More than 50% of GEP-NET patients have bilobar and synchronous liver metastases [15]. In these cases, an aggressive surgical approach is widely accepted. The indication for surgical resection of both the primary tumor and liver metastases depends on the site of the primary NET and the type of NELMs (Table 13.1).

Generally, synchronous or staged resections are performed if the primary NET and liver metastases are both amenable to potentially curative resection with acceptable morbidity and mortality rate. Kianmanesh et al. reported a series of 41 patients with synchronous NELMs treated by a two-stage approach [16]: the first step consisted in the resection of the primary NET (small bowel, pancreas tail or rectum), wedge resections of left-sided metastases (segments 1–4) and right portal vein ligation; in the second step a right hepatectomy was performed with a sufficient FLR due to achieved hypertrophy of the left liver. The 5-year overall survival and disease-free survival were 95% and 50%, respectively. These data were also confirmed by Gaujoux et al. who found that concomitant resection of the primary NET and liver metastases can be performed with low mortality (3%) and acceptable morbidity [17].

Another factor to consider for synchronous resection is the location of the primary NET: even though most authors found no significant differences in postoperative complications between pancreatic and digestive tract surgery, perioperative risk should always be taken into account before surgical planning, especially if the primary NET is located in the pancreatic head and a pancreatoduodenectomy is performed in combination with a major hepatectomy: in this case, patients should be carefully selected based on both operative risk and tumor biological behavior [17].

In cases of unresectable synchronous NELMs, because of the relatively prolonged life expectancy of patients with a well-differentiated slow-growing GEP-NET treated by non-surgical therapies, resection of the primary tumor could be performed to avoid tumor-related complications or hormone secretion symptoms. Some authors suggest resection of the primary NET with unresectable liver metastases even for non-functioning and asymptomatic tumors because it may facilitate liver-directed therapies during active follow-up, improving progression-free survival [18]. These findings have been confirmed in the UKINETS study that found resection of the primary tumor to be one of independent factors of prolonged survival for midgut NETs with unresectable liver metastases [19].

Neoadjuvant and adjuvant treatments have been introduced in single case reports and smaller series. Few authors reported a benefit of neoadjuvant treatment by immunochemotherapy or peptide receptor radionuclide therapy or both, resulting in enhanced tumor resectability. No difference in survival was found between patients treated by adjuvant chemotherapy with streptozocin and fluorouracil versus non-treated patients after liver resection or transplantation [20].

Thus, neoadjuvant and adjuvant chemotherapy are not currently recommended for treating resectable liver metastases from NETs.

Unresectable NELMs, unlike other metastatic malignancies, have been accepted as an indication for orthotopic liver transplantation (OLT). However clear evidence is lacking regarding the role of OLT in the treatment of unresectable NELMs because of the low incidence of the disease and the wide variety of alternative treatments without adequate available data comparing transplantation for unresectable liver metastases to other treatment modalities. Liver transplantation would ideally allow otherwise unfeasible removal of all metastases of the liver, but it is burdened with higher operative mortality and donor shortage. Moreover, the experience is scarce because liver transplantation for metastatic NETs represents 0.3% and 0.2% of overall liver transplants [21].

OLT for NELMs can offer a significant survival benefit when patients are selected properly. Mazzaferro et al. [22] emphasized the role of patient selection on post-transplant outcomes and proposed the following criteria (Table 13.2): young recipient (less than 55 years), low-grade tumor, Ki-67 index ≤10%, metastatic disease involvement no more than 50% of hepatic volume, primary tumor drained by the portal vein system (midgut and pancreatic NETs) and removed with all extrahepatic deposits (perihilar lymph nodes) before OLT, no metastatic spread to other organs, absence of right heart insufficiency. This approach led to 5-year survival close to 90% and these criteria are currently accepted by most transplantation centers with comparable results. The same Milan group compared two homogeneous groups of patients treated with transplant versus non-transplant techniques and found a significant long-term advantage of liver transplantation over other options (10-year overall survival, 88.8% vs 22.4%; 10-year progression-free survival, 13.1% vs 89%) [23]. These excellent results were not confirmed by other multicenter series (Table 13.3). Le Treut et al. recently reported the largest retrospective multicenter review of 213 patients undergoing OLT for NELMs and found a 5-year survival of 52%, with a 3-month mortality of 10%, and identified three poor prognostic factors affecting postoperative deaths: age more than 45 years, hepatomegaly and concomitant surgery in addition to OLT [24]. The same authors argued that if patients without those risk factors were selected, the 5-year survival rate would reach 80%. These survival rates were also confirmed in another recent review of the UNOS database with a 5-year survival rate of 49.2% in the pre-MELD era that increased to 57.8% after the application of the MELD score (since 2002) [25].

Other authors reported similar results even without applying a strict patient selection according to the aforementioned Milan criteria. Olausson et al. performed 15 transplants (10 OLT and 5 multivisceral) for metastatic NETs with expanded criteria: high proliferation rate (Ki-67 >10%), large tumor burden exceeding 50% of the hepatic volume and older age (>64 years) [26]. However, this experience is too limited for proper evaluation, and other non-surgical treatments should be considered.

Although the short- and long-term outcomes of liver transplantation for NELMs have improved during the last two decades, the recurrence rate after liver transplantation for metastatic NETs still remains high. A recent systematic review of the literature identified the following risk factors for recurrence after OLT: older age (>50 years), symptomatic tumor, primary tumor in the pancreas or a non-gastrointestinal location, non-carcinoid primary tumor, a high Ki-67 index, large liver involvement (>50%) and poor tumor differentiation (G3) [27]. However, in most of the reported series, liver transplantation was offered only after failure of other techniques, whereas it should be considered as a curative option. As for hepatocellular carcinoma, downstaging procedures while on the waiting list should be included in the general protocol of liver transplantation for NELMs [4].

The use of organs from the deceased donor pool for neuroendocrine metastatic disease presents ethical and cost-effectiveness problems. The absence of precise guidelines for liver transplantation for NELMs and the shortage of donors hamper access of these patients to this therapeutic option, as in hepatocellular carcinoma or other rare hepatic tumors. Living donor liver transplantation (LDLT) may be a solution to overcome this obstacle, providing a life-saving organ with acceptable mortality and morbidity. However, large experiences of living donation for NET are not available and current data are limited to very small case series.

Surgical resection represents the first-choice approach for the liver metastases from NETs with an overall survival of approximately 114 months as compared to 25 months for the unresectable non-treated lesions [33]. Chemotherapy presents a limited role for NETs with a modest rate of disease control ranging from 0–38% [34]. Preliminary data for novel therapeutic agents demonstrate a promising progression-free survival [35].