Prognostic role of minute ventilation/carbon dioxide production slope for perioperative morbidity and long-term survival in resectable patients with non-small-cell lung cancer: A prospective study using propensity score overlap weighting.

Abstract

The role of minute ventilation/carbon dioxide production (/CO2) slope, a ventilation efficiency marker, in predicting short- and long-term health outcomes for patients with non-small-cell lung cancer (NSCLC) undergoing lung resection has not been well investigated. This prospective cohort study consecutively enrolled NSCLC patients who underwent a presurgical cardiopulmonary exercise test from November 2014 to December 2019. The association of /CO2 slope with relapse-free survival (RFS), overall survival (OS), and perioperative mortality was evaluated using the Cox proportional hazards and logistic models. Covariates were adjusted using propensity score overlap weighting. The optimal cut-off point of the E/CO2 slope was estimated using the Receiver Operating Characteristics curve. Internal validation was completed through bootstrap resampling. A cohort of 895 patients (median age [interquartile range], 59 [13] years; 62.5% male) was followed for a median of 40 (range, 1-85) months. Throughout the study there were 247 relapses or deaths and 156 perioperative complications. The incidence rates per 1000 person-years for relapses or deaths were 108.8 and 79.6 among patients with high and low E/CO2 slope, respectively (weighted incidence rate difference per 1000 person-years, 29.21 [95% CI, 7.30 to 51.12]). A E/CO2 slope of ≥31 was associated with shorter RFS (hazard ratio for relapse or death, 1.38 [95% CI, 1.02 to 1.88], P=0.04) and poorer OS (hazard ratio for death, 1.69 [1.15 to 2.48], P=0.02) compared to a lower /CO2 slope. A high E/CO2 slope increased the risk of perioperative morbidity compared with a low E/CO2 slope (odds ratio, 2.32 [1.54 to 3.49], P<0.001). In patients with operable NSCLC, a high E/CO2 slope was significantly associated with elevated risks of poorer RFS and OS, and perioperative morbidity.