A scoring system to predict the technical difficulty of endoscopic resection for cardial submucosal tumors

Area of Science:

• Gastroenterology
• Surgical Oncology
• Medical Technology

Background:

• Endoscopic resection is effective for digestive submucosal tumors (SMTs).
• The cardia presents unique challenges for endoscopic resection due to anatomical constraints.
• Predicting technical difficulty is crucial for successful cardial SMT removal.

Purpose of the Study:

• To develop a clinical scoring model for grading the technical difficulty of endoscopic resection for cardial SMTs.
• To provide clinicians with a tool for better procedural planning and risk assessment.

Main Methods:

• Retrospective analysis of 246 patients undergoing endoscopic resection for cardial SMTs.
• Patients were divided into training (n=123) and validation (n=123) cohorts.
• Logistic regression identified predictors (male gender, extraluminal growth, tumor diameter ≥3 cm) for a scoring system.

Main Results:

• The developed scoring model incorporates male gender (2 points), extraluminal growth (3 points), and maximum diameter ≥3 cm (3 points).
• The model demonstrated strong discriminatory power (AUC=0.860) and good fit (Hosmer-Lemeshow P=0.979).
• The model accurately predicted varying probabilities of technical difficulty across different score categories in both training and validation cohorts.

Conclusions:

• The established clinical scoring system is a valuable tool for predicting the technical difficulty of endoscopic resection for cardial SMTs.
• This model can assist clinicians in anticipating challenges and optimizing surgical strategies.