Diagnosis and Prediction of Neuroendocrine Liver Metastases: A Protocol of Six Systematic Reviews

Background: Patients with hepatic metastases from neuroendocrine tumors (NETs) benefit from an early diagnosis, which is crucial for the optimal therapy and management. Diagnostic procedures include morphological and functional imaging, identification of biomarkers


Neuroendocrine Tumors
Neuroendocrine tumors (NETs) arise from the diffuse neuroendocrine system and therefore appear widespread over the whole body, especially in the gastrointestinal tract and the bronchopulmonary system [1,2]. NETs secreting hormones lead to a symptomatic disease. Nonsecreting NETs may occur initially asymptomatic or with delayed symptoms due to progressive increase in tumor mass [3,4]. Therefore, differences in functional behavior are the basis of a classification system categorizing functioning and nonfunctioning NETs [4]. Other reported classification systems are based on embryological origin or histopathological findings. In 2010, The World Health Organization (WHO) presented a new classification on the basis of tumor grading using histopathological criteria such as Ki67 index, mitotic count, and presence or absence of necrosis [5].
NETs is a relatively rare disease with an incidence of 1-3 per 100,000 [6,7]. The large range of reported incidence might be due to the fact that NETs often present initially asymptomatic and are often found accidentally or in autopsies [4]. Predominantly, NETs emerge sporadically (>90%) and are traditionally assigned to multiple endocrine neoplasia type 1 (MEN1), neurofibromatosis-type 1 (NF1), and Von-Hippel-Lindau syndrome [1,4]. The clinical picture of NETs spans over different effects of excessive hormone secretion such as hypergastrinemia in Zollinger-Ellison Syndrome (ZES) with hyperchlorhydria, hyperinsulinemia in insulinoma, flushing and diarrhoea in the serotoninergic carcinoid syndrome. In the case of nonsecreting NETs, symptoms present due to the adverse effects of the growing primary tumor or metastases [8].

Biochemical Markers
Hormones secreted from NETs can be used as specific markers for NETs. Moreover, NETs express, store, and secrete characteristic neuronal proteins such as acid glycoprotein chromogranin A (a component of the membrane neurosecretory granula), neuron-specific-enolase (NSE), and synaptophysin [3,9]. These proteins derived from neuronal structures could serve as markers and are even positive in nonfunctioning NETs [1,3]. Since more than one half of NETs are nonsecreting, these proteins play a crucial role [4]. Assessment of different biochemical markers depends on various parameters, such as threshold cut-off level, detecting method of urine, serum or plasma as well as location of the primary tumor or metastases and extension of the disease. Due to the large variety and number of evaluation parameters, it is difficult to compare the studies [10,11].

Histopathological Prognostic Markers
Ki67 is a monoclonal antibody, which was introduced in 1984 by Gerdes et al [11]. It detects a growth rate depending on the nuclear antigen Ki67 which is only expressed during active cell cycle phases (S, G2, and M-phase). Ki67 is completely absent during the resting phase G0. Therefore, cell proliferation is assessed by the immunohistologic presence of Ki67 positive cells per area in stained tissue blocks [11].
For various human neoplasms such as breast, lung, and solid cancers, Ki67 proliferation index has been successfully established as a predictive marker [12,13]. The higher the cell proliferation, the greater is the probability for metastases resulting in decreased patient survival. The primary location of NETs metastases is the liver [14][15][16][17]. The occurrence of hepatic metastases is a prognostic factor which strongly influences the survival of patients suffering from NET [18][19][20].

