S904 I-SEE Score Components in Real-World Eosinophilic Esophagitis Care: TARGET-GASTRO

Introduction: Eosinophilic esophagitis (EoE) is a chronic disease typified by inflammation, difficulty swallowing, food impaction, and other esophageal symptoms. The Index of Severity for EoE (I-SEE) is a clinical severity score covering four domains: symptoms, complications, inflammatory features, and fibrostenotic features. However, it is unclear how well these components are captured in real world records, potentially limiting the determination of I-SEE. We evaluated I-SEE component frequency in EoE patients in the real world. Methods: From a multicenter observational gastrointestinal patient registry, TARGET-GASTRO, we selected adult patients with both an EoE diagnosis (International Classification of Diseases-10 K20.0) and endoscopy on or after their first documented visit. We descriptively analyzed components of the I-SEE leading up to and surrounding that endoscopy (index) to assess the extent of documentation. Results: At index, 1,743 patients qualified, their median age was 41 years, 55% of subjects were men, 90% White, and 7% Black. The presence of any symptom was documented in 91% of all patients with 45% of patients describing both esophageal-specific and non-specific symptoms. Symptom frequency was noted in 81% of patients with documented symptoms, including 39% experiencing weekly symptoms and 34% reporting symptoms occurring multiple times per day or disrupting social functioning. Overall, 18% of all patients had EoE-related complications; the most common was food impaction requiring emergency room visit or endoscopy (10%). Only 5% had persistent inflammation requiring elemental formula, systemic corticosteroids, or immunomodulators. Inflammatory features seen on endoscopy were reported in 29% (furrows as the most common in 23%) and among these, 47% had localized vs 36% diffuse features. Nearly half (49.7%) of patients with histology data lacked quantitative assessment of eosinophils. Fibrostenotic features (rings or stricture) were reported in 44% and of these, 42% underwent dilation at some point. Conclusion: Although I-SEE domains are present and in part reported in real-world practice, clinical documentation often lacks details needed to calculate a full I-SEE score. In particular, incomplete documentation of symptom frequency and eosinophil counts hinder I-SEE score calculation. More detailed documentation of EoE-specific symptoms, endoscopic features, and histology, potentially assisted by the I-SEE scoring app, could enhance the utility of I-SEE in treatment decisions.