Updated Analysis of the Impact of Pulmonary Function Tests on Complications After Lung Resection

B. Wallen¹, N. Kapula², L. L. Tsai³, B. Guenthart⁴, D. Liou², N. Lui⁵, L. Backhus⁶, M. Berry⁵, J. Shrager², I. Elliott^7
1Stanford Medical School, Stanford, California ²Stanford University School of Medicine, Stanford, California ³Stanford, Menlo Park, California ⁴Stanford Health Care, San Jose, California ⁵Stanford University, Stanford, California ⁶Stanford University, Dept. of Cardiothoracic Surgery, Stanford, California ⁷Stanford, Stanford, California

Purpose: Most data regarding the prognostic value of pulmonary function tests (PFTs) are >10 years old, and guidelines for preoperative pulmonary evaluation have not changed since 2013.1,2 With increased adoption of minimally-invasive surgery and sub-lobar resection, updated outcomes data are important for counseling higher risk patients undergoing lung surgery.

Methods: Using our institution’s General Thoracic Surgery STS Database, we performed a retrospective analysis of patients who underwent elective lung resection from 2013-2023. A composite “respiratory complication” outcome was created, which included air leak, pneumonia, pleural effusion or pneumothorax requiring drainage, ARDS, respiratory failure, bronchopleural fistula, ventilatory support >48 hours, and other pulmonary events as coded in the STS Database. FEV1 and DLCO were stratified in to 20% interval quantiles ( < 40, 40-60, 60-80, and >80) and postoperative outcomes were examined. Then, using FEV1 and DLCO as continuous variables, and including clinically relevant patient and operative characteristics, multivariable models were created to determine the influence of various factors on postoperative respiratory complications, length of hospital stay, and 30-day mortality.

Results: We identified 1,532 patients undergoing pulmonary resection; baseline characteristics are shown in Table 1. Postoperative respiratory complications occurred in 8.6%, median length of stay (LOS) was 4 days, and 30-day mortality was 1.1%.

Examination of outcomes by FEV1 and DLCO by 20% interval quantiles suggested a linear increase in respiratory complications with decreasing FEV1 and DLCO (Fig. 1a). However, in multivariable analysis, lower FEV1 (p < 0.001) predicted respiratory complications while DLCO was not statistically significant. Other predictors were older age (p=0.007), male gender (p=0.007), current smoking status (p=0.020), and open surgical approach (p=0.010).

Examination of LOS by 20% intervals suggested a linear decrease in LOS with increasing FEV1 but not DLCO (Fig. 1b). In multivariable analysis, significant predictors of shorter LOS included higher FEV1 (p < 0.001), benign pathology (p=0.030), wedge resection (p < 0.001), and minimally-invasive approach (p < 0.001), but not DLCO.

Examination of mortality by 20% intervals suggested a relationship with both FEV1 and DLCO, but only at < 40% (Fig. 1c). There were 17 mortality events, limiting the number of variables for inclusion in the adjusted model. A model with FEV1 and approach created the best fit model, and mortality was significantly associated with lower FEV1 (p=0.020) and open surgical approach (p=0.029).

Conclusion: This study shows that in the modern era of primarily minimally-invasive surgery, FEV1 remains associated with postoperative morbidity and mortality, while DLCO is not. Despite increased risks, some clinical situations may merit resection in well-selected high-risk patients; these data provide updated risk-assessment and inform patient counseling given current surgical techniques.

Identify the source of the funding for this research project: No funding source.