Elsevier

European Journal of Cancer

Volume 168, June 2022, Pages 80-90
European Journal of Cancer

Original Research
The European Neuroendocrine Tumour Society registry, a tool to assess the prognosis of neuroendocrine neoplasms

https://doi.org/10.1016/j.ejca.2022.03.007Get rights and content

Highlights

  • The European Neuroendocrine Tumour Society registry is a multinational database, under the umbrella of European Neuroendocrine Tumour Society.

  • We analysed 10,102 patients, providing data on grade, stage, treatment and survival.

  • Survival is influenced by clinical stage 4, pathological grade 3 and organ of origin.

  • A 55% Ki67 cut-off distinguishes 2 groups of patients with G3 and with different prognosis.

  • Concerning patients with G2, 3 Ki67 cut-offs lead to distinctive overall survival.

Abstract

Background

Neuroendocrine neoplasms (NENs) are rare tumours with variable clinical behaviour. Their natural history is ideally best approached in large, multicentre and multinational registries with long-term patients’ follow-up. The European Neuroendocrine Tumour Society registry aims to obtain information regarding NEN outcomes and prognostic factors in a European frame.

Patients and methods

We collected data from 7 national NEN registries (Belgium, Czech Republic, Germany, Greece, Poland, Spain, Switzerland), representing 10,102 patients. Anonymised/pseudonymised data were collected in a secured server. Descriptive statistical methods were applied, as well as Kaplan–Meier survival curves and multivariable analyses for prognostic factors of overall survival (OS).

Results

median age of the study population was 60 years (range: 18–102), 48% were female. Common primary tumour sites were pancreas (27%) and small intestine (21%). Stage 4 disease was found in 47% of patients, while 26/10/16% had stage 1/2/3 disease, respectively. Grading (n = 6952) was G1/2/3 in 48/37/15% of the patients, respectively. Surgery was the main treatment, provided to 71% of patients, followed by somatostatin analogues (32%), chemotherapy (20%), Peptide receptor Radionuclide Therapy (PRRT) (9%) and targeted therapies (8%). OS at 5 years was 74%, influenced by grade, stage and tissue of origin in multivariate analysis. A Ki67 cut-off value set at 55% within the G3 group allowed to separate 2 groups with a meaningful different OS.

Conclusion

We report the first analysis of the European Neuroendocrine Tumour Society registry, comprising 10,102 patients with NEN from 7 European countries. This large cohort study describes prognostic factors for the survival of NENs throughout Europe, including primary tumour site, grade, stage and treatment.

Introduction

Neuroendocrine neoplasms (NENs) are a heterogeneous group of tumours increasing incidence and concern [1]. Their widespread anatomic location and specific endocrinological clinical features require the close collaboration of many different specialists for adequate diagnosis and clinical management. Although traditionally considered rare indolent tumours, their biological behaviour varies widely depending on primary tumour site, histological differentiation and proliferation rate, the ability to secrete different peptides/amines and extent of disease. Although the majority are well-differentiated, a subset is poorly-differentiated [2].

Given the low incidence and relatively better prognosis of NENs than their exocrine counterparts, their epidemiology is best studied in large registries with long-term follow-up. To date, the largest source of population-based epidemiological information on NENs is the National Cancer Institute's Surveillance, Epidemiology and End Results (SEER) Program, which covers approximately 30% of the US population and involves 64,971 cases of NENs. A recent analysis published in 2017 [1] showed a 6.4-fold increase in age-adjusted NEN incidence from 1973 (1.09/100,000) to 2012 (6.98/100,000) across all sites, stages and grades. Overall survival (OS) of all NENs has significantly improved between 2000–2004 and 2009–2012 (hazard ratio [HR], 0.79), including patients with metastatic gastroentero-pancreatic NENs. This improved prognosis may be partial due to increased access to better diagnostic tools, leading to earlier detection and stage migration and more effective administered therapy.

In Europe, the largest published database included 20,994 NENs, diagnosed from 1978 to 2002, provided by the project ‘Surveillance of rare cancers in Europe (RARECARE)’, which included 76 population-based cancer registries from 18 countries, 9 of which covering the entire national population [3]. The overall incidence rate for NENs was 2.5/100,000 inhabitants per year; 5-year OS was 50% for well-differentiated NETs and 12% for poorly-differentiated neuroendocrine carcinomas. Survival by primary tumour site was consistent with previously published data, although relevant prognostic factors such as tumour stage, grade or proliferative index were lacking, reflecting suboptimal pathological assessment per current standards. In addition, as more precise diagnostic procedures and effective treatments have evolved since 2002, these figures likely do not properly reflect the current status of NENs in Europe.

The European Neuroendocrine Tumour Society (ENETS) was founded in 2004 to improve the care, research and medical education in NENs [4]. ENETS is a multidisciplinary scientific society that encompasses a wide variety of NEN experts [5]. In this context, the ENETS Registry was created in 2007 to collect information regarding NEN outcomes and prognostic indicators in Europe. This is an institution-based registry, including ENETS Centres of Excellence and several NEN national networks, and aims to provide information on current standards of pathological classification and staging, diagnostic procedures, therapeutic strategies and patients’ outcomes.

