Intestinal Ultrasound Predicts Primary Non‐Response and Maintenance Relapse in Upadacitinib‐Treated Active Crohn's Disease: A Multicenter, Prospective Cohort Study

ABSTRACT Background Intestinal ultrasound (IUS) is increasingly used to monitor activity and predict treatment outcomes in Crohn's disease (CD), but its role in predicting primary non‐response (PNR) to Upadacitinib (UPA) and identifying maintenance relapse remains unclear. Methods In this multicenter prospective cohort study, 122 patients with active CD treated with UPA were followed through week 52. Demographic, clinical, biochemical, endoscopic and IUS data were collected longitudinally. Univariable logistic regression, ROC analysis and likelihood ratio test (LRT) identified the best PNR predictors. Survival analysis evaluated the association between post‐induction ultrasound targets and the risk of relapse during 15‐mg maintenance. Results Among baseline IUS parameters, IBUS‐SAS > 60.20 demonstrated the highest odds ratio for predicting UPA PNR (OR 16.63, 95% CI 6.54–42.29, p < 0.001). Compared with clinical and endoscopic parameters, the inclusion of IBUS‐SAS significantly improved PNR prediction (AUC 0.865 vs. 0.670, LRT p < 0.001, 95% CI 0.802–0.929, Sensitivity 82.98%, Specificity 77.33%, PPV 69.64% and NPV 87.88%). Early at week 4, worse IUS findings were significantly correlated with PNR, with IBUS‐SAS > 47.80 showing the highest odds ratio (OR 26.25, 95% CI 2.46–280.20, p = 0.007). Among clinical and endoscopic responders, failure to achieve IUS response/remission was associated with an increased risk of relapse during 15‐mg maintenance. (Log‐rank p = 0.005, p < 0.001). Conclusion Early IUS monitoring provides incremental value for predicting UPA PNR beyond clinical and endoscopic parameters. Failure to achieve IUS response/remission may identify patients at increased risk of relapse during 15‐mg maintenance.