A Novel Predictive Model for Cancer-Specific Surgical Risk Stratification of Patients Undergoing High-Risk Upper...

Area of Science:

• Oncology
• Surgical Oncology
• Gastroenterology

Background:

• High-risk oncologic resections are crucial for curative treatment of upper gastrointestinal (GI) tumors.
• Existing risk-stratification models lack cancer-specificity for these complex operations.

Purpose of the Study:

• To develop and validate cancer-specific nomograms for predicting postoperative complications and length of stay (LOS) after upper GI oncologic resections.
• To provide clinicians with tools for better risk estimation and outcome prediction in patients undergoing major GI surgery.

Main Methods:

• Retrospective analysis of 2823 patients undergoing upper GI oncologic resections (gastrectomy, esophagectomy, pancreaticoduodenectomy) for adenocarcinoma from 2015-2022.
• Development of multivariable models to identify independent risk factors for LOS and complications, followed by nomogram construction.
• Internal validation using bootstrap resampling to assess model performance (discrimination and calibration).

Main Results:

• The median LOS was 7.0 days, with a 31% overall 30-day postoperative complication rate.
• Independent predictors for increased LOS included age, ASA class >3, COPD, renal failure, and longer operation duration; neoadjuvant chemotherapy (NAC) and pre-incision antibiotics reduced LOS.
• The complication nomogram achieved a C-index of 0.66, and the LOS nomogram showed a Kendall's Tau of 0.39. Key predictors included procedure type, age, COPD, NAC, race, and BMI.

Conclusions:

• Developed cancer-specific pre- and postoperative nomograms offer valuable tools for estimating risks and outcomes.
• These nomograms aid in patient counseling and optimizing treatment strategies for high-risk upper GI oncologic resections.