Validation of PERI-Gastric Models for Predicting Peritoneal Recurrence Post-Gastrectomy

Gastric cancer remains a significant global health issue, with radical gastrectomy being a common curative approach. However, peritoneal recurrence poses a major challenge post-surgery. This study aimed to validate the PERI-Gastric 1 and PERI-Gastric 2 prediction models using a large Korean cohort.

Study Design and Methods: The study retrospectively analyzed data from 8564 patients who underwent gastrectomy for stage Ib-IIIc gastric cancer at the Yonsei Cancer Center between 1998 and 2018. The performance of the PERI-Gastric models was evaluated through AUROC, calibration plots, and decision curve analysis.

Results:

1. Discrimination: PERI-Gastric 1 and PERI-Gastric 2 models demonstrated AUROC values of 0.766 and 0.767, respectively, indicating good discriminatory power.
2. Calibration: The models exhibited satisfactory calibration, particularly within the risk range of 24-33% for PERI-Gastric 1 and 35-47% for PERI-Gastric 2.
3. Clinical Utility: Decision curve analysis confirmed a positive net benefit within specific risk ranges (10-42% for PERI-Gastric 1 and 15-45% for PERI-Gastric 2).

Discussion: The external validation study confirms that the PERI-Gastric models are robust tools for predicting peritoneal recurrence in Eastern populations. The models’ ability to identify high-risk patients allows for tailored therapeutic strategies, potentially incorporating more aggressive perioperative treatments and stringent postoperative surveillance.

Conclusion: The PERI-Gastric 1 and 2 models provide reliable risk stratification for peritoneal recurrence after radical gastrectomy. Their implementation in clinical practice can enhance prognostic accuracy and guide treatment decisions, ultimately improving patient outcomes.

For detailed insights, refer to the full study here.