We reviewed our experience in the treatment of 299 patients with metachronous colorectal pulmonary metastases and investigated the factors affecting survival.

A retrospective analysis of the records of 299 patients with metachronous pulmonary metastases from CRC was performed. All patients had undergone treatments between January 1984 and Fabruary 2011 at a single cancer center.

The univariate and multivariate analyses demonstrated the factors significantly associated with disease free interval (DFI) were pathologic stage, age, chemotherapy and R classification of primary cancer (all P<0.05). Additionally, lung metastasectomy and DFI were predictors for overall survival (OS). Kaplan-Meier survival curves showed significantly survival difference among the subgroups in accordance with these independent factors. To construct a suitable model for assessing both categorical and continuous variables, multivariate logistic regression was used to create a scoring method. The score value for every CRC patient could be calculated using corresponding clinical and pathologic parameters. The score accurately predicted the risk of recurrence (the AUC is 0.732). A novel prognostic contour map was generated using score and age for the risk assessment of pulmonary metastases.

Old age, advanced stage and positive resection margin characterize a novel poor prognostic group in patients with CRC. The risk-stratification scheme assessed could identify patients who are likely to suffer pulmonary metastases and facilitate treatment. The prognostic contour map is appropriate for estimation of individualised outcomes.