Development and validation of a prediction model for primary non-response to IL-17A inhibitors in psoriasis.

Background: Interleukin-17A (IL-17A) inhibitors have revolutionized the treatment of moderate-to-severe plaque psoriasis, but a substantial proportion of patients experience primary non-response during the early treatment phase. Objectives: To identify clinical predictors of primary non-response to IL-17A inhibitors and to develop and validate a prediction model for individualized risk assessment. Methods: A total of 617 patients initiating IL-17A inhibitors in the Shanghai Psoriasis Effectiveness Evaluation Cohort (SPEECH) were randomly allocated into a training set ( Results: In the training cohort, 23.2% of patients met the definition of primary non-response. Clinical predictors included higher BMI, absence of family history of psoriasis, prior biologic exposure, and higher baseline PASI score. In addition to these clinical factors, failure to achieve PASI40 at week 4 was a strong indicator of subsequent non-response. The final prediction model exhibited strong discrimination (AUC 0.766, 95% CI, 0.690-0.842), good calibration between predicted and observed probabilities, and meaningful clinical usefulness as demonstrated by DCA. Performance remained robust in both the internal validation cohort (AUC 0.706) and the external validation cohort (AUC 0.708). An online calculator was developed to facilitate individualized risk estimation. Conclusion: This study identified key baseline and early-response predictors of primary non-response to IL-17A inhibitors and established a validated prediction model with clinical utility. The web-based tool may support precision decision-making in psoriasis management by enabling early identification of high-risk patients.