The intrathoracic lymph node ratio seems to be a better prognostic factor than the level of lymph node involvement in lung metastasectomy of colorectal carcinoma

Abstract

OBJECTIVES Data on thoracic lymph node involvement (LNI) in lung metastasis of colorectal cancer (CRC) are conflicting, with a 5-year overall survival (OS) ranging from 6 to 40%. We aimed to evaluate whether there are subgroups of patients according to the lymph node ratio (LNR). METHODS We retrospectively reviewed the data from 106 patients who underwent a thoracic procedure for CRC lung metastasis with pathologically proven thoracic LNI. RESULTS In the univariate analysis, the median OS was significantly poorer for a pN2 location of LNI (26 vs 16 months, P = 0.04), LNR ≥ 50% (30 vs 17 months, P = 0.005), high preoperative CEA (32 vs 16 months, P = 0.02), hepatic metastases (27 vs 11 months, P <0.0001) and disease-free survival < 24 months (32 vs 17 months, P = 0.05). When pN1 and pN2 patients were staged according to the LNR, the median OS was significantly better for an LNR <50% (27 vs 17 months for pN1, 32 vs 12 months for pN2, P = 0.01). In the multivariate analysis, a high preoperative CEA [hazard ratio (HR): 2.256 (1.051-4.841), P = 0.04], pN1 status [HR: 0.337 (0.162-0.7), P = 0.004] and the absence of hepatic metastases [HR: 0.395 (0.180-0.687), P = 0.02] remained significant prognostic factors. There was an upward trend for patients with LNR <50% [HR: 0.565 (0.296-1.082), P = 0.08]. Otherwise, low LNR was significantly associated with a decreased risk of loco-regional recurrence (HR: 0.36, 95% confidence intervals: 0.14-0.96, P = 0.04). CONCLUSIONS The LNR seems to be a more reliable prognostic factor than LNI for CRC lung metastasis. Prospective studies are necessary.