Objectives: In studies of adult patients with attention deficit hyperactivity disorder (ADHD), atomoxetine has demonstrated substantial improvements in ADHD symptomatology using the Conners’ Adult ADHD Rating Scales (CAARS). The pattern suggests that there is an incremental response over time, with no clear plateau of response. This analysis was conducted to identify patterns of response to atomoxetine in adult ADHD patients and to describe those trajectories over time, and to determine if patients have distinct response trajectories using CAARS in 2 populations during short-term (12 weeks) and long-term (24 weeks) atomoxetine treatment.
Methods: Data from 13 studies of 2502 atomoxetine-treated patients, with a CAARS-Investigator Rated: Screening Version (CAARS-Inv:SV) total score at baseline and at either a >= short-term or >= long-term time point, were included in the analyses. Data were obtained from Phase 2, 3, or 4 double-blind or open-label studies. Nine of the source studies were long term. Percentages of responders were calculated based on a>=30% improvement in CAARS-InV:SV total scores from baseline and a Clinical Global Impressions for Severity (CGI-S) of ADHD score ≤3. The numbers of trajectory clusters based on response rate for short-term (n=2502) and long-term (n=1139) data were identified using hierarchical clustering methods. Linear mixed modeling was used to describe mean last observation carried forward change to end point in CAARS-Inv:SV, CGI-S, and Adult ADHD Quality of Life (AAQoL) scores.
Results: Using CAARS total, 4 trajectory clusters were identified in short-term treated patients and 5 in long-term treated patients. Baseline characteristics appeared to be similar in each cluster with the exception of ADHD subtypes, CAARS-Inv:SV total score, and CGI-S scores, which differed numerically. Three out of 4 short-term (representing 84% of patients) and 4 out 5 long-term (representing 96% of patients) trajectory clusters showed more successful trajectories. In the long-term trajectory clusters, there was a rapid increase in the proportion of responders in 2 of the clusters, and almost all of the patients had responded after 24 weeks of treatment. A further 2 longterm trajectory clusters showed a more gradual increase, with the proportion of responders continuing to increase over the 24 weeks. In general, clusters with less improvement were those with the worst baseline CAARS-Inv:SV and minimal initial improvement. Distinct trajectory patterns of response were found that were incremental over time in all clusters. Changes in CAARS-Inv:SV total scores were consistent with changes seen in CGI-S and AAQoL total scores.
Conclusion: Adult ADHD patients receiving atomoxetine have individual trajectories of response that can be divided into 4 short-term and 5 long-term trajectories. The proportion of responders increased over time, with no plateau in response