Genetic Signatures and the Presence of Circulating Tumor Cells
To stratify outcomes in patients undergoing resection of primary NET, a simple scoring system using tumor size, histological grade, nodal metastases, and resection margin status has been introduced [21]. Nevertheless, current classification systems for NETs other than positron emission tomography (PET) fail to predict the clinical course and the response to treatment [22]. The discrepancy might be explained either by an insufficient accuracy of these classification systems or an adaptive NET behavior [23]. These limitations of the pathologic classifications have led to the investigation of other predictive parameters based on genetic signatures as well as the presence of circulating tumor cells [24,25]. These novel predictive parameters have to be included in the classification systems in order to account for the biological behavior, the likelihood for developing metastases as well as the choice of treatment [25].

Imaging Methods
Imaging methods are used to diagnose neuroendocrine tumors (NETs) and their metastases [26]. Beside conventional morphologic imaging methods, functional imaging modalities have been introduced in order to improve accuracy in detecting NETs and liver metastases [27]. Functional imaging methods have their limitations with a great impact on a possible therapeutic strategy, where differentiation between pancreatic foci and neighbouring lymph nodes as well as exact demarcation of a suspicious focus to a liver segment is crucial [28]. Advanced techniques such as contrast-enhanced ultrasound may assist in earlier detection of hepatic metastases, and could therefore offer a wider therapeutic range either surgically, with radiofrequency thermal ablation, or with systemic chemotherapy [29].

Liver Biopsy
The most common site of neuroendocrine tumor (NET) metastases is the liver [30]. The presence of hepatic metastases is a strong prognostic factor for the survival of patients with NETs, regardless of the primary tumor site [31]. Histologic examination is the most sensitive diagnostic method and forms the basis for treatment decisions [32]. However, the value of the biopsy for treatment decision making involving primary NETs and their liver metastases is not well defined [33,34].

Objective
The aim of these six systematic reviews is to assess the predictive value of Ki67 index and other biomarkers, to compare the diagnostic accuracy of morphological and functional imaging, and to define the role of biopsy in the diagnosis and prediction of neuroendocrine tumor liver metastases.

Systematic Reviews
Our reviews were prospectively registered at the International Prospective Register of Systematic Reviews (PROSPERO) with the following IDs: CRD42012002644, CRD42012002647, CRD42012002648, CRD42012002649, CRD42012002650, CRD42012002651.
The above six systematic reviews dealing with the diagnosis and prediction of neuroendocrine liver metastases attempt to address the following questions in Table 1. Table 1. Scientific questions on diagnosis and prediction of neuroendocrine liver metastases.

Should patients with low Ki67 index be followed up for the detection of liver metastases?
In patients with a primary NET, what is the predictive value of Ki67 index, mitotic count, or tumor grading, obtained from the primary tumor, in predicting the development of liver metastases?

Should genetic signatures and the presence of circulating tumor cells be used in the prediction of liver metastases and to inform treatment decisions?
In patients with a primary NET, what is the predictive value of genetic signatures obtained from the primary tumor, in predicting the development of liver metastases?
In patients with a primary NET, what is the predictive value of circulating tumor cells obtained from the primary tumor, in predicting the development of liver metastases?
In patients with a primary NET, should genetic signatures be used in the treatment decision (surgery, locally ablative techniques, liverdirected techniques, peptide receptor radionuclide treatment, chemotherapy, targeted therapy, and biotherapy)?
In patients with a primary NET, should the presence of circulating tumor cells be used in the treatment decision (surgery, locally ablative techniques, liver-directed techniques, peptide receptor radionuclide treatment, chemotherapy, targeted therapy, and biotherapy)?

Which biochemical markers should be used for detection and post treatment follow-up of liver metastases?
In patients with a primary NET, what is the diagnostic accuracy of the available biochemical markers (eg, chromogranin A and B, Serotonin, neuron-specific-enolase (NSE), tumor specific hormones) in detecting liver metastases?
In patients receiving a liver resection, what is the diagnostic accuracy of the available biochemical markers (eg, chromogranin A and B, serotonin, NSE, tumor specific hormones) obtained during follow-up, in detecting recurrent disease or disease progression?

