An exploratory study on the usefulness of pre-albumin and CONUT score in predicting the prognosis of gastrointestinal cancer

Abstract

This study aimed to predict the short-term prognosis of patients with unresectable advanced recurrent digestive cancers. We analyzed the correlation between BMI, pre-albumin, zinc, CRP, total cholesterol, CONUT score, PNI and survival time using Spearman’s rank correlation coefficient and a multiple regression analysis in patients who visited our palliative outpatient clinic. The results of the analysis showed that the pre-albumin level and CONUT score were correlated. Using the obtained multiple regression equation values, a cutoff value was calculated from the ROC curve and validated in patients who were receiving chemotherapy in an oncology outpatient clinic. The results showed that it was statistically possible to predict the prognosis on a monthly basis. In clinical practice, chemotherapy for advanced solid tumors is sometimes continued until near the end of life owing to the lack of clear prognostic criteria. Our findings suggest that nutritional indices such as pre-albumin and CONUT score may serve as objective indicators to support timely decision-making regarding the discontinuation of chemotherapy.