(P-539) The Association Between Trajectories of Use of Systemic Agents and Depression in Patients with Psoriasis: A Sequence and Cluster Analysis

Background: Moderate-to-severe psoriasis is managed with conventional systemic agents (CSA) and/or biologic agents when CSA are ineffective or inappropriate. Failure of these treatments may trigger depression and depression in turn may exacerbate psoriasis.

Objectives: To determine the longitudinal treatment patterns of systemic agents’ use (or trajectories) and assess their association with depression in patients with moderate to severe psoriasis.

Methods: We conducted a cross-sectional study using Quebec health administrative databases (1997 - 2015). Individuals ages ≥ 20 years were included if they received a first psoriasis diagnosis and initiated at least one CSA in 2002-2013 and did not have depression. Treatment patterns were assessed by 1-month intervals over the following two years. Psoriasis treatments were classified in each interval according to five treatment categories: 1) combination of biologics and CSA, 2) ≥2 CSAs, 3) biologic (no CSA), 4) CSA (no biologics) and 5) other treatments (topical agents, phototherapy or no treatment). We measured the dissimilarity between individual treatment trajectories using the dynamic hamming distance, a sequence analysis measure. Patients with similar trajectories were grouped into clusters using agglomerative hierarchical clustering with Ward’s linkage criterion. Depression (yes/no) was also assessed during the 2-year study period. Multivariable logistic regression was conducted to measure the association between the different trajectory clusters and depression.

Results: We included 953 patients (51.8% females, mean age 60±15.5 years). Five trajectory clusters were identified: persistent users (40.5%), early discontinuation (35.9%), intermediate-to-late discontinuation (12.4%), switch/augmentation to a biologic agent or a different CSA (5.2%), and discontinuation and restart of therapy (6.0%). In the two-year study period, 30 patients had depression. When compared to the trajectory persistent users, switch/augmentation to a biologic agent or a different CSA was associated with a non-significant increase in the likelihood of depression (OR = 2.24, 95% CI: 0.60-6.75).

Conclusions: Our study identified five longitudinal treatment patterns in moderate to severe psoriasis using a novel methodological approach. Our study lacked statistical power for a conclusive determination of the association between these patterns and depression.