# Validation of the predictive value of combined prealbumin and lymphocyte score for prognosis of stage II/III gastric cancer following curative resection

## TL;DR

This study validates that combining prealbumin and lymphocyte levels can predict survival outcomes in gastric cancer patients after surgery.

## Contribution

The Co-PaL score is validated as a new prognostic tool for stage II/III gastric cancer patients post-surgery.

## Key findings

- Co-PaL score, TNM stage, complications, and chemotherapy are independent predictors of survival.
- The nomogram based on these variables outperformed the TNM staging system in predicting survival.
- The model showed strong accuracy with a C-index of 0.701 and high AUC values for 1-5 year survival.

## Abstract

Our previous study found that the combined prealbumin and lymphocyte (Co-PaL) score could accurately classify patients into severe, mild to moderate malnutrition and good nutrition, and might be a predictor for prognosis of patients undergoing gastrectomy for stage II/III gastric cancer (GC). The aim of the present study was to validate these findings.

The medical records of stage II/III GC patients undergoing curative resection in our hospital from January, 2017 to December, 2023 were retrospectively reviewed. Basing on whether the lymphocyte count was <1.5 ×109/L and/or the prealbumin concentration <180 mg/L, patients were assigned a Co-PaL score of 0, 1 or 2, respectively. A nomogram was established basing on independent predictors for OS identified by univariate and multivariate Cox regression analyses. Concordance index and calibration curves were used to evaluate the nomogram. Clinical utility and predictive accuracy were further assessed by net reclassification index (NRI), integrated discrimination improvement (IDI) and decision curve analysis (DCA).

A total of 890 consecutive patients were recruited. Multivariate regression analyses revealed that Co-PaL score, TNM stage, post-operative complications and adjuvant chemotherapy were independent predictors for OS. A nomogram based on these four variables was established. The C-index value obtained for the model was 0.701 (95%CI: 0.672-0.729). The area under the curve (AUC) values to predict the 1- 3- and 5-year survival probabilities were 0.709 (95%CI: 0.662-0.756), 0.728 (95%CI: 0.692-0.764) and 0.734 (95%CI: 0.695-0.7772), respectively. The calibration curves represented fine consistency between the actual and predicted 1-, 3- and 5-year survival probabilities. Compared with TNM staging system, our model demonstrated strong accuracy, discriminative ability, and clinical utility.

The Co-PaL score was a simple and promising predictor for prognosis of patients undergoing gastrectomy for stage II/III GC. The established nomogram showed superiority over TNM staging system in predicting OS.