Morphology‐Based Cytosponge Cytology Versus Concurrent Endoscopic Biopsy for Detecting Dysplasia in Barrett's Oesophagus Surveillance: A Prospective, Single‐Centre Paired Study
Vincent Labador, Cherry Galorport, Hamid Masoudi, Robert Enns, Eric Lam, Neal Shahidi, Pam Wong, Wei XiongABSTRACT
Background
Barrett's oesophagus (BE) is the principal precursor of oesophageal adenocarcinoma, and guidelines recommend endoscopic surveillance to detect dysplasia. Endoscopy is invasive and resource‐intensive. The Cytosponge is a non‐endoscopic cell‐collection device. We evaluated Cytosponge cytology against concurrent endoscopic biopsy for detecting dysplasia in a BE surveillance population, and explored factors associated with sample adequacy.
Methods
In this prospective, single‐centre study, patients with established BE underwent a Cytosponge procedure with a concurrent same‐visit endoscopic biopsy. Cytosponge cell blocks were stained with H&E and Alcian Blue 2.5/Periodic Acid‐Schiff. Two blinded gastrointestinal pathologists recorded goblet cells and dysplasia by consensus. Sensitivity, specificity, and predictive values with 95% confidence intervals (CI) were calculated against biopsy, and age‐adequacy association tested with Fisher's exact test.
Results
Fifty‐nine participants were analyzed (median age 70 years; 93.2% male). Cytosponge samples were adequate in 46/59 (78.0%; 95% CI 65.9–86.6). Among 44 adequate paired samples, sensitivity for dysplasia was 76.5% (95% CI 52.7–90.4), specificity 92.6% (76.6–97.9), positive predictive value 86.7%, and negative predictive value 86.2%. Sensitivity for definite (low‐ plus high‐grade) dysplasia was 85.7% (12/14); the lower overall figure reflected indefinite‐for‐dysplasia cases, of which only one of three was detected. Inadequacy was higher in patients aged ≥ 71 than ≤ 70 years (30% vs. 16%; p = 0.22). No procedural complications occurred.
Conclusions
Cytosponge cytology showed high specificity and moderate sensitivity for dysplasia, with a positive result strongly predicting biopsy‐confirmed dysplasia. These pilot data support further evaluation of Cytosponge cytology, reported by trained pathologists, as a non‐endoscopic triage tool in BE surveillance.