Clinical utility of the modified Glasgow Prognostic Score to classify cachexia in patients with advanced cancer in palliative care

Background & aims: It is a challenge in clinical practice to identify and classify cancer cachexia. Currently, it has been extensively discussed if the presence of alterations in inflammatory biomarkers implies the presence of cachexia. This study aimed to evaluate the clinical relevance of cachexia classification through modified Glasgow Prognostic Score (mGPS) in advanced cancer patients in palliative care. Methods: Observational prospective cohort study conducted at a Palliative Care Unit in Brazil. Cachexia classification was performed according to mGPS (based on albumin and C-reactive protein) in four different stages: no cachexia (NCa), undernourished (Un), pre cachexia (PCa), and refractory cachexia (RCa). Logistic regression models were used to test the association between cachexia stages and clinical, nutritional and functional domains. Kaplan-Meier curve and Cox multivariate model were used to analyze overall survival (OS). Results: A total of 1166 patients were included in the study. According to the cachexia framework 37.5% were NCa, 32.3% Un, 3.9% PCa and 26.4% RCa. Significant differences were observed among cachexia stages for most of the outcome measures. This classification was able to predict mortality in 90 days [Un (HR, 1.55; 95% CI, 1.25; 1.93); PCa (HR, 2.00; 95% CI, 1.34; 2.98); RCa (HR, 2.45; 95% CI, 1.34; 2.98)]. Conclusion: Cachexia stages were associated with significant differences in poor clinical outcomes and were also capable of predicting OS. This framework based on simple and objective criteria can be used as part of the routine to characterize the presence and stages of cachexia in advanced cancer patients. (C) 2019 Elsevier Ltd and European Society for Clinical Nutrition and Metabolism. All rights reserved.