RESUMEN
Objectives: Recently, some cohorts have looked into the use of Global Leadership Initiative on Malnutrition (GLIM) criteria in cancer patients. The objective of the current meta-analysis was to determine its utility in predicting clinical and survival outcomes for cancer patients. Method: Searching and screening literature from PubMed, Web of Science and Embase until September 13, 2022 was performed by two researchers independently. According to the exclusion and inclusion criteria, articles reporting the impact of malnutrition diagnosed by GLIM on long-term survival and clinical outcomes were included. Data of interest were also extracted from the included papers. The stability of the pooled results was evaluated using sensitivity analysis. With the aid of subgroup analysis, heterogeneity was revealed. To assess publication bias, Egger's and Begg's tests were conducted. The influence of publication bias on the pooling risk estimate was examined using a trim-and-fill analysis. Results: 15 studies that qualified for our study were identified. Pooled hazard ratio (HR) from both multivariate and univariate regression analysis showed a worse overall survival in GLIM-defined malnourished cancer patients than those in well-nourished status. Meanwhile, disease-free survival was also poorer in malnourished patients. Moreover, pooled odds ratio (OR) demonstrated that malnourished cancer patients were more likely to develop overall postoperative complications, complications ≥ Clavien-Dindo grade IIa and complications ≥ Clavien-Dindo grade IIIa. Two articles reported negative relation between GLIM-defined malnutrition and 30-day readmission/mortality. Conclusion: GLIM-defined malnutrition possesses value in predicting poorer survival and clinical outcomes for cancer patients. Systematic review registration: [https://www.crd.york.ac.uk/PROSPERO/display_record.php?RecordID=321094], identifier [CRD42022321094].
