Multicenter study on the outcomes of patients with NET and high-grade transformation

GradeTrans-NET

Project description

Coordinating center:

Beaujon Hospital, APHP Université Paris-Cité, Clichy, France.

Steering Committee members:

• Rachel Riechelmann (Sao Paulo, Brazil),
• Halfdan Sorbye (Bergen, Norway)
• Joakim Crona (Uppsala, Sweden)
  
  

Ethical Approval:

ENETS DB Core approval and local IRB approval as per local ethical regulation.

Type of research:

Retrospective study

Background:

High grade tumor transformation (HGT) has been described in patients with grade 1 or grade 2 neuroendocrine tumors (NET) which become G3, or low G3 which substantially increase Ki-67 index. HGT is mostly observed in pancreatic NET following multiple treatment lines, especially with alkylating agents (ALK) and/or radioligand therapy (RLT), being reported in 30-60% of pancreatic NET1–3. HGT is generally associated with very aggressive disease evolution, sometimes with neuroendocrine carcinoma-like clinical and histological features, rapid tumor growth, extensive progression, intrinsic therapy resistance and dismal prognosis 1,2. The biological reasons for such aggressive evolution are unknown; yet, because a subset of tumors acquire high tumor mutation burden (TMB-h), accumulation of unrepaired neomutations induced by DNA-damaging treatments may be one of the triggering mechanisms of HGT4,5.

For TMB-h pancreatic NETs, a recent multicenter retrospective series demonstrated that immune checkpoint inhibitors could benefit to selected patients with acquired mismatch repair deficiency (MMRd), which may nevertheless account for a limited proportion of patients only6. However, several other aspects of HGT remained currently unexplored. For example, the treatment outcomes of patients without TMB-h have not been reported, precluding to identify the optimal strategy in these patients. In patients with TMB-h, the efficacy of treatments given alternatively or beyond immunotherapy has not been reported either.

Still, these unmet needs appear to be of great importance given the strong prognostic impact of HGT, and because the prevalence of HGT will likely increase in the upcoming years because of increased number of treatment modalities (e.g., new RLTs) leading to prolonged survival and higher impact on tumor biology and heterogeneity. Therefore, the GradeTrans-NET study will aim to generate high-quality data on the characteristics and clinical outcomes of patients with NET and HGT, collected retrospectively from expert NET centers.

Aim

This is a retrospective, multicenter cohort study aiming to enroll patients with advanced or metastatic NET, whose disease underwent histologically confirmed high-grade transformation (HGT), defined as a G3 NET in a previous G1/G2 NET, or an increase of Ki-67 ≥ 20% in a previous G3 NET. The study period covers patients diagnosed and treated from January 1, 2017, to January 1, 2027. From eligible patients, medical charts will be evaluated to collect relevant clinical data.

Primary objective:

describe overall survival from HGT onset

Secondary objectives:

• Clinical characteristics of patients with HGT
• Progression-free survival to therapies used at HGT onset
• Time to treatment failure to therapies used at HGT onset
• Objective response rate to therapies used at HGT onset
• Factors associated with overall survival following HGT onset

Exploratory analyses may be performed, depending on the number of patients, to investigate the prognostic effect of certain subgroups of molecular or IHC markers.

Study Endpoints

Primary endpoint: overall survival from HGT, defined as the date interval between date of “HGT sample” and the date of death – or last follow up (censure).

Secondary endpoints:

• Progression-free survival: time interval between the date of first dose of a treatment and date or radiological progression or death
• Time to treatment failure: time interval between the date of first dose of a treatment and the date of its discontinuation for any reason, such as radiological disease progression, adverse events, patient withdrawal, start of new therapy, or death
• Objective response rate: as per local evaluation, based RECIST 1.1
  
  

Statistical Analyses

Descriptive statistics will be used to summarize the population characteristics at baseline and at HGT onset. Comparisons between groups will be performed according to type of outcome variable (Chi2 for binary variables and Mann-Whitney for continuous variables).

