Saliva versus Plasma for Pharmacokinetic and Pharmacodynamic Studies of Fentanyl in Patients with Cancer

Abstract Purpose Fentanyl is widely used to relieve cancer pain. However there is great interpatient variation in the dose required to relieve pain and little knowledge about the pharmacokinetic and pharmacodynamic (PK/PD) relationship of fentanyl and pain control. Patients with cancer are fragile and there is reluctance on the part of health professionals to take multiple plasma samples for PK/PD studies. The relationship between plasma and saliva fentanyl concentrations was investigated to determine whether saliva could be a valid substitute for plasma in PK/PD studies. Methods One hundred sixty-three paired plasma and saliva samples were collected from 56 patients prescribed transdermal fentanyl (Durogesic, Janssen-Cilag Pty Limited, NSW, Australia) at varying doses (12–200 µg/h). Pain scores were recorded at the time of sampling. Fentanyl and norfentanyl concentrations in plasma and saliva were quantified using HPLC-MS/MS. Findings Saliva concentrations of fentanyl (mean = 4.84 μg/L) were much higher than paired plasma concentrations of fentanyl (mean = 0.877 μg/L). Both plasma and saliva mean concentrations of fentanyl were well correlated with dose with considerable interpatient variation at each dose. The relationship between fentanyl and norfentanyl concentrations was poor in both plasma and saliva. No correlation was observed between fentanyl concentration in plasma and saliva ( r2 = 0.3743) or free fentanyl in plasma and total saliva concentrations ( r2 = 0.1374). Pain scores and fentanyl concentration in either of the matrices were also not correlated. Implications No predictive correlation was observed between plasma and saliva fentanyl concentration. However the detection of higher fentanyl concentrations in saliva than plasma, with a good correlation to dose, may allow saliva to be used as an alternative to plasma in PK/PD studies of fentanyl in patients with cancer.