Development and Validation of an Individualized Nomogram to Identify Undifferentiated-Predominant Mixed-Type Early...

Area of Science:

• Gastroenterology
• Oncology
• Medical Diagnostics

Background:

• Diagnosing undifferentiated-type-predominant mixed-type early gastric cancer (UM-EGC) is challenging due to its complex histology.
• Accurate preoperative diagnosis is crucial for effective endoscopic treatment planning.

Purpose of the Study:

• To develop and validate a novel nomogram for predicting UM-EGC.
• To identify key clinicopathological and endoscopic features associated with UM-EGC.

Main Methods:

• A retrospective study included 808 early gastric cancer (EGC) patients undergoing endoscopic submucosal dissection.
• Logistic regression analysis identified predictors for UM-EGC.
• A predictive nomogram was constructed and validated in training and external cohorts.

Main Results:

• Key predictors for UM-EGC included specific atrophic gastritis patterns, lesion types (IIb, IIc), discolored appearance, location, and prior H. pylori eradication.
• The nomogram showed good discriminative ability (AUC 0.83 in training, 0.82 in validation).
• UM-EGC was associated with a higher risk of metachronous cancer.

Conclusions:

• The developed nomogram effectively predicts UM-EGC risk preoperatively.
• This tool can assist clinicians in managing EGC patients.
• Further validation in diverse populations may enhance its clinical utility.