Why Atomoxetine Fails — and How to Fix It

Your patient’s genetics may explain why atomoxetine isn’t working

STUDY: Tobin KV et al, J Clin Pharmacol 2026;66(4):e70168

STUDY TYPE: Population pharmacokinetic modeling study

FUNDING: Eunice Kennedy Shriver National Institute of Child Health and Human Development

Background

Atomoxetine is a norepinephrine reuptake inhibitor approved for ADHD, and is one of the few medications where the FDA has warned about clinically significant effects from altered drug metabolism. Two liver enzymes, CYP2D6 and CYP2C19, metabolize atomoxetine. Whether through drug interactions or genetic differences, poor metabolizers can end up with over ten times the drug exposure of normal metabolizers at the same dose, causing extreme fatigue and other side effects.

The Study

Researchers combined pharmacokinetic data from 86 children and adolescents (ages 6–17) across three studies, each with genotyping for CYP2D6 and CYP2C19. Participants fell into four metabolizer categories: poor (n=6), intermediate (n=41), normal (n=35), and ultra-rapid (n=4). The team built a population pharmacokinetic model incorporating body weight, CYP2D6, and CYP2C19 phenotype to predict drug exposure at standard and individualized doses.

The model confirmed dramatic exposure differences by phenotype. Poor metabolizers had 130% higher peak concentrations and 624% higher overall drug exposure compared to normal metabolizers at the same weight-based dose — and their drug took over two hours longer to reach peak levels. Intermediate metabolizers had about 1.3-fold higher bioavailability than normal metabolizers, and CYP2C19 poor metabolizers showed roughly 2.3-fold higher bioavailability, a clinically meaningful difference even though only two participants fell into that category.

Limitations: Sample was small — especially for poor and ultra-rapid metabolizers — which limits how confidently these results can be generalized.

Practice Implications

1. If you know the patient’s CYP genotype, follow the titration guidelines for atomoxetine at the Clinical Pharmacogenetics Implementation Consortium (scroll to Table 1). They recommend checking peak (not trough) serum levels and aiming for close to 400 ng/ml.
2. For drug interactions, treat “strong” or “potent” CYP inhibitors as if they are poor metabolizers (strong inhibitors are in bold in table below). Treat “moderate” inhibitors as if they are intermediate metabolizers (using dosing from Table 1).

Common drugs that inhibit or induce CYP-2D6 and -2C19 enzymes. Risks of inhibition are listed as “Inh:”. Strong inhibitors and inducers are in bold. From Prescribing Psychotropics, C Aiken et al, 2020.