Symptoms correlate only modestly with endoscopic and histological activity in Eosinophilic Oesophagitis (EoE). Real-world data show that symptoms detect histological remission with only ~60% accuracy (AUC 0.60–0.61) [1]. Persistent histological activity is potentially associated with progression to fibrostenotic disease, highlighting the importance of detecting subclinical inflammation. UK, US, and European guidelines therefore recommend reassessment after induction or treatment change to detect unrecognised persistent inflammation [2-4]. Whether all symptom-free patients require routine gastroscopy in real-world practice remains uncertain.

This study aims to determine whether clinical remission predicts histological remission in treated EoE, and whether a targeted rather than universal reassessment strategy may be appropriate.

We conducted a prospective single-centre cohort study from September 2023 to November 2025. Adults with confirmed EoE underwent reassessment after ≥8 weeks of stable therapy (food elimination diet, proton pump inhibitor, budesonide orodispersible tablets (BOTs), or combinations). All patients completed the Dysphagia Symptom Questionnaire (DSQ) and underwent same-day capsule sponge cytology (Endosign) and gastroscopy with biopsies, as part of another study. Clinical remission was defined as DSQ=0. Histological remission was defined as highest peak eosinophil count <15 eos/hpf on capsule sponge or histology. Fisher’s exact test, Mann–Whitney U testing, sensitivity, specificity, predictive values and Cohen’s kappa were calculated.

A total of 66 patients underwent paired symptom and tissue assessment. Median age was 50 years (IQR 36.5–57.0); 64% were male and 91% were White or White British. Of the 66 patients, 29 had DSQ=0 and 37 had DSQ≥1. Histological remission occurred in 55/66 (83.3%).

Among DSQ=0 patients, 24/29 (82.8%) were in histological remission and 5/29 (17.2%) had active eosinophilia (median 20 eos/hpf, IQR 19–30). In the DSQ=0 group, the median total EREFS score was 2 in histological remission, and 4 in those with active disease. Of the 5 DSQ=0 patients with active eosinophilia at follow-up, 2 reported variable compliance with Jorveza, 2 had marked symptomatic improvement despite persistent inflammation, and 1 had insufficient dietary elimination.

Among DSQ≥1 patients, 31/37 (83.8%) were in histological remission and 6/37 (16.2%) were histologically active (median 30 eos/hpf, IQR 27–37.5). In the DSQ≥1 group, the median total EREFS score was 2 in both remission and active disease. Within the DSQ≥1 group, DSQ scores were significantly higher in histologically active patients compared with those in histological remission (median 34.5 [IQR 18–42] vs 9 [IQR 3–14]; Mann–Whitney U p=0.016).

Median treatment duration was 16.3 weeks (IQR 13.0–19.6) for DSQ=0 patients and 14.9 weeks (IQR 10.4–21.9) for DSQ≥1 patients.

Clinical remission did not correlate with histological remission (Fisher’s exact p=0.76; κ=–0.04). Using DSQ≥1 to detect histologic activity gave a sensitivity of 55%, specificity 44%, PPV 16%, and NPV 83%.

Despite clinical remission, 17.2% of asymptomatic patients had persistent eosinophilic inflammation. As these are a minority of DSQ=0 patients demonstrated active disease, these findings could support a targeted reassessment strategy, reserving gastroscopy for individuals with higher-risk features i.e. Higher EREFS at index endoscopy. The results also highlight the potential role of less invasive monitoring tools, such as capsule sponge cytology, and the need for biomarkers beyond eosinophil counts alone.