Depth-predicting score for differentiated early gastric cancer

Area of Science:

• Gastroenterology
• Oncology
• Endoscopic Imaging

Background:

• Differentiated early gastric cancers (EGCs) confined to the intramucosal or minute submucosal layer (M-SM1; <500 μm) are candidates for endoscopic resection due to low metastasis risk.
• Distinguishing M-SM1 EGCs from those with deeper submucosal invasion (SM2; ≥500 μm) endoscopically remains challenging.
• Limited research exists on endoscopic features differentiating M-SM1 and SM2 differentiated gastric cancers.

Purpose of the Study:

• To identify endoscopic features associated with submucosal invasion depth in differentiated EGCs.
• To develop and validate a predictive scoring model for estimating EGC invasion depth.
• To aid in treatment strategy decisions for M-SM1 differentiated gastric cancers.

Main Methods:

• Retrospective analysis of 853 differentiated EGCs (derivation group) and 211 validation samples.
• Evaluation of endoscopic images for correlations between invasion depth and features: tumor location, macroscopic type, size, remarkable redness, uneven surface, margin elevation, ulceration, and enlarged folds.
• Development of a depth-predicting score based on logistic regression analysis.

Main Results:

• Tumor size >30 mm, remarkable redness, uneven surface, and margin elevation were significantly linked to SM2 invasion.
• A scoring model was created: 2 points for margin elevation and tumor size >30 mm; 1 point for remarkable redness and uneven surface.
• For lesions with a score ≥3, sensitivity ranged from 29.7-45.9%, specificity from 93.1-93.7%, and accuracy from 82.5-84.8% in predicting SM2 invasion.

Conclusions:

• The developed depth-predicting score is a valuable tool for guiding treatment decisions in differentiated M-SM1 EGCs.
• Accurate endoscopic assessment of invasion depth is crucial for selecting appropriate resection strategies.
• Further validation may enhance the clinical utility of this scoring system.