A Novel Exhaled Gas Detection Model for Diagnosis and Prognosis Prediction of Colorectal Cancer

Abstract

Abstract Background: Exhalation determination can detect intestinal metabolism-related gases. Colorectal cancer causes a significant proportion of global cancer morbidity and mortality. Our aim was to analyse the predictive value of a novel diagnostic model based on exhaled gas composition for detecting colorectal cancer. Methods: We recruited 40 patients diagnosed with colorectal cancer as an observation group and 40 healthy volunteers as a control group. The patients underwent surgical treatment at the Department of General Surgery, Huashan Hospital affiliated to Fudan University, from June 2018 to November 2019 and were diagnosed with colorectal cancer based on preoperative pathology. Exhaled gas components (EGCs) were collected using a respiratory analyser and analysed in both colorectal cancer patients and healthy controls. Results: The concentrations of four exhaled gases (H2, CH4, H2S and NO) were significantly higher in the colorectal cancer group than in the control group. We obtained the receiver operating characteristic (ROC) curves of these four gases. In addition, we constructed a new predictive model using these four gases according to logistic regression analysis with an area under the curve (AUC) of 0.962, a sensitivity of 92.5%, and a specificity of 97.5%. Moreover, multivariate Cox analysis showed that this model can serve as an independent prognostic method for colorectal cancer. Conclusion: Exhaled gas assessment has predictive value for colorectal cancer prognosis. The new model constructed using exhaled gases is a valuable noninvasive testing method that can be used as an auxiliary judgement tool before more invasive examinations.