A decision support framework for postoperative radiotherapy in patients with pathological N2 non-small cell lung cancer.

INTRODUCTION: Postoperative radiotherapy (PORT) plays a highly controversial role in pathological N2 (pN2) non-small cell lung cancer (NSCLC) disease. Recent studies reveal that not all patients can benefit from PORT. Further research is needed to identify predictors of PORT. METHODS: A total of 1044 pathologic stage T1-3N2M0 NSCLC patients were analyzed. Risk factors of distant metastasis were identified by the log-rank tests and the multivariable Cox models. We integrated risk factors of distant metastasis and our previously published loco-regional recurrence (LRR) related prognostic index into a decision support framework (DSF) to predict the outcomes of PORT. An independent cohort was used to validate the DSF. RESULTS: We defined patients with more than two of three identified LRR-related features (heavy cigarette smoking history, clinical N2 status, and more than four positive lymph nodes) as a high LRR risk group. We found the high-intermediate-risk histological type (with micropapillary and/or solid components) was associated with a higher risk of distant metastasis (HR = 1.207, 95 % CI 1.062 to 1.371, P =  0.0038), but not LRR. We built the DSF by combining these two types of features. Patients were stratified into four groups by using the DSF. PORT significantly improved OS only in the subgroup without high-risk histological features (without micropapillary or solid components) and with a high risk for LRR (three-year OS: 66.7 % in the PORT group vs 50.2 % in the non-PORT group; P = 0.023). CONCLUSIONS: A particular pN2 subgroup with a high risk of LRR and without micropapillary or solid components could benefit from PORT.