Rate and predictors of unforeseen PN1/PN2-disease in surgically treated cN0 NSCLC-patients with primary tumor > 3 cm: nationwide results from Italian VATS-group database

Background. Since no robust data are available on the real rate of unforeseen N1-N2 disease (uN) and the relative predictive factors in clinical-N0 NSCLC with peripheral tumours > 3 cm, the usefulness of performing a (mini)invasive mediastinal staging in this setting is debated. Herein, we investigated these issues in a nationwide database. Methods. From 01/2014 to 06/2020, 15,784 thoracoscopic major lung resections were prospectively recorded in the “Italian VATS-Group” database. Among them, 1982 clinical-N0 peripheral solid-type NSCLC > 3 cm were identified, and information was retrospectively reviewed. A mean comparison of more than two groups was made by ANOVA (Bonferroni correction for multiple comparisons), while associations between the categorical variables were estimated with a Chi-square test. The multivariate logistic regression model and Kaplan–Meyer method were used to identify the independent predictors of nodal upstaging and survival results, respectively. Results. At pathological staging, 229 patients had N1-involvement (11.6%), and 169 had uN2 disease (8.5%). Independent predictors of uN1 were SUVmax (OR: 1.98; CI 95: 1.44–2.73, p = 0.0001) and tumour-size (OR: 1.52; CI: 1.11–2.10, p = 0.01), while independent predictors of uN2 were age (OR: 0.98; CI 95: 0.96–0.99, p = 0.039), histology (OR: 0.48; CI 95: 0.30–0.78, p = 0.003), SUVmax (OR: 2.07; CI 95: 1.15–3.72, p = 0.015), and the number of resected lymph nodes (OR: 1.03; CI 95: 1.01–1.05, p = 0.002). Conclusions. The unforeseen N1-N2 disease in cN0/NSCLCs > 3 cm undergoing VATS resection is observable in between 12 and 8% of all cases. We have identified predictors that could guide physicians in selecting the best candidate for (mini)invasive mediastinal staging.