Surgical resection of neuroendocrine tumor liver metastases as part of multimodal treatment strategies: A propensity score matching analysis.

Background: It remains unclear whether liver resection as part of multimodal therapy of neuroendocrine liver metastases (NELM) is superior to non-surgical (interventional and medication-based) treatment alone. This study should determine if patients with NELM undergoing hepatic surgery in addition to non-surgical treatment have improved overall survival compared to patients undergoing non-surgical therapy alone.

Methods: 123 patients undergoing treatment of NELM between 1995 and 2014 were included in this retrospective cohort study. Two groups were formed: (A) surgery and non-surgical therapy and (B) non-surgical treatment alone. To minimize the bias of patient selection propensity score matching was used.

Results: There was significantly better overall survival for group A (152 months, 95%CI: 119-185) compared to group B (63 months, 95%CI: 45-81) measured from the initial diagnosis of the metastases (P = 0.003). After propensity score matching, 37 patients undergoing surgical resection of NELM within a multimodal treatment were compared to 37 patients undergoing non-surgical treatment. Under these circumstances, surgery had no significant influence on survival (group A: 134 months, 95% CI: 94-173; group B: 76 months, 95% CI: 53-99, P = 0.23). Based on a multivariate Cox proportional hazard model, only Ki-67 of primary tumor >20% (HR, 50.776; 95%CI, 4.056-635.71; P = 0.002) and no resection of primary tumor (HR, 10.464; 95%CI, 1.873-58.448; P = 0.007) remained independent risk factors.

Conclusion: After minimizing patient selection bias, patients with hepatic resection as integral of multimodal therapy of NELM do not have better overall survival than those receiving non-surgical treatment alone.