Neuroendocrine tumors (NET) arise from neuroendocrine cells and are a heterogeneous group of neoplasms [1-3]. NETs originate from a wide range of anatomic sites, which are mainly located in the gastroenteropancreatic system (60%) and the bronchopulmonary system (>25%) [4,5]. With an incidence of 5.25 per 100,000 each year in the United States, NETs are considered to be rare tumors [4].

Patients with NET liver metastases either complain of abdominal pain due to the mass effect of the tumor or excessive hormone production leading to the carcinoid syndrome, which consists of diarrhea, cutaneous flushing, various hemodynamic alterations, and wheezing [6,7]. Moreover, up to 75% of patients with NETs (including midgut or hindgut origin) present with liver metastasis at the time of diagnosis [8].

A 5-year survival rate of 22% for patients with NET liver metastases has been described [5]. Surgical approach shows a benefit concerning overall survival as compared with the nonsurgical approach; however, curative surgery is only applicable in 10% of the patients [9]. Therefore, noninvasive alternatives, such as local ablation techniques, percutaneous liver-directed techniques (chemoembolization, bland embolization, and selective internal radiation therapy), peptide receptor radionuclide technique, chemotherapy, targeted therapy, and biotherapy are of great importance and their value has to be determined.

Mechanistically, local ablative techniques such as cryotherapy, radiofrequency ablation (RFA), and percutaneous alcohol injection (PEI) rely on the cytotoxic effects of chemicals and induce nonphysiologic temperatures into the hepatic tissue [10,11]. Therefore, tumor location and extension in the liver influences the choice of treatment modality [12].

Treatment modalities involved in percutaneous liver-directed techniques, include bland embolization (BE), transarterial chemoembolization (TACE), and selective internal radiation therapy (SIRT). The principle of BE consists of inducing regional ischemia to the tumor. In TACE, chemotherapeutic substances are locally applied causing a cytotoxic effect. Due to the local embolization the intratumoral concentration of the cytostatic is as much as 20 times higher using TACE than in systemically applied chemotherapy [13]. In addition, systemic side effects can be minimized and cytotoxicity can be maximized [14]. SIRT uses intracorporal radiation through microspheres made of glass (Thera-Spheres) or of resins (SIR-Spheres). These microspheres are loaded with radioactive Yttrium-90 [2]. By virtue of their size, the microspheres obliterate the vessels and irradiate the tumor with a high radiation dose. The adjacent healthy tissue receives minimal dosage [2]. Unfortunately, these treatment modalities are difficult to compare due to the small number of patients and to heterogeneous inclusion criteria (eg, tumor staging, primary tumor location, etc). [15].

Peptide receptor radionuclide therapy (PRRT), a combination of a somatostatin analog with a radioligand, a beta-emitter, is used to detect and treat NETs expressing somatostatin receptors. After systemic injection, the radioligand is internalized into the cells and transported to the lysosomes. The effect on tumor cell proliferation is based on the radiotoxicity of the radionuclide in the deoxyribonucleic acid of the target cell [16]. However, radionuclides should be applied cautiously since side effects, such as bone marrow toxicity, hepatic insufficiency, myelodysplastic syndrome, renal insufficiency, or hematological toxicity might occur. Secondary malignancies such as leukemia are rare, but may also occur [17-19]. Functional imaging (ie, octreoscan or gallium 68 Positron Emission Tomography [PET]) is required to identify the subgroup of patients eligible for PRRT [20,21].

The role of systemic chemotherapy for NETs with liver metastases has been discussed vigorously. In a prospective study, Moertel et al [22] evaluated streptozotocin as a chemotherapeutic monotherapy and found a significant response; however, the benefit was strongly limited by the renal and hematologic toxicity of streptozotocin, and therefore is not an acceptable treatment option. Studies combining streptozotocin with other agents have been conducted with the aim to decrease the dosage of streptozotocin, and thus reduce its toxicity [23]. For the therapy of metastatic pancreatic NETs, Kouvaraki et al [24] reported that a combined multidrug chemotherapy with fluorouracil, doxorubicin, and streptozotocin showed an acceptable response rate of 39% with responders having both increased progression-free survival and overall survival. However, patients with metastatic midgut NETs treated with this multidrug chemotherapy regimen showed the same survival rates as interferon-based therapy concepts [25]. Since poorly differentiated (G3) gastrointestinal NETs behave like lung neuroendocrine carcinomas (small-cell carcinomas) a platin-based chemotherapy is discussed [12].

