by Christos Evangelou, MSc, PhD – Medical Writer and Editor 

A UK real-world implementation study involving 910 patients demonstrated that a simple capsule sponge test coupled with biomarker analysis can safely identify more than half of the patients with Barrett’s esophagus who could replace routine endoscopic surveillance with less invasive monitoring. According to the authors, this prospective implementation study represents a paradigm shift toward precision medicine in Barrett’s surveillance, offering clinicians a validated tool to stratify patients into meaningful risk categories while potentially transforming patient care pathways. 

The study was published in The Lancet. 

Study Rationale

Although endoscopic monitoring aims to detect dysplasia and early cancer when treatment outcomes are optimal, corresponding author Rebecca Fitzgerald, MD, of the University of Cambridge, explained that endoscopy for Barrett’s surveillance is not effective. “It is often performed by non-specialists and recent trials show that around 10% cases of dysplasia and cancer are missed,” she said. “This means that some patients re-present within a year of their surveillance procedure with a symptomatic cancer that should have been diagnosed earlier.” 

In addition, the progression rates from non-dysplastic Barrett’s disease to adenocarcinoma are low, making repetitive surveillance endoscopies costly for healthcare systems and burdensome for patients. Some healthcare systems are already abandoning surveillance for lower-risk patients, yet this approach risks missing treatable disease. 

Methodology

The DELTA consortium study recruited 910 patients with non-dysplastic Barrett’s esophagus from 13 UK hospitals between August 2020 and December 2024.1 The capsule sponge device was administered by trained nurses in under 10 minutes to collect cells from across the entire esophageal surface. 

The risk stratification algorithm combines clinical factors (age, sex, and segment length) with two biomarkers: p53 protein abnormalities and glandular atypia. Patients are classified into three groups: low risk (negative biomarkers, no clinical risk factors), moderate risk (negative biomarkers and positive clinical factors), and high risk (any positive biomarker regardless of clinical factors).1 

“We have previously shown that the capsule sponge test, coupled with a centralized lab test for p53 and atypia, can risk-stratify patients into low-, moderate-, and high-risk groups,” noted Fitzgerald, referencing their earlier validation study.2 

The current study aimed to confirm these findings in an unenriched, real-world population. 

Key Findings

Among the 489 patients (54% of the cohort) classified as low risk, only two developed high-grade dysplasia, with no cancers reported in this group. This prevalence of 0.4% (95% confidence interval [CI], 0.1-1.6) is below the 3% clinical threshold for primary care cancer referral.1 

The negative predictive value for any dysplasia in the low-risk group was 97.8% (95% CI, 95.9%-98.8%), suggesting that these patients could safely undergo capsule sponge surveillance instead of endoscopy.1 

“Our results showed that in the low-risk group, the risk of high-grade dysplasia or cancer was 0.4% suggesting that these patients could be offered follow-up with the capsule sponge test,” Fitzgerald stated. 

The high-risk group (138 patients; 15% of the cohort) demonstrated enrichment for dysplasia, with a positive predictive value of 37.7% (95% CI, 29.7%-46.4%) for any dysplasia.1 Patients with both glandular atypia and aberrant p53, designated as ‘high risk tier 1’, had an 85.2% prevalence of any dysplasia and 55.6% prevalence of high-grade dysplasia or cancer. 

“The high-risk group with a double biomarker positive, p53 and atypia, had an 85% risk of finding dysplasia or cancer; we call this a tier 1 or ultra-high risk, and this suggests these cases merit a specialist endoscopy in a center that could treat the dysplasia or cancer,” Fitzgerald emphasized. 

Technological Enhancement Through AI

The researchers incorporated an AI algorithm to streamline pathology workflows and address scalability concerns. The AI system achieved 100% sensitivity for detecting p53 overexpression while eliminating the need for pathologist review in 68% of cases by accurately identifying negative samples.1 

“The AI would assist the pathologist reporting the capsule sponge results to make it much faster,” Fitzgerald stated. “Any positive areas are pre-scored and highlighted so that the pathologist knows which areas to focus on.”  

Clinical Implications and Potential Implementation Timeline

According to the authors, the findings of the study support the need for restructuring surveillance protocols for Barrett’s esophagus. Rather than subjecting all patients to regular endoscopy regardless of risk, the capsule sponge enables personalized surveillance strategies. Low-risk patients could undergo nurse-led capsule sponge testing, whereas high-risk patients receive prioritized specialist endoscopy. 

 “Most patients with Barrett’s won’t develop cancer,” Fitzgerald noted. “Repeated endoscopy monitoring is stressful, and a simple non-endoscopic capsule sponge done nearer to home is less scary and could be less operator dependent.” 

The study showed that capsule sponge testing enabled clinicians to triage high-risk patients for endoscopy sooner (median 1.5 months) compared to low-risk patients (median 13.1 months). Fitzgerald emphasized that, by reducing endoscopy demand for low-risk patients, resources can be redirected toward high-risk cases requiring specialist intervention.  

Fitzgerald noted that guideline committees in the UK are already considering this evidence for their upcoming 2025 guidelines.

“The GI community is realizing that we need a better approach to managing patients with Barrett’s,” she said. “By reducing the burden of endoscopy in patients at very low risk, we can focus more on the patients at higher risk.” 

Fitzgerald also highlighted that the real-world design of the study and the implementation of the capsule sponge test across diverse hospital settings, including tertiary referral centers and district general hospitals, enhance the generalizability of the findings. Training requirements for staff proved manageable, and the centralized laboratory processing ensures standardized results. 

The study received financial support from Innovate UK, Cancer Research UK, and the National Health Service England Cancer Alliance. 

References

1. 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. 2025;406(10500):271-282. doi:10.1016/S0140-6736(25)01021-9 

1. Pilonis ND, Killcoyne S, Tan WK, et al. Use of a Cytosponge biomarker panel to prioritise endoscopic Barrett’s oesophagus surveillance: a cross-sectional study followed by a real-world prospective pilot. Lancet Oncol. 2022;23(2):270-278. doi:10.1016/S1470-2045(21)00667-7