Patterns of long-term and short-term responses in adult patients with attention-deficit/hyperactivity disorder in a completer cohort of 12 weeks or more with atomoxetine

Abstract Background Atomoxetine is a well-established pharmacotherapy for adult ADHD. Long-term studies show incremental reductions in symptoms over time. However, clinical experience suggests that patients differ in their response patterns. Methods From 13 Eli Lilly-sponsored studies, we pooled and analyzed data for adults with ADHD who completed atomoxetine treatment at long-term (24 weeks; n = 1443) and/or short-term (12 weeks; n = 2830) time-points, and had CAARS-Inv:SV total and CGI-S data up to or after these time-points and at Week 0 (i.e. at baseline, when patients first received atomoxetine). The goal was to identify and describe distinct trajectories of response to atomoxetine using hierarchical clustering methods and linear mixed modelling. Results Based on the homogeneity of changes in CAARS-Inv:SV total scores, 5 response clusters were identified for patients who completed long-term (24 weeks) treatment with atomoxetine, and 4 clusters were identified for patients who completed short-term (12 weeks) treatment. Four of the 5 long-term clusters (comprising 95% of completer patients) showed positive trajectories: 2 faster responding clusters (L1 and L2), and 2 more gradually responding clusters (L3 and L4). Responses (i.e. ≥ 30% reduction in CAARS-Inv:SV total score, and CGI-S score ≤ 3) were observed at 8 and 24 weeks in 80% and 95% of completers in Cluster L1, versus 5% and 48% in Cluster L4. Conclusions While many adults with ADHD responded relatively rapidly to atomoxetine, others responded more gradually without a clear plateau at 24 weeks. Longer-term treatment may be associated with greater numbers of responders.