Abstract
BACKGROUND: Guidelines recommend primary surgical resection for non-small cell lung cancer (NSCLC) patients with clinical N1 disease and adjuvant chemotherapy if nodal disease is confirmed after resection. We tested the hypothesis that induction chemotherapy for clinical N1 (cN1) disease improves survival. METHODS: Patients treated with lobectomy or pneumonectomy for cT1-3 N1 M0 NSCLC from 2006 to 2011 in the National Cancer Data Base were stratified by treatment strategy: surgery first vs induction chemotherapy. Propensity scores were developed and matched with a 2:1 nearest neighbor algorithm. Survival analyses using Kaplan-Meier methods were performed on the unadjusted and propensity-matched cohorts. RESULTS: A total of 5,364 cN1 patients were identified for inclusion, of which 565 (10.5%) were treated with induction chemotherapy. Clinical nodal staging was accurate in 68.6% (n = 3,292) of patients treated with surgical resection first, whereas 16.3% (n = 780) were pN0 and 10.7% (n = 514) were pN2-3. Adjuvant chemotherapy was given to 60.9% of the surgery-first patients who were pN1-3 after resection. Before adjustment, patients treated with induction chemotherapy were younger, with lower comorbidity burden, were more likely to be treated at an academic center and to have private insurance (all p < 0.001), but were significantly more likely to have T3 tumors (28.7% vs 9.9%, p < 0.001) and to require pneumonectomy (23.5% vs 18.5%, p = 0.005). The unadjusted and propensity-matched analyses found no differences in short-term outcomes or survival between groups. CONCLUSIONS: Induction chemotherapy for cN1 NSCLC is not associated with improved survival. This finding supports the currently recommended treatment paradigm of surgery first for cN1 NSCLC.