Targeted therapy includes multikinase inhibitors, mammalian target of rapamycin (mTOR) inhibitors, and monoclonal antibodies, which interact with various molecular pathways [26]. Sunitinib, an orally applied multikinase inhibitor, targets vascular endothelial growth factor (VEGF) receptors as well as platelet-derived growth factor receptors, which are often expressed in NETs [27,28]. Everolimus, an mTOR inhibitor, has also shown antitumor activity [29]. Bevacizumab, a monoclonal antibody against VEGF, inhibits angiogenesis in tumors and seems to reduce tumor perfusion [26,30].

Biotherapy, using interferon-α and somatostatin analogues such as octreotide and lanreotide, prevents the synthesis of the polypeptide hormones and biogenic amines produced by functional NETs. This provides relief from endocrine symptoms associated with the carcinoid syndrome in 80% of patients [31-34]. The Placebo-Controlled, Double-Blind, Prospective, Randomized Study on the Antiproliferative Efficacy of Octreotide LAR in the Control of Tumor Growth in Patients with Metastatic Neuroendocrine Midgut Tumors reports treatment with octreotide (long-acting release) essentially delays the period to tumor progression in patients with both functionally active and inactive metastatic midgut NETs compared with the placebo-treated group. However, survival analysis could not be performed due to a small number of observed deaths [35].

The aim of these four systematic reviews is to determine evidence for the noninvasive treatment options in terms of symptom relief and tumor control in patients with nonresectable liver metastases of NETs.

The following four systematic reviews dealing with the nonsurgical treatment options of neuroendocrine liver metastases attempt to address the following questions represented in Textbox 1.

Our research results will be reported in accordance with the standards of the Preferred Reporting Items for Systematic reviews and Meta-Analyses (PRISMA)[36].

The eligibility criteria for inclusion as well as for exclusion of studies are illustrated in Tables 1-4. Furthermore, the count and reason of exclusion will be revealed in a flow diagram, which will comply with the PRISMA Statement 2009 (Figure 1) [36]. The study types that will be included are randomized controlled trials (RCTs), prospective and retrospective comparative cohort studies, noncomparative cohort studies, case-control-studies, and case series.

These studies will provide the basis for the qualitative synthesis of this systematic review. Single-cohort studies will be collected in a separate database and will only be used for descriptive purposes. No publication date or language restrictions will apply.

The electronic search strategy to scan the databases and detect all relevant articles was developed by the librarians of the Medical Library Careum (University of Zurich, Switzerland). The search will be performed on the MEDLINE, EMBASE, and The Cochrane Library (Cochrane Database of Systematic Reviews, Database of Abstracts of Reviews of Effects, and Cochrane Central Register of Controlled Trials [CENTRAL]) databases. An endnote file, comprising all identified titles and, if accessible, the corresponding abstracts will be prepared for the investigators. Supplementary publications will be found by manual search or by reviewing reference lists. Two independent review group members will peruse titles and/or abstracts from studies, which were identified using the search profile. Afterwards, the entire text of these potentially eligible studies will be re-examined for eligibility. Any uncertainties will be discussed and resolved with a third member. A specially created Web-based, predetermined protocol will be used to extract data from the included studies for the study quality expertise and synthesis of medical findings.

The data extraction parameters, include manuscript title, name of journal, first author's name, publication year, total number of patients, number of patients in the chemotherapy/biotherapy/targeted therapy group, number of patients in the nontreatment group, name of used substances, age (mean, standard deviation, median), male to female ratio, progression-free survival, overall survival, quality of life (containing side effects), study design, and targeting objective 1-5. The Grading of Recommendations Assessment, Development and Evaluation (GRADE) will be used to grade the quality (level) of evidence and the strength of recommendations [37].

We will prepare a narrative synthesis of the findings from the included studies. A quantitative synthesis will be used if the included studies are sufficiently homogenous. We anticipate that there will be a limited scope for meta-analysis of a relatively large number of studies because of the range of outcomes measured across the small number of existing trials (such tumors are rare). Nevertheless, where studies have used the same type of intervention and comparator, with the same outcome measure, we will pool the results using a random-effects meta-analysis. We calculate a 95% CI and two-sided P values for each outcome.