Overall survival, progression free survival and time to treatment failure will be estimated by the Kaplan-Meier method. The follow up period will be estimated the Kaplan-Meier reverse method. The overall survival from HGT onset will be compared between different therapies (e.g., with or without immunotherapy) using the log rank test.

Prognostic factors for overall survival at HGT will be investigated by a Cox proportional hazard multivariable analysis, considering the following independent variables: age, Ki-67 index on the “HGT sample”, TMB, treatments prior HGT (e.g., ALK, RLT), morphological de-differentiation, MMRd on NGS. Non-collinear, clinically relevant variables resulting in p<0.2 in univariable analyses will be tested in the multivariable model. Hazard ratios and respective 95% Confidence intervals will be reported.

All results of inferential tests will be considered significant when a two-sided p < 0.05 is observed.

A planned sample size of 200 patients was determined based on convenience and feasibility to investigate the study endpoints.

Study Population

Inclusion criteria:

• Histologically confirmed metastatic neuroendocrine tumor (NET) of gastro-enteropancreatic (GEP) origin, thoracic (bronchial/thymic), or unknown primary origin.
• Availability of pathological diagnostic report of ≥1 baseline tissue specimen (core biopsy or resection) graded G1, G2, or G3 per WHO 2022/2026 criteria, obtained during routine diagnostic workup.
• ≥1 metachronous (during the follow-up) pathology diagnostic report (biopsy/resection) from a progressive lesion, acquired outside the initial diagnostic/therapeutic window.
• Confirmed HGT on a tumor sample performed during follow-up (“HGT sample”) in comparison with the “initial sample” (i.e., the sample performed at the diagnosis of metastatic NET; or primary tumor sample if performed less than 1 year before the diagnosis of metastases), defined as:
  • Progression from G1/G2 NET to G3 NET, or
  • Pre-existing G3 NET (Ki-67 <55%) with absolute Ki-67 increase ≥20% (e.g., 40% to 60%).
• Minimum data completeness: information on types of therapies prior to HGT, OS status, description of treatments administered following HGT
  

Exclusion criteria:

• Primary neuroendocrine carcinoma (NEC; small/large cell), mixed neuroendocrine-non-neuroendocrine neoplasm (MiNEN), or G3 GEP-NET with upfront Ki-67 ≥55%
• Patients declining data use/consent where required by local regulations
• Duplicate cases across centers.

Methods

Data Collection and Management

From eligible patients, medical charts will be evaluated to collect relevant clinical data. Data entry uses the centralized, REDCap-based ENETS platform hosted by the Coordinating Center for Clinical Trials at Philipps-University Marburg, Germany. Data will be collected by local investigators and filled into the ENETS database (level 2, extended version: 250 items).

Data will be extracted from electronic medical records by local investigators. ENETS level 2 fields are mandatory; HGT-specific variables will be added as customized modules.

Study activation follows protocol approval by the ENETS DB Core Working Group; data lock occurs 3 months after the last case entry, with interim data freezes at 50 cases for quality review.

As an optional sub-study, pathology review will be undertaken by expert pathologists to evaluate digitalized tumor paraffin-embedded tumor slides to quantify Ki-67, determine cell morphology and review immunohistochemistry (IHC) expression markers. Additionally, depending on the number of cases with somatic NGS at HGT, a pooled data analysis will be conducted to explore HGT signatures.

Baseline Data (prior to HGT)

Baseline is defined as the time of diagnosis of metastatic disease.

• Demographics/Clinical data: Date of birth, sex, ECOG PS (0-4), race and ethnicity.
• Tumor characteristics: Date of diagnosis of metastatic disease, primary NET site, timing of metastases (synchronous or metachronous), metastatic sites (liver, bone, nodes, peritoneum), NET functionality (per local evaluation), known hereditary syndrome
• Pathology on the “initial sample” (i.e., the sample performed at the diagnosis of metastatic NET; or primary tumor sample if performed less than 1 year before the diagnosis of metastases): site of sample (primary tumor, metastases), type of sample (biopsy, surgery), Ki-67 index, NET grade (G1-3)
• Molecular biology: done or not (if yes: to be recovered anonymously apart from ENETS DB), TMB (high ≥10 mut/Mb), MMR status (MMRd is defined as pathological variant on NGS, and/or negative immunostaining)
• Nuclear Imaging: uptake on SSTR-PET (Krenning score), uptake on 18F-FDG PET
  
