Significance of the Glasgow prognostic score for short-term surgical outcomes: A nationwide survey using the Japanese National Clinical Database

Aim: Preoperative inflammation-based Glasgow prognostic score (GPS) is a useful tool for predicting long-term prognosis in cancer patients. However, its association with postoperative short-term outcomes remains unknown. The aim of this study is to investigate the association between GPS and postoperative morbidity and mortality among patients undergoing surgery for various gastrointestinal malignancies. Methods: Using the Japanese National Clinical Database, we analyzed the records of 312 357 patients with gastrointestinal malignancy who underwent six typical elective surgeries, including esophagectomy, distal gastrectomy, total gastrectomy, right hemicolectomy, low anterior resection, and pancreaticoduodenectomy, between January 2015 and December 2018. We assigned GPS of 0, 1, or 2 to patients with no, one, or both decreased albumin and elevated C-reactive protein levels, respectively. We investigated the relationship of GPS with operative morbidity and mortality for each procedure with adjustments for patients' demographics, preoperative status, comorbidities, and cancer stages. Results: Crude operative morbidity was significantly higher for GPS 1 and 2 than GPS 0 patients in all procedures except pancreaticoduodenectomy. The postoperative length of hospital stay was significantly longer for GPS 1 and 2 patients in all procedures (P <.001). Operative mortality was also higher for GPS 1 and 2 patients in all procedures. The associations remained significant after adjustments for potential confounders of age, sex, physical status, tumor classification, use of preoperative therapy, and comorbidities. Conclusion: This nationwide study provides solid evidence on the strong association between GPS and postoperative outcomes.