Different DLCO Parameters as Predictors of PostoperativePulmonary Complications in Mild Chronic Obstructive Pulmonary Disease Patients with Lung Cancer
- Author:
Mil Hoo KIM
1
;
Joonseok LEE
;
Joung Woo SON
;
Beatrice Chia-Hui SHIH
;
Woohyun JEONG
;
Jae Hyun JEON
;
Kwhanmien KIM
;
Sanghoon JHEON
;
Sukki CHO
Author Information
- Publication Type:Clinical Research
- From: Journal of Chest Surgery 2024;57(5):460-466
- CountryRepublic of Korea
- Language:English
-
Abstract:
Background:Numerous studies have investigated methods of predicting postoperative pulmonary complications (PPCs) in lung cancer surgery, with chronic obstructive pulmonary disease (COPD) and low forced expiratory volume in 1 second (FEV1 ) being recognized as risk factors. However, predicting complications in COPD patients with preserved FEV 1 poses challenges. This study considered various diffusing capacity of the lung for carbon monoxide (DLCO ) parameters as predictors of pulmonary complication risks in mild COPD patients undergoing lung resection.
Methods:From January 2011 to December 2019, 2,798 patients undergoing segmentectomy or lobectomy for non-small cell lung cancer (NSCLC) were evaluated. Focusing on 709 mild COPD patients, excluding no COPD and moderate/severe cases, 3 models incorporating DLCO , predicted postoperative DLCO (ppoDLCO ), and DLCO divided by the alveolar volume (DLCO /VA) were created for logistic regression. The Akaike information criterion and Bayes information criterion were analyzed to assess model fit, with lower values considered more consistent with actual data.
Results:Significantly higher proportions of men, current smokers, and patients who underwent an open approach were observed in the PPC group. In multivariable regression, male sex, an open approach, DLCO <80%, ppoDLCO <60%, and DLCO /VA <80% significantly influenced PPC occurrence. The model using DLCO /VA had the best fit.
Conclusion:Different DLCO parameters can predict PPCs in mild COPD patients after lung resection for NSCLC. The assessment of these factors using a multivariable logistic regression model suggested DLCO /VA as the most valuable predictor.