Here we present the first analysis of the registry, comprising 10,102 patients with lung and gastroentero-pancreatic–NENs from 7 countries, consisting of the largest European study to date reporting the survival of NENs stratified by major prognostic factors. It provides some insights regarding regional disparities and may help understand differences in epidemiology, clinical presentation and survival of NENs across different European countries.

The ENETS registry collects patient data from several European countries since 07/2015 according to national and institutional data protection rules and ethical requirements. The Oracle database was programmed, hosted and maintained by the company Lohmann&Birkner (Berlin, Germany). Data were provided to the database either by direct pseudonymised data entry (Greece) or pseudonymised (Belgium, Czech Republic, Poland) or anonymised (Germany, Spain, Switzerland) data transfer. Quality data have been priorly provided to the scientific community by several national registries [[6], [7], [8], [9], [10], [11], [12], [13], [14], [15], [16]].

A thorough data checking and cleaning process of the initial data set (12,327 patients) was performed to identify the missing or inconsistent variables. A series of logic tests as well as monovariate and bivariate analysis were performed, and inconsistent data were set to missing; data imputation was not applied. Data elimination was performed for cases with missing outcome variables (any of the time or event variables); extensive bias analysis was carried out, which did not reveal any significant distortion in the remaining data caused by excluded cases. The final data set included 10,102 patients.

Cancers of unknown primary (CuP, n = 1244) were excluded from bivariate and multivariate analysis because of significant inconsistencies in the definition of this entity and differences in diagnostic procedures between recruiting institutions.

As descriptive statistics mean, median and standard deviation were used for continuous variables and simple proportions were used in bivariate analysis. At initial data assessment, bias analysis was performed, and in preparation for the multivariate analysis, the appropriate statistical tests (including t-test, F-Test and Chi-square tests) were applied.

Kaplan–Meier statistics were applied to estimate survival function and rates, with t being time from diagnosis to either death due to any cause or last contact. To compare two survival functions either log-rank or Tarone–Ware tests were used.

To assess the impact of different risk factors on OS, a multivariate Cox regression model was built using step-forward method finally including age, gender, WHO-grade (per Ki-67 proliferation rate), tumour stage (by ENETS or UICC depending on national specifications) and primary tumour site.

All statistical analyses were performed in IBM SPSS Statistics v.23 (Chicago, Ill., USA). Statistical significance was assumed for p-values <0.05.

Section snippets

Study population

The main characteristics of our study population (10,102 patients) are summarised in Table 1. The median age at diagnosis was 60 years (range: 10–102), 48% were female, and most common primary tumour sites were pancreas (n = 2722, 27%) and small intestine (Si) (n = 2132, 21%). Carcinoid syndrome was the most commonly reported functional syndrome (2184/2716, 80%).

Clinical stage

Overall, 47% of patients had metastases at diagnosis, 49% Pan-NEN and 59% Si-NEN (Fig. 1). Among less common NENs, the tumour sites

Discussion

This is the first European comprehensive analysis from a disease-specific registry of NEN, presenting data from 7 countries. To our knowledge, this is the largest comprehensive European NEN patient data set with detailed information on grade, stage and treatment. It provides relevant information on survival stratified by major prognostic factors. Overall 5-and 10-year survival of the whole cohort was 74.5 and 61%, respectively.

Pancreatic (27%) and Si (21%) NEN were the most common primary

Conclusion

We report for the first time results of a large European NEN database, the ENETS registry, and provide relevant data to improve the prognostic stratification of patients with NEN that shall help in clinical decisions, adequate patient selection and stratification for clinical trials. Further analyses on an increasing number of patients are planned, including broader coverage of European nations, and more in-depth subanalysis of specific patient populations.

Author contributions

IB has been involved in conceptualisation, data curation, formal analysis, funding acquisition, investigation, methodology, project administration, resources, supervision, validation, visualisation, writing the original draft, review & editing.

RGC has been involved in investigation, resources, data curation, formal analysis, validation, visualisation, writing the original draft, review & editing.

DB has been involved in data curation, Formal analysis, Methodology, Software, writing the original

Role of the funder

The work was supported by unrestricted grants from Ipsen, Novartis and Pfizer, specifically dedicated to the design, building, implementation and use of the ENETS registry, in good collaboration with the company Lohmann&Birkner. Authors received no personal funding.

Conflict of interest statement

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this article.

Acknowledgements

The authors wish to acknowledge all the physicians that participated in the collection and quality of the data; Ursula Plöckinger and Cheryl Berg, respectively, initiator and first data manager of the registry; members of the Registry working group (Barbro Eriksson, Eric Baudin, Pierre Goudet, Eric Van Cutsem); the ENETS executive committees and their successive chairs (Bertram Wiedenmann, Guido Rindi, Kjell Öberg, Wouter De Herder, Philippe Ruszniewski, Martyn Caplin, Massimo Falconi, Dermot

References (19)

There are more references available in the full text version of this article.

Cited by (10)

View all citing articles on Scopus
View full text