  

Follow-up data (between baseline and HGT onset)

• History of treatments: watch-and-wait, long-acting SSAs, targeted therapies (everolimus, sunitinib, cabozantinib, lenvatinib), PRRT (177Lu-DOTATATE), chemotherapy [alkylating-based, oxaliplatin-based)
• Outcomes of treatments: start/end dates, number of cycles, best RECIST 1.1 response, date of progression
• Additional pathological samples (i.e., other than “initial sample” or “HGT sample”): site of sample (primary tumor, metastases), type of sample (biopsy, surgery), Ki-67 index, NET grade (G1-3)
• Molecular biology: done or not (if yes: to be recovered anonymously apart from ENETS DB), TMB (high ≥10 mut/Mb), MMR status (MMRd is defined as pathological variant on NGS, and/or negative immunostaining)
  
  

At HGT onset (within 30 days)

• Clinical: ECOG stage, symptoms (e.g., pain), clinical progression (per local evaluation), indication for “HGT sample” (routinely performed upon tumor progression, guided by unexpected disease behavior, clinical trial, surgical resection)
• Pathology: site of sample (primary tumor, metastases), type of sample (biopsy, surgery), Ki-67 index, NET grade (G1-3), morphological features of de-differentiation (per local evaluation), other IHC markers (chromogranin, synaptophysin, Rb, p53, DLL3)
• Molecular biology: done or not (if yes: to be recovered anonymously apart from ENETS DB), TMB (high ≥10 mut/Mb), MMR status (MMRd is defined as pathological variant on NGS, and/or negative immunostaining)
• Nuclear Imaging: uptake on SSTR-PET (Krenning score), uptake on 18F-FDG PET
• Treatments: long-acting SSAs, targeted therapies (everolimus, sunitinib, cabozantinib, lenvatinib), PRRT (177Lu-DOTATATE), chemotherapy [alkylating-based, oxaliplatin-based), immunotherapy (single, double), best supportive care
• Outcomes of treatments: start/end dates, number of cycles, best RECIST 1.1 response, date of progression (per local evaluation)
• Outcomes: last follow-up/death date

References:

1. Bourdeleau P, Pokossy-Epée J, Hentic O, et al. Temporal increase in Ki-67 index in patients with pancreatic neuroendocrine tumours. Endocrine-Related Cancer. 2025;32:e240321.

2. Botling J, Lamarca A, Bajic D, et al. High-Grade Progression Confers Poor Survival in Pancreatic Neuroendocrine Tumors. Neuroendocrinology. 2020;110:891–898.

3. Mollazadegan K, Botling J, Skogseid B, et al. The impact of re‐characterizing metastatic pancreatic neuroendocrine tumors: A prospective study. J Neuroendocrinology. 2025;e70040.

4. de Mestier L, Cohen D, Masliah Planchon J, et al. Temozolomide treatment induces an MMR-dependent hypermutator phenotype in well differentiated pancreatic neuroendocrine tumors (1182O). Annals of Oncology. 2023;34 (suppl.2):S701–S710.

5. Backman S, Botling J, Nord H, et al. The evolutionary history of metastatic pancreatic neuroendocrine tumours reveals a therapy driven route to high‐grade transformation. The Journal of Pathology. 2024;264:357–370.

Publication policy:

Authorship will follow the ICMJE recommendations and the specific ENETS rules outlined in the ENETS DB Authorship and Publication Policy.

Study timelines:

• April 2026: Study proposal submitted to ENETS database committee.

• June 2026 –  March 2027: Inclusion of patients into the ENETS database.

• April 2027 – June 2027: Quality check of entered cases.

• July – October 2027: Data analyses and manuscript preparation

• November 2027: Manuscript/ENETS abstract submittance