Surgery for NSCLC stages T1-3N2M0 having preoperative pathologically verified N2 involvement: A prospective randomized multinational phase III trial by the Nordic Thoracic Oncology Group.

Background: Surgery is not generally considered standard of care in preoperative pathological verified spread to N2 mediastinal lymph nodes in NSCLC. Methods: Previously untreated histologically verified NSCLC stages T1-3N2M0 were randomized to reg. A (Paclitaxel 225 mg/m2 + Carboplatin AUC6 day 1 q 3 wks for 3 courses, followed by surgery with ipsilateral mediastinal lymph node sampling followed by radiotherapy 2Gy x 30 fractions, 5F/W) or reg. B: same as A without surgery (sequential chemo-radiotherapy). 406 pts were needed to detect a 10% 5-year survival increase with 80% power and type 1-error of 5%. The study was approved by ethical committees. Pts gave informed consent. Results: 170 pts were randomized to A and 171 to B from 1998-2009 when study closed due to concomitant chemo-radiotherapy becoming standard instead of sequential treatment. Median age was 61 years (range 33-76 yrs), 59% were males, 43% had performance status 0. Stages T1N2M0, T2N2M0, and T3N2M0 occurred in 19%, 60%, and 21%, respectively. Adenocarcinoma (ADC) and squamous cell carcinoma occurred in 50% and 29%, respectively. In reg. A, surgery was possible in 132 out of 170 pts (78%), 121 pts (71%) had complete resection while 11 pts (6%) had incomplete resection. Pathological-surgical stage pT0 occurred in 4%. Median progression free survival (PFS), OS and 5-years survival rate were 10 mths, 17 mths, and 20% for A (+ surgery) compared to 8 mths (p=0.144), 15 mths (p=0.172), and 16% (p=0.310) for B, respectively. ADC pts had better OS in A than in B (HR 0.60; p=0.002), and 5-year survivals 20% and 7% (p=0.017) respectively. Stage T1N2 had better OS in A than in B (HR 0.47; p=0.010), 5-year survivals 36% and 17%. Conclusions: There were no statistical overall significant advantage for surgery in addition to chemo-radiotherapy (A) compared to chemo-radiotherapy alone (B) but ADC pts and pts with T1N2 had significantly improved OS and 5-year survival rates in the surgery arm. Current standard treatment for T1-3N2M0 NSCLC is concomitant chemo-radiotherapy which was not used in this study, hence conclusions should be further tested with use of such treatment as reference arm.