Treatment and recurrence of loco-regional (Stage I-III) NET G3

Level: Level 2

Launch date: 1 March 2026

Task force chairs: ENETS high-grade task force

Principle investigators: Halfdan Sorbye, Bergen, Norway.

Project description:

Coordinating center:

Cancer Clinic, Haukeland University Hospital, Bergen, Norway

Type of research:

International multicentric retrospective real-world cohort

Rationale/ justification of the study:

Neuroendocrine tumors (NET) grade 3 is a relatively new entity, with a well-differentiated morphology, but a Ki-67 greater than 20%. It was included in the WHO classifications of 2017 for pancreas and 2020 for digestive NET. The evolution of NET G3 is incompletely understood, and many cases have a prior diagnosis of NET G1/2. Most of NET G3 are located in the pancreas (50-60%), but it can occur in other organs, such as small bowel, lung, stomach and rectum. NET G3 express somatostatin receptor 2 (SSTR2) in 70-85%, which is similar or less to that of NET G1/G2. Median Ki-67 is 30%, but still some cases have a Ki-67 >55%. The majority of NET G3 have metastases at diagnosis, up to 85% in the NORDIC NEC 2 study. The prognosis of patients with metastatic NET G3 is worse than for NET G1/G2, but better than for neuroendocrine carcinoma (NEC). Data on NET G3 are limited and data on patients with initially non-metastatic NET G3 are lacking. Non-metastatic NET G3 are too rare to perform a prospective study and the data quality in general registry studies may vary due to difficulties in the diagnostic work-up of NET G3 and their recent recognition in classifications. We therefore propose this study to generate new high-quality data through a dedicated retrospective collected cohort from expert NET centers.

Reference:

McNamara MG, et al. Controversies in NEN: An ENETS position statement on the treatment of patients with Grade 3 well-differentiated neuroendocrine tumours of the gastro-enteropancreatic tract. J Neuroendocrinol. 2025 Dec;37(12):e70080.

Objectives

Primary:

• Describe the clinical features and treatments in patients with stage I-III NET G3
• Describe the recurrence rate after surgery

Secondary:

• Explore the prognostic factors for recurrence
• Time to recurrence
• Use of adjuvant treatment
• Treatment of recurrence
• Compare the prognosis of NET G3 stage I-III with NEC stage I-III through matching with patients from the ENETS stage I-III digestive NEC cohort.
• Compare histological and NGS data from NET G3 stage I-III to NET G3 diagnosed with synchronous metastatic disease (data from NORDIC NEC 2 and ENETS SYNERGY NEN G3 study)
  
  

Population:

Inclusion criteria:

• Patients diagnosed from 2015 to 2023
• Digestive or other primary tumor site
• Initial stage I-III with radiological imaging demonstrating M0 disease.
• Any kind of treatment (resected or non-resected can both be included)
• NET G3 on the initial (i.e., at diagnosis) tumor sample (well-differentiated tumor with Ki-67 >20%).
• Diagnostic pathology must be done on the primary tumor

Exclusion criteria:

• NEC
• MiNEN
• Patients with secondary NET G3

Methods

Recruitment through invited expert NET centers (20-30), hopefully 50-60 cases.

Data will be collected by local investigators and filled into the ENETS database (level 2, extended version: 200 items)

• Clinical characteristics
• Date of diagnosis and method
• Location of primary tumor
• Pathology staging
• Tests used for staging
• Tumor functionality
• Immunohistochemical staining results if available (DAXX ATRX, MEN-1)
• NGS results if available
• Ki-67% value
• Treatments
• Recurrence after radical treatment (date, site(s), and new Ki-67 if new biopsy)
• Date of last follow up or death

 

Primary outcome:

Recurrence rate after treatment, survival and cancer specific survival

Morphological sub-study (optional- but very important if possible)

Histological slides (HE, Ki-67, CgA, SYN, and if avaiable Rb and P53) should be digitalized to be later uploaded to a platform for pathological re-assessment (so a kind of pathological second opinion request).

Statistical analysis plan:

Descriptive cohort.

Survival curves: Kaplan Meier.

Univariable/multivariable analysis of clinically relevant pre-defined characteristics and their influence on recurrence and OS (Cox model).

Research questions:

1. What treatment is given
2. Recurrence after curative treatment (usually surgery)
   1. How frequent are recurrences
   2. How frequent are distant vs local recurrences
3. Does optimal (PET) imaging before surgery better OS?
4. Use and possible benefit of (neo)adjuvant chemotherapy
5. Other preoperative treatments
6. Possible prognostic factors for recurrence
   1. Primary tumor site
   2. Stage
   3. Ki-67 level (< 30% vs >30%)
   4. Other pathological data (IHC, NGS etc)
   5. Number of reclassified cases after central pathology review
7. Follow-up
   1. When does recurrence occur? How long follow-up is needed.
   2. Location of recurrence
8. Comparison of locoregional NET G 3 vs locoregional NEC.
9. Possible histopathological/genetic differences between initial NET G3 stage I-III compared to NET G3 diagnosed with synchronous metastatic disease.

Publication policy:

The principal investigator will be first author.

Centers providing >5% of included patients with sufficient data quality will be co-authors.

Study timelines:

• December 2025-March 2026: ENETS database contracting with centers.
• January 2026: Study proposal submitted to ENETS database committee.
• March - December 2026: Inclusion of patients into the ENETS database.
• April 2026 – January 2027: Quality check of entered cases.
• Nov 2026: Abstract with preliminary data for ENETS 2027
• January- March 2027: Final data analyses and manuscript preparation
• June 2027: Manuscript submittance

Introduction video – REDCap data entry (Locoregional NET G3)

This short video provides step-by-step guidance on how to enter data in REDCap, including key study-specific requirements.

Watch the video: https://youtu.be/cbHw1mav55g