Factors associated with incomplete resection for large, locally invasive non-small cell lung cancer
Whitney S Brandt,Zhizhou Yang,Brendan T Heiden,Pamela P Samson,Daniel Morgensztern,Saiama N Waqar,Bryan F Meyers,Ruben G Nava,G Alexander Patterson,Benjamin D Kozower,Varun Puri
DOI: https://doi.org/10.21037/jtd-23-989
2024-05-31
Abstract:Background: Large, node-negative but locally invasive non-small cell lung cancer (NSCLC) is associated with increased perioperative risk but improved survival if a complete resection is obtained. Factors associated with positive margins in this population are not well-studied. Methods: We performed a retrospective cohort study using National Cancer Database (NCDB) for adult patients with >5 cm, clinically node-negative NSCLC with evidence of invasion of nearby structures [2006-2015]. Patients were classified as having major structure involvement (azygous vein, pulmonary artery/vein, vena cava, carina/trachea, esophagus, recurrent laryngeal/vagus nerve, heart, aorta, vertebrae) or chest wall invasion (rib pleura, chest wall, diaphragm). Our primary outcome was to evaluate factors associated with incomplete resection (microscopic: R1, macroscopic: R2). Kaplan-Meier analysis and cox multivariable regression models were used to evaluate overall survival (OS), 90-day mortality, and factors associated with positive margins. Results: Among 2,368 patients identified, the median follow-up was 33.8 months [interquartile range (IQR), 12.6-66.5 months]. Most patients were white (86.9%) with squamous cell histology (47.3%). Major structures were involved in 26.4% of patients and chest wall invasion was seen in 73.6%. Four hundred and seventy-eight patients (20.2%) had an incomplete resection. Multivariable analysis revealed that black race [hazard ratio (HR) 1.568, 95% confidence interval (CI): 1.109-2.218] and major structure involvement (HR 1.412, 95% CI: 1.091-1.827) was associated with increased risk of incomplete resection and surgery at an academic hospitals (HR 0.773, 95% CI: 0.607-0.984), adenocarcinoma histology (HR 0.672, 95% CI: 0.514-0.878), and neoadjuvant chemotherapy (HR 0.431, 95% CI: 0.316-0.587) were associated with decreased risk of incomplete resection. The 5-year OS was 43.7% in the entire cohort and 28.8% in patients with positive margins and 47.5% in patients with an R0 resection. Positive margin was also associated with a significantly higher 90-day mortality rate (9.9% versus 6.7%). Conclusions: For patients with large, node-negative NSCLC invading nearby structures, R0 resection portends better survival. Treatment at academic centers, adenocarcinoma histology, and receipt of neoadjuvant chemotherapy are associated with R0 resection in this high-risk cohort.