ABSTRACT
BACKGROUND: Lymph node (LN) involvement is a poor prognostic factor for malignant pleural mesothelioma (MPM). However, to our knowledge, postresection outcomes of node-negative (cN0/pN0), occult pathologic nodal disease (cN0/pN+), and clinical node-positive disease (cN+) have not been compared to date. PATIENTS AND METHODS: The National Cancer Data Base was queried for newly diagnosed, resected MPM with known clinical/pathologic LN information. Three cohorts were compared: cN0/pN0, cN+, and cN0/pN+. Multivariable logistic regression examined predictors of pathologic nodal upstaging. Kaplan-Meier analysis with propensity matching assessed overall survival (OS); multivariate Cox proportional hazards modeling examined predictors thereof. RESULTS: Of 1369 patients, 687 (50%) had cN0/pN0, 457 (33%) cN+, and 225 (16%) cN0/pN+ disease. Median follow-up was 29 months. In patients with cN0 disease, factors associated with pathologic nodal upstaging were younger age, greater number of examined LNs, and nonsarcomatoid histology (P < .05 for all). Relative to pN0 cases, occult LN involvement (65% being pN2) was associated with 51% higher hazard of mortality on multivariate analysis (P = .005). Following propensity matching, the OS of cN0/pN+ was similar to cN+ cases (P = .281). On multivariate analysis, the number of involved LNs (continuous variable, P = .013), but not nodal tumor, node, metastasis (TNM) classification or LN ratio (P > .05 for both), was associated with OS. CONCLUSION: Detecting occult nodal disease during resection for cN0 MPM is associated with poorer prognosis, with similar survival as cN+ cases, underscoring the importance of routine preoperative pathologic nodal assessment for potentially resectable MPM. The number of involved LNs (rather than current location-based classification) may provide more robust prognostic stratification for future TNM staging.