Pharmacogenetic CYP2D6 variability, phenoconversion and treatment outcomes: A Danish population-based cohort study of 6,798 individuals initiating atomoxetine treatment

Background: Atomoxetine is used in individuals with Attention Deficit Hyperactivity Disorder. Atomoxetine blood concentration is affected by genetic variation of the Cytochrome P450 (CYP) drug metabolizing enzyme CYP2D6 and phenoconversion by drug-drug-CYP2D6 interactions. The effect of CYP2D6 status on clinical outcomes of atomoxetine use in the population has rarely been reported. PGx testing in clinical practice in Denmark is low.

Objectives: We investigated wide-ranging clinical outcomes in atomoxetine users in association with CYP2D6 variability according to PGx and phenoconversion.

Methods: Register-based cohort study of individuals born in Denmark between 1981 and 2006 with psychiatric disorders diagnosed at psychiatric hospitals. We included individuals who had redeemed a prescription for atomoxetine (new user) between 1995 and 2016. We linked their newborn blood spot array-based single nucleotide polymorphism (SNP) data with Danish medical and socio-demographic registers. The CYP2D6 genotypes were translated into extreme metabolizer (EM), intermediate metabolizer (IM), and poor metabolizer (PM). Use of CYP2D6 inhibitors was considered in a time varying manner (phenoconversion) during follow-up (FU) of one year. FU started at the first atomoxetine prescription and ended at the date of outcome (switching, discontinuation, inpatient-, outpatient-, emergency-contact, suicide attempt/self-harm, sleep problems, depression), death, emigration, end of FU, whichever came first. Incidence rate ratios (IRR) were estimated using Poisson regression analyses comparing IM/PM versus EM (reference) and adjusted for potential confounders and covariates.

Results: The cohort included 6798 children ( < 12 years) and adolescents (55.4%, females 26.5%) and adults (44.6% adults, females 41.0%) who initiated atomoxetine. CYP2D6 status was determined based on 4/8 known actionable SNPs; deletions and duplications could not be determined. The distribution of CYP2D6 phenotypes was similar across age groups (~31% IM, ~4.2% PM). Intermediate and strong CYP2D6 inhibitors were used by 6.6% of children and adolescents and by 14.6% of adults during FU. Children and adolescents with a genetic IM/PM status had a 1.25 (1.01-1.54) increased risk of sleep problems. Considering phenoconversion by intermediate and strong CYP2D6 inhibitors, IM/PM children had a significantly increased risk of sleep problems compared with EM children (IRR 1.86, 95% CI 1.31-2.63). Analyses for other clinical outcomes did not yield any significant differences between CYP2D6 phenotypes.

Conclusions: CYP2D6 PM/IM phenotype status was associated with increased risk of sleep problems particularly in children, further increased by phenoconversion by drugs inhibiting CYP2D6.