Hepatectomy for noncolorectal non-neuroendocrine metastatic cancer: a multi-institutional analysis.

BACKGROUND Although hepatic metastasectomy is well established for colorectal and neuroendocrine cancer, the approach to hepatic metastases from other sites is not well defined. We sought to examine the management of noncolorectal non-neuroendocrine liver metastases. STUDY DESIGN A retrospective review from 4 major liver centers identified patients who underwent liver resection for noncolorectal non-neuroendocrine metastases between 1990 and 2009. The Kaplan-Meier method was used to analyze survival, and Cox regression models were used to examine prognostic variables. RESULTS There were 420 patients available for analysis. Breast cancer (n = 115; 27%) was the most common primary malignancy, followed by sarcoma (n = 98; 23%), and genitourinary cancers (n = 92; 22%). Crude postoperative morbidity and mortality rates were 20% and 2%, respectively. Overall median survival was 49 months, and 1, 3, and 5-year Kaplan-Meier survival rates were 73%, 50%, and 31%. Survival was not significantly different between the various primary tumor types. Recurrent disease was found after hepatectomy in 66% of patients. In multivariable models, lymphovascular invasion (p = 0.05) and metastases ≥5 cm (p = 0.04) were independent predictors of poorer survival. Median survival was shorter for resections performed between 1990 and 1999 (n = 101, 32 months) when compared with resections between 2000 and 2009 (n = 319, 66 months; p = 0.003). CONCLUSIONS Hepatic metastasectomy for noncolorectal non-neuroendocrine cancers is safe and feasible in selected patients. Lymphovascular invasion and metastases ≥5 cm were found to be associated with poorer survival. Patients undergoing metastasectomy in more recent years appear to be surviving longer, however, the reasons for this are not conclusively determined.