(P-421) Inadequate Response Rates and Associated Factors to Psoriasis Therapies Using a Claims-based Algorithm

Background: Limited research has been done to identify patient characteristics and factors associated with variability in responses to psoriasis (PsO) therapies using real-world, large-scale data sources such as health plan claims.

Objectives: To estimate frequency of inadequate response (IR) to PsO therapies (biologic and phosphodiesterase 4inhibitor [PDE4i]) using a claims-based algorithm and to describe characteristics and treatment patterns among patients with IR.

Methods: This retrospective observational cohort study included PsO patients (aged 18+), identified using US administrative claims data from HealthCore Integrated Research Database®, initiating biologics or PDE4i (Apremilast) therapies between Jul 1, 2016 to Aug 31, 2018. Baseline period was 6 months pre-index date (i.e. date of initiation and follow-up was 12 months. Patients were stratified based on treatment response and IR was defined using a modified algorithm from Curtis et al. Patient characteristics were compared between responders and inadequate responders. Inadequate responders were stratified based on treatment switching to non-index therapies during follow-up. Multivariable logistic regression models were fitted to identify key baseline patient characteristics associated with IR to initial therapy.

Results: Among 4952 eligible patients, 70% were inadequate responders (95% CI: 68–71%). Low adherence (64%) and switching or addition of biologic/PDE4 therapy (14%) were primary reasons for IR. Overall, mean (SD) age was 46.8 (13.64) years; 52% male; and Quan–Charlson Comorbidity Index score was 0.28 (0.82). The proportions of patients treated with index tumor necrosis factor inhibitor (TNFi) biologics, non-TNFi biologics, and PDE4i were 36%, 29%, and 36%, respectively. Non-TNFi biologics were associated with significantly lower IR rates (64%) vs. TNFi biologic (69%, odds ratio [OR]=0.84) and PDE4i therapy (74%, OR=0.60) (all p<0.05). Patients who were female (OR=1.21), had previously used TNFi biologics (OR=2.26), and had existing comorbidities (bacterial infection, OR=1.22; mental health issues, OR=1.35; and diabetes without chronic complications, OR=1.22) were associated with higher IR rates (all p<0.05). Among switchers, patients using TNFi agents (50%) were more than non-TNFi agents (15%) or Apremilast (35%); between-categories switching was almost 5 times more frequent than within-categories switching.

Conclusions: Non-TNFi biologics had the lowest IR rate compared with other therapy agents. Patient characteristics, treatment history, and type of index therapy affected treatment response.