs have been evaluated for salivary therapeutic drug monitoring since the 1970s. The most studied drugs are the anticonvulsants phenytoin, phenobarbital, and carbamazepine, demonstrating strong correlations between serum and saliva concentrations. No published data exist for levetiracetam (LEV) to the authors' knowledge. This study's objective is to determine the correlation between LEV serum and saliva concentrations.
Investigators identified subjects seen in neurology clinics at the University of Kentucky. Patients were eligible if they agreed to participate in this study, were taking LEV for a minimum of 4 weeks, and if a serum LEV concentration had been ordered by their physician. Patients spit a minimum of 0.25 mL into a cup to obtain saliva samples. Blood samples were obtained by phlebotomy.
Serum and saliva LEV concentrations were determined via high-performance liquid chromatography (HPLC) in two separate reference laboratories. Linear regression analysis was used to evaluate correlations. Serum and saliva samples were obtained from 40 patients (22 female, 18 male), ranging from 3 to 57 years of age. The mean ± SD serum LEV concentration for reference laboratory A was significantly lower (P < 0.0001) than reference laboratory B, 23.6 ± 13.8 μg/mL and 27.0 ± 16.9 μg/mL, respectively. The mean ± SD saliva to serum concentration fraction was also different for the two laboratories, i.e., 41.0% ± 0.15% for lab A and 36.0% ± 0.15% for lab B (P = 0.001). The correlation coefficients for the two laboratories were similar, 0.87 and 0.86 (both P < 0.0001) for labs A and B, respectively.
A significant positive correlation exists between LEV saliva and serum concentrations. The ability to monitor LEV therapy using saliva may provide benefits that include facilitating sample collection and improving the quality of life for persons with epilepsy. Patients with poor venous access, such as children and elderly patients, and persons afraid of needles may particularly benefit from this method.