OBJECTIVETo explore the pattern of lymph node metastasis and evaluate the modes and extent of mediastinal lymph node dissection in patients with ≤ 3 cm, clinical stage I primary non-small cell lung cancer (NSCLC).

METHODSData of 270 eligible patients who underwent pulmonary resection with systematic lymph node dissection in our hospital between March 2012 and August 2013 were retrospectively analyzed in order to investigate the relationship between the clinicopathological features and lymph node metastatic patterns. Patients with multiple primary carcinomas or non-primary pulmonary malignancies and those who received any chemotherapy or radiotherapy or did not undergo systematic nodal dissection were excluded. The criteria of systematic nodal dissection included the removal of at least six lymph nodes from at least three mediastinal stations, one of which must be subcarinal. The data were analyzed and compared using Chi-square test.

RESULTSThe postoperative morbidity rate was 14.8% and no death occurred in this series. The imaging findings showed 34 cases of pure ground glass opacity lesions, 47 partial solid nodules, and 189 solid nodules. Apart from 34 p-GGO lesions, among the other 236 cases, ≤ 1 cm lesions were in 22 cases, 1 cm- ≤ 2 cm lesions in 138 cases, and >2 cm- ≤ 3 cm lesions in 76 cases based on radiologic findings. The pathological types included adenocarcinoma (n = 245), squamous cell carcinoma (n = 18) and other rare types (n = 7). The overall lymph node metastasis rate was 18.9% (51/270), and the incidence of lymph node involvement was 0(0/34) in cancers with p-GGO, 2.1% (1/47) in mixed solid nodules, 26.5% (50/189) in solid nodules, 18.2% (4/22) in nodules ≤ 1 cm, 14.5% (20/138) in 1 cm < nodules ≤ 2 cm, and 35.5% (27/76) in 2 cm < nodules ≤ 3 cm. The metastasis rates of non-specific tumor-draining region lymph nodes detected in the patients with positive and negative lobe-specific lymph node involvement were 20.0%-50.0% vs. 0-2.9% (P < 0.001).

CONCLUSIONSUsually NSCLC with p-GGO nodules has no lymph node metastasis, therefore, systematic nodal dissection may be not necessary. The larger the tumor size is, the higher the lymph node metastatic rate is for mixed or solid nodules. Intraoperative frozen-section examination of the lobe-specific lymph nodes should be performed routinely in patients with ≤ 2 cm stage I NSCLC, and systematic nodal dissection should be done if positive, but it may be not necessary if negative. However, the effectiveness of the systematic selective lymph node dissection still needs to be further confirmed.