Which morphological imaging modality should be used to assess resectability of liver metastases with a curative intent?
In patients with NET liver metastases, what is the diagnostic accuracy of different morphological imaging modalities (US, CT, MRI) in identifying liver lesions and extrahepatic disease?
In patients with NET liver metastases, what is the diagnostic accuracy of different morphological imaging modalities (US, CT, 3D-CT, MRI) in detecting vascular and biliary invasion, in order to assess resectability (R0/R1)?

Which functional imaging modality should be used to assess resectability of liver metastases with a curative intent?
In

Do we need a biopsy of both the primary and liver metastases for the treatment decision of liver metastases?
In patients with a primary NET and synchronous liver metastases, what is the agreement between the biopsy of the primary and the liver metastases with regards to tumor grading?
In patients with metachronous liver metastases, what is the agreement between the biopsy of the primary and the liver metastases with regards to tumor grading?
In patients with liver metastases, what is the agreement between single vs multiple liver biopsies with regards to tumor grading?
In patients with NET liver metastases, do we need additional biopsies from normal parenchyma to detect micrometastases?
The systematic review inclusion and exclusion criteria are listed in Tables 2-7. There were no restrictions in the literature search regarding the publication language or by publication date. The following study types were included: randomized controlled trials (RCTs), prospective and retrospective comparative cohort and case-control studies and case series (Figure 1).   Patients with primary neuroendocrine tumors and patients who underwent surgery for primary liver tumors with a curative intent and were followed up for the detection of potential liver metastases

Test of interest
The different biochemical markers Chromogranin A and B, Serotonin and tumor specific hormones will be compared

Reference standard
The histopathological diagnosis of the resected specimen or a tumor biopsy will be considered as the reference standard

Search
Librarians of the Medical Library Careum, University of Zurich, Switzerland, develop the electronic search strategy to query databases and to identify all potentially relevant articles. The following databases will be searched: MEDLINE, EMBASE and The Cochrane Library (Cochrane Database of Systematic Reviews, Cochrane Central Register of Controlled Trials (CENTRAL), Database of Abstracts of Reviews of Effects). The investigators will be provided with an Endnote file containing all identified titles and, if available, the corresponding abstracts. Additional articles will be retrieved through manual search or scanning of reference lists. Titles and/or abstracts of all identified records will be independently screened by two review team members to ascertain their relevance and to identify studies that potentially meet the inclusion criteria outlined in Tables 2-5. The full text of each of these potentially relevant studies will then be assessed for eligibility. Any disagreement will be resolved through discussion with a third review team member. A predefined protocol will be used to extract data from the included studies for assessment of study quality and evidence synthesis.

Data Extraction
The parameters for data extraction will be the following: first author's name, publication year, answering scientific questions, study design, total number of patients, number of patients in the study group, and number of patients in the comparison group. The Grading of Recommendations Assessment, Development and Evaluation (GRADE) will be used to grade the quality (level) of evidence and the strength of recommendations [35].
A narrative synthesis of the findings from studies included will be provided. A quantitative synthesis will be used for studies that are sufficiently homogenous from a clinical (comparability of populations, interventions and outcomes) and from a statistical perspective (heterogeneity, eg, I2<50%). We anticipate that there will be a limited scope for meta-analysis despite a relatively large number of studies due to the different outcome measurements of the existing trials (ie, since such tumors are rare). However, results from studies using the same type of intervention and comparator, with the same outcome and measurements will be pooled using a random-effects meta-analysis. In addition risk ratios for binary outcomes, 95% confidence intervals and two-sided P values will be calculated for each outcome.

Discussion
There are several modalities for the diagnosis and prediction of neuroendocrine liver metastases; however, there is a lack of consensual data on the subject. The six systematic reviews described in this protocol will elucidate the role and compare histopathological prognostic and biochemical markers, biopsies of the primary neuroendocrine tumor and NET liver metastases, morphological and functional imaging modalities. They will help to define clinical guidelines.