Benchmarking Detachable-string Magnetic Capsule Endoscopy against the Right Standard for Early Esophageal Squamous Cell Cancer

Letter to: Exploratory evaluation of magnetic capsule endoscopy with detachable-string for detecting esophageal lesions including early squamous cell carcinoma Endosc Int Open 2026; 14(CP): DOI: 10.1055/a-2823-7782 To the Editor, We read with interest the exploratory study by Dai et al. evaluating detachable-string magnetically controlled capsule endoscopy (DS-MCCE) for esophageal lesions including early esophageal squamous cell carcinoma (ESCC).[ 1 ] The authors should be congratulated for testing a minimally invasive, patient-friendly platform in a clinically important setting. Their safety, comfort, and preference data are encouraging. We are concerned, however, that the diagnostic performance of DS-MCCE is positioned against an inappropriately weak benchmark. In the Discussion, the authors state that the 66.67% sensitivity of DS-MCCE for early ESCC “approached” that of conventional white-light endoscopy (55%).[ 1 ] However, the relevant comparator in contemporary practice is image-enhanced endoscopy and not historical white-light imaging alone. In the authors’ own study, the reference examination included narrow-band imaging (NBI) in every patient, with Lugol chromoendoscopy applied to suspicious lesions. In the randomized trial cited by the authors, NBI detected superficial ESCC substantially more often than white-light imaging (97% vs. 55%).[ 2 ] Meta-analytic data likewise report per-patient sensitivities of approximately 88% for NBI and 92% for Lugol chromoendoscopy.[ 3 ] Against that standard, a sensitivity of 66.67% (95% CI 38.38–88.18%) cannot yet be regarded as competitive for screening. The authors further suggest that missed lesions were mainly smaller or morphologically subtle.[ 1 ] However, their own Table 3 shows that the limitation is not confined to subcentimeter subtle abnormalities. One 2.0 cm high-grade dysplasia (HGD) lesion was interpreted as “Negative,” and another 1.0 cm HGD lesion was classified as reflux esophagitis. The latter is particularly concerning, because a false-benign label may reduce the perceived need for confirmatory endoscopy. An additional interpretive caution is that 27 examinations without suspicious findings on DS-MCCE were excluded from reference verification.[ 1 ] The authors appropriately acknowledge this limitation, but it means additional false-negatives could have remained undetected. DS-MCCE should therefore be viewed as an interesting exploratory platform rather than a clinically benchmarked screening tool. The next step should be a prospective head-to-head study against contemporary image-enhanced endoscopy, with explicit accounting for false-negative and false-benign classifications. This is especially important because nonendoscopic alternatives are advancing rapidly; for example, AI-assisted sponge cytology has already shown high sensitivity and specificity at screening scale.[ 4 ] Publication History Received: 23 April 2026 Accepted: 02 June 2026 Article published online: 25 June 2026 © 2026. The Author(s). This is an open access article published by Thieme under the terms of the Creative Commons Attribution License, permitting unrestricted use, distribution, and reproduction so long as the original work is properly cited. (https://creativecommons.org/licenses/by/4.0/). Georg Thieme Verlag KG Oswald-Hesse-Straße 50, 70469 Stuttgart, Germany Bibliographical Record Ervin Toth, Anastasios Koulaouzidis. Benchmarking Detachable-string Magnetic Capsule Endoscopy against the Right Standard for Early Esophageal Squamous Cell Cancer. Endosc Int Open 2026; 14: a28890416. DOI: 10.1055/a-2889-0416