Impact of L4 lymph node dissection on long-term survival in left-side operable non-small-cell lung cancer: a propensity score matching study

OBJECTIVES We investigated the impact of level 4 (L4) lymph node dissection (LND) on overall survival (OS) in left-side resectable non-small-cell lung cancer (NSCLC), with the aim of guiding lymphadenectomy. METHODS A total of 1929 patients with left-side NSCLC who underwent R0 resection between 2001 and 2014 were included in the study. The patients were divided into a group with L4 LND (L4 LND+) and a group without L4 LND (L4 LND−). Propensity score matching was applied to minimize selection bias. The Kaplan–Meier method and Cox proportional hazards model were used to assess the impact of L4 LND on OS. RESULTS A total of 317 pairs were matched. Of the cohort of patients, 20.3% (391/1929) had L4 LND. Of these patients, 11.8% (46/391) presented with L4 lymph node metastasis. L4 lymph node metastasis was not associated with the primary tumour lobes (P = 0.61). Before propensity score matching, the 5-year OS was comparable between the L4 LND+ and L4 LND− groups (69.0% vs 65.2%, P = 0.091). However, after propensity score matching, the 5-year OS of the L4 LND+ group was much improved compared to that of the L4 LND− group (72.9% vs 62.3%, P = 0.002) and L4 LND was an independent factor favouring OS (hazard ratio 0.678, 95% confidence interval 0.513–0.897; P = 0.006). Subgroup analysis suggested that L4 LND was an independent factor favouring OS in left upper lobe tumours. CONCLUSIONS In patients with left-side operable NSCLC, L4 lymph node metastasis was not rare and L4 LND should be routinely performed.