Early diagnosis of cancer is important for improving patient outcomes and reducing the treatment costs. Less invasive diagnostic methods, such as the use of biomarkers in direct tissue samples, have shown promise in detecting precancerous conditions that are more amenable to treatment.
Barrett’s oesophagus, which is a precursor lesion for oesophageal adenocarcinoma, is a challenging and often fatal cancer with a poor 5-year overall survival rate. Detecting early cancer or dysplasia during endoscopy is difficult and depends on the quality of the endoscopic equipment and the operator’s ability. Therefore, there is a requirement to alter the surveillance method, which depends on an operator-independent and high-quality method that is acceptable to patients and cost-effective for healthcare management.
A novel approach using a non-endoscopic cell sample device (capsule sponge) combined with centralized, pathology-based biomarker analysis has been developed and tested. This capsule sponge aims to stratify patients into different risk categories to guide the appropriate timing of follow-up endoscopy. Its implementation in routine clinical practice was assessed in this study.
This prospective, multicentre, pragmatic implementation study was conducted in the United Kingdom among patients with a confirmed diagnosis of non-dysplastic Barrett’s oesophagus who underwent subsequent endoscopy using the capsule sponge. Participants were consecutive patients undergoing Barrett’s oesophagus surveillance at 13 hospitals as part of the National Health Service England (NHSE) diagnostic solution for the early detection of oesophageal cancer (DELTA) research.
Inclusion criteria included patients aged at least 18 years with a diagnosis of non-dysplastic Barrett’s oesophagus undergoing surveillance per UK guidelines. All patients had to swallow the capsule sponge successfully and had a confirmatory endoscopy. Exclusion criteria were a history of high-grade dysplasia, previous surgical or endoscopic treatment of the oesophagus, lack of capacity to give informed consent, or contraindications to the capsule sponge as per the manufacturer’s guidelines.
Patients were classified as low risk (negative capsule sponge and clinical biomarkers), mild risk (positive clinical biomarker and negative capsule sponge biomarker – sex, segment length, and age), or high risk (glandular atypia or p53 abnormality independent of clinical biomarker or both). The primary result was the diagnosis of cancer requiring treatment and high-grade dysplasia, as correlated with assigned risk categories.
About 910 patients were enrolled from August 2020 to December 2024. Four hundred eighty-nine patients (54%) were at low risk, 138 patients (15%) were at high risk, and 283 patients (31%) were at mild risk. The positive risk factor for any dysplasia or greater in the high-risk group was 37.7% (95% confidence interval [CI] 29.7 to 46.4). Patients who have both abnormal p53 and atypia had a significantly elevated risk of cancer or high-grade dysplasia (135.8% [95% CI 32.7 to 564.0] compared to the low-risk group. The rate of cancer or high-grade dysplasia in the low-risk group was 0.4% (95% CI 0.1% to 1.6%), with a negative risk score for cancer or any dysplasia of 97.8% (95% CI 95.9 to 98.8) by using machine learning algorithm in digital pathology process decrease proportion which needs p53 pathology evaluation to 32% without losing any positive case.
Limitations of the study included the inability to differentiate between incidental and prevalent dysplasia, the potential to misclassify patients, and the lack of centralized pathology evaluation. Future study needs to focus on improving risk stratification and establishing a unified computational pipeline for atypia prediction. This study can clear the path for other precancer initiatives. The risk panel identifies dysplasia and suggests capsule sponge-based monitoring as a viable alternative to endoscopy in low-risk Barrett’s oesophagus.
Reference: Tan WK, Ross-Innes CS, Somerset T, et al. Biomarker risk stratification with capsule sponge in the surveillance of Barrett’s oesophagus: prospective evaluation of UK real-world implementation. Lancet. 2023. doi:10.1016/s0140-6736(25)01021-9
