Construction and validation of a nomogram for predicting pathological upgrading/progression of gastric low-grade...

To analyze the clinical, endoscopic, and pathological characteristics of gastric low-grade intraepithelial neoplasia (LGIN), identify independent risk factors for pathological upgrading/progression, and construct and internally validate a nomogram for individualized risk stratification.

We retrospectively enrolled 278 patients with biopsy-confirmed gastric LGIN who underwent regular endoscopic follow-up at the Digestive Endoscopy Center of the Affiliated Hospital of Qingdao University between November 2016 and June 2023. The primary endpoint was pathological upgrading/progression, defined as either (1) true longitudinal progression from LGIN to high-grade intraepithelial neoplasia (HGIN) or gastric cancer (GC) during surveillance biopsy follow-up, or (2) histological upgrading to HGIN or GC in endoscopic submucosal dissection (ESD) specimens after an initial biopsy diagnosis of LGIN, reflecting possible biopsy underestimation. Univariable analyses were first performed, and variables with P < 0.05 were entered into a multivariable logistic regression model using backward stepwise selection. A nomogram was developed based on the final independent predictors. Discrimination and calibration were evaluated using the area under the receiver operating characteristic curve (AUC), bootstrap internal validation (1,000 resamples), calibration plots, and the Hosmer-Lemeshow test.

Pathological upgrading/progression occurred in 49 (17.63%) patients during follow-up. Multivariate analysis identified four independent risk factors for LGIN upgrading/progression: advanced age (OR = 1.07, 95% CI: 1.01-1.12, P = 0.018), larger lesion diameter (OR = 1.08, 95% CI: 1.03-1.15, P = 0.004), enlarged gastric folds (OR = 3.23, 95% CI: 1.03-10.11, P = 0.044), and irregular microvascular pattern (OR = 13.26, 95% CI: 1.70-103.31, P = 0.014). The nomogram constructed based on these factors showed good discriminative performance, with an area under the ROC curve (AUC) of 0.84 (95% CI: 0.78-0.89). Bootstrap internal validation (1,000 repetitions) and Hosmer-Lemeshow test (P = 0.333) confirmed high consistency between predicted and observed outcomes, indicating satisfactory model accuracy.

A nomogram based on age, lesion diameter, enlarged gastric folds, and irregular microvascular pattern provides good predictive value for pathological upgrading/progression in gastric LGIN. This tool may help clinicians optimize individualized surveillance and treatment strategies. Nevertheless, because the composite endpoint includes both true progression and histological upgrading after ESD, the model should be interpreted as predicting the risk of clinically significant pathological upgrading/progression rather than pure natural-history progression alone.