Comments on the J‐GESS Score: Exposure Classification, Follow‐Up, and Gastric Atrophy Assessment

To the Editor,

1.

We read with great interest the article by Watanabe et al., “Five‐Year Unified Risk Prediction Model for Gastric and Esophageal Cancer Based on Nationwide Health Check‐up Data in Japan: J‐ GESS Score” [1]. The authors deserve recognition for compiling a robust nationwide cohort to address long‐term patient surveillance optimization. Their effort to translate massive multicenter screening data into an accessible 5‐year prediction framework is highly noteworthy. While the study makes a valuable contribution, we would like to highlight several limitations to enhance the applicability of the clinical findings of the study.

First, smoking was modeled only as ever/never, which likely discarded clinically meaningful dose–response information. This coarse categorization can attenuate predictive performance because it treats light and heavy exposure as equivalent, despite strong evidence that tobacco exposure is dose‐related for upper gastrointestinal cancer risk and that dichotomizing predictors reduces information content in regression models. In a screening model intended to guide surveillance intensity, pack‐years or finer smoking categories would likely have preserved more prognostic signal and improved risk stratification at the upper end of the distribution [2, 3].

Eligibility required at least one follow‐up EGD, and more than 6000 screened individuals were excluded because follow‐up was unavailable. This may have introduced selection bias, because participants who returned for repeat EGD could differ systematically from nonreturners in baseline risk, health awareness, and access to care. As a result, the J‐GESS score may have been derived from a surveillance‐adherent subgroup rather than the full screening population. A model intended for population‐level screening should therefore be validated in a cohort with more complete outcome ascertainment or with methods that explicitly address differential follow‐up [4].

The J‐GESS score depends heavily on endoscopic gastric atrophy, which contributes 5 of the 9 total points, yet the classification was performed locally by site endoscopists without central adjudication or any reported interobserver agreement assessment. This is important because the score's performance depends on a subjective visual variable that may be prone to observer variability and misclassification. In a model intended for nationwide implementation, reproducibility of such a heavily weighted endoscopic predictor should be demonstrated more explicitly [5].

In summary, while the J‐GESS score is a practical tool for upper gastrointestinal cancer risk stratification, future validation studies addressing exposure granularity, follow‐up selection bias, and reproducibility of gastric atrophy grading may further strengthen its reliability and generalizability across broader screening populations.

Funding

The authors have nothing to report.

Ethics Statement

The authors have nothing to report.

Consent

The authors have nothing to report.

Conflicts of Interest

The authors declare no conflicts of interest.

Data Availability Statement

The authors have nothing to report.