The continuing controversy about surgery for non-colorectal non-neuroendocrine liver metastases (NCRNNE) necessitates identifying risk factors of worsened outcomes to improve patient selection and survival. Prospectively collected data of 167 patients undergoing hepatectomy for NCRNNE were analyzed, and a comparison to a matched population of colorectal liver metastases (CLM) was performed. Overall survival (OS) (35 vs. 54 months; P = 0.008) and recurrence-free survival (RFS) (15 vs. 29 months; P = 0.004) of NCRNNE patients were significantly shorter compared to those with CLM. The best survival was found in the genitourinary (GU; OS, 45 months; RFS, 21 months) NCRNNE subgroup, whereas survival for gastrointestinal (GI) metastases was low (OS, 8 months; RFS, 7 months). Patients with renal cell carcinoma (RCC) showed excellent outcomes when compared to CLM (OS, 50 vs. 51 months; P = 0.901). Extrahepatic disease (EHD) was identified as independent prognostic factor for reducing both RFS (P = 0.040) and OS (P = 0.046). The number of liver lesions (P = 0.024), residual tumor (P = 0.025), and major complications (P = 0.048) independently diminished OS. The degree of survival advantage by surgery is determined by the primary tumor site, EHD, the number of metastases, and residual tumor. Thus-even more than in CLM-these oncological selection criteria must prevail. GU metastases, especially RCC, represent a favorable subgroup.

No abstract