Background: Retrospective studies have shown that primary tumor resection improves the prognosis of patients with colorectal cancer with unresectable metastasis (mCRC). Prognostic significance of lymph node dissection (LND) in mCRC has not been examined previously. The aim of this study was to investigate the prognostic impact of primary tumor resection and LND in mCRC.
Methods: A total of 1,982 patients with mCRC from January 1997 to December 2007 were retrospectively studied. The impact of primary tumor resection and LND on overall survival (OS) was analyzed using Cox proportional hazards model and propensity score analysis to mitigate the selection bias. Covariates in the models for propensity scores included treatment period, institution, age, sex, carcinoembryonic antigen, tumor location, histology, depth, lymph node metastasis, lymphovascular invasion, and number of metastatic organs.
Results: In a multivariate analysis, primary tumor resection and treatment in the latter period were associated with an improved OS, and age over 70 years, female sex, lymph node metastasis, and multiple organ metastasis were associated with a decreased OS. In the propensity-matched cohort, patients treated with primary tumor resection showed a significantly better OS than those without tumor resection (median OS 13.8 vs. 6.3 months; p = 0.0001). Furthermore, among patients treated with primary tumor resection, patients treated with D3 LND showed a significantly better OS than those with less extensive LND (median OS 17.2 vs. 13.7 months; p < 0.0001).
Conclusions: It was suggested that primary tumor resection with D3 LND improves the survival of patients with mCRC.