Relationship Between Amphetamine Concentrations in Saliva and Serum in Children and Adolescents With Attention-Deficit/Hyperactivity Disorder.

Background: Therapeutic drug monitoring (TDM) is a valid tool for the optimization of psychopharmacotherapy; however, in child and adolescent psychiatry, uncomfortable intravenous sample collection is the main challenge and restricts the use of TDM. Therefore, it is important to evaluate alternate specimens to facilitate TDM. The aim of this study was to evaluate the feasibility of using saliva for the TDM of amphetamine in children and adolescents with attention-deficit/hyperactivity disorder.
Methods: In this study, 28 patient samples (mean age, 11.3 years; boys, 23; and girls, 5) treated with lisdexamfetamine were included. The active compound amphetamine was extracted and derivatized before quantification by high-performance liquid chromatography with fluorescence detection. Nonparametric Spearman rank correlations were used for correlation analyses; for clinical validation, Bland-Altman analysis was applied.
Results: The median amphetamine concentrations in saliva were 2.7 times higher (range 0.7-23.6) than those in serum (257.8 ng/mL versus 77.2 ng/mL; z = -4.51, P < 0.001). A strong positive linear correlation was observed between saliva and serum concentrations (ρ = 0.628, P < 0.001). The ratio of saliva-to-serum concentration was strongly pH dependent (ρ = -0.712, P < 0.001). Therefore, a transformation formula, factoring in salivary pH, to calculate serum concentrations from the measured saliva concentrations was applied. Theoretical and measured serum amphetamine concentrations were subjected to Bland-Altman analysis. Using an acceptance limit of 20%, only 21% (n = 6) of samples fulfilled this criterion.
Conclusions: Amphetamine paired saliva-to-serum concentrations were highly variable and strongly affected by salivary pH, indicating that saliva is an inappropriate sampling matrix for TDM of amphetamine. Furthermore, Bland-Altman analysis did not support saliva as a suitable matrix for TDM.
Competing Interests: The authors declare no conflict of interest.
(Copyright © 2020 Wolters Kluwer Health, Inc. All rights reserved.)