Appropriate use of tissue sampling and somatostatin receptor PET imaging in the diagnosis of pancreatic neuroendocrine tumors: results of an International Delphi Consensus
- Open Access
- 02.05.2025
- Guidelines
Abstract
Materials and methods
Initial survey design
Expert panel recruitment
Delphi methodology
Statistical analysis
Results
N (%) | |
|---|---|
Total | 38 (100) |
Country | |
United States | 29 (76) |
Belgium | 5 (13) |
Canada | 2 (5) |
Italy | 1 (3) |
Norway | 1 (3) |
Specialty | |
Surgical Oncology | 13 (34) |
Medical Oncology | 8 (21) |
Gastroenterology | 8 (21) |
Pathology | 4 (10) |
Nuclear Medicine | 3 (8) |
Radiology | 2 (5) |
Practice Setting | |
Academic | 33 (87) |
Private | 3 (8) |
Community | 1 (3) |
Federal/Government Hospital | 1 (4) |
Years in practice, mean (range) | 13.4 (3–34) |
Practice at High Volume Institution (> 20 PNET patients per year)? | |
Yes | 36 (95) |
No | 2 (5) |
First-round survey
Second-round survey
Third-round survey
Diagnostic Workup 1: For a patient who is a surgical candidate with a > 2 cm pancreatic mass suspicious for nonfunctional PNET based on multiphase CT findings and initial laboratory evaluation: | Round | Consensus |
|---|---|---|
Statement 1: Somatostatin Receptor (SSTR) PET imaging, if available, is the most appropriate next step for workup and staging | 1 | 86% |
Statement 2: If the mass is SSTR PET-avid without aggressive features or radiological features of high-grade nature, and there is no evidence of metastatic disease, the patient should proceed to surgical resection without preoperative tissue sampling | 1 | 88% |
Statement 3: Tissue sampling via EUS-FNA or FNB is warranted if there is concern for aggressive clinical or laboratory features, or radiological features of high-grade nature (i.e., low SUV max), or if there is a discrepancy between CT and SSTR PET findings | 1 | 91% |
Statement 4: If SSTR PET imaging demonstrates metastases, tissue sampling of the largest, most aggressive-appearing lesion with the lowest risk of tissue sampling-related complication should be performed to determine grade and possible genetic subtyping | 1 | 86% |
Diagnostic Workup 2: For a patient who is not a surgical candidate with a > 2 cm pancreatic mass suspicious for nonfunctional PNET based on multiphase CT findings and initial laboratory evaluation: | Round | Consensus |
|---|---|---|
Statement 1: SSTR PET imaging should be obtained to determine extent of disease | 1 | 89% |
Statement 2: Tissue sampling should be performed to confirm the diagnosis and determine grade prior to initiating medical treatment | 1 | 97% |
Diagnostic Workup 3: For an asymptomatic patient with a ≤ 2 cm pancreatic mass suspicious for nonfunctional PNET based on multiphase CT findings and initial laboratory evaluation: | Round | Consensus |
|---|---|---|
Statement 1: If there is no concern for metastases or high-grade features based on multiphasic CT/MRI findings, options include 1) short-term observation (especially for lesions < 1 cm); 2) SSTR PET imaging and 3) tissue sampling. This decision should be made based on discussions with the patient, lesion size, and available resources | 2 | 89% |
Statement 2: If the mass is SSTR-avid without aggressive features or radiological features of high-grade nature, and there is no evidence of metastatic disease, radiological observation with a follow-up multiphase CT or MRI is appropriate in lieu of immediate tissue sampling • Comment 2a: For lesions greater than 1.5 cm that are SSTR-avid, surgical resection may be considered based on surgical risk/tumor location and discussion with the patient | 2 2 | 97% 92% |
Statement 3: If the mass exhibits growth on surveillance imaging, the patient should be referred for surgical evaluation and discussed in a multidisciplinary tumor board. Tissue sampling to rule out high-grade differentiation (G3) and SSTR PET imaging to rule out occult metastatic disease not seen by conventional imaging should be considered | 2 | 89% |
Diagnostic Workup 4: For a patient who is a surgical candidate with a pancreatic mass suspicious for functional PNET based on multiphase CT findings and initial laboratory evaluation: | Round | Consensus |
|---|---|---|
Statement 1: SSTR PET imaging is the most appropriate next step for workup and staging | 1 | 88% |
Statement 2: Tissue sampling is not required prior to surgical resection unless there is concern for aggressive features or radiological features of high-grade nature | 1 | 88% |
Diagnostic Workup 5: For a patient who is a surgical candidate with multifocal pancreatic masses and a clinical picture concerning for nonfunctional PNETs: | Round | Consensus |
|---|---|---|
Statement 1: SSTR PET imaging, if available, is the appropriate next step for workup and staging | 1 | 91% |
Statement 2: If a specific lesion exhibits aggressive or high-grade features, EUS evaluation and tissue sampling of that lesion should be considered | 1 | 88% |
Statement 3: Germline genetic testing should be obtained to evaluate for MEN1 or other genetic syndromes depending on the clinical context | 2 | 100% |
Final imaging statements | Round | Consensus |
|---|---|---|
Statement 1: Chest imaging is low yield for patients with small tumors and those with no evidence of hepatic or other abdominal metastases | 2 | 94% |
Statement 2: Low SUV max and poor correlation with multiphase CT findings are features on SSTR PET imaging that are concerning for high-grade PNET. This should prompt tissue sampling of the mass and possible FDG PET/CT for further evaluation of the lesion • Comment 2a: If both FDG PET and tissue sampling are deemed necessary for a suspected PNET, FDG PET should be performed before tissue sampling to avoid false positives due to biopsy-related inflammation and to help guide the choice of lesion to biopsy in the case of multiple lesions | 1 2 | 97% 88% |
Statement 3: For patients with pancreatic tail lesions that are suspicious for accessory spleen on multiphase CT and that may be SSTR-avid, a heat-damaged RBC scan, abdominal MRI with and without IV contrast, or EUS with or without tissue sampling should be considered to further characterize the lesion • Comment 3a: In centers with significant experience, EUS without tissue sampling may help differentiate between a PNET and accessory spleen | 2 2 | 84% 81% |
Statement 4: If SSTR PET imaging is u navailable, one should consider either referral to a center with SSTR PET imaging capability or an EUS to characterize suspected PNETs | 2 | 86% |
Statement 5: If SSTR PET imaging demonstrates uptake in the uncinate process of unclear etiology, or if the location of uptake is unclear (pancreas versus small bowel versus lymph node), EUS with or without tissue sampling or multiphase MRI can be used to further characterize the region of uncertainty • Comment 5a: Tissue sampling via EUS should be done if indicated. However, EUS without tissue sampling also has a role in characterizing the lesion | 1 2 | 82% 81% |
Statement 6: In situations where concern for multifocality exists, it is reasonable to obtain an EUS without tissue sampling prior to resection of suspected PNET to assess for additional lesions not seen on cross-sectional imaging | 2 | 85% |
Final tissue sampling statements | Round | Consensus |
|---|---|---|
Statement 1: If tissue sampling of a suspected PNET is deemed necessary, fine needle biopsy should be performed over fine needle aspiration to ensure adequate tissue sampling and avoid sampling error | 1 | 87% |
Statement 2: In-room cytology evaluation, if available, should be performed for EUS biopsies of suspected PNETs to ensure adequate sampling • Comment 2a: If available, in-room cytology is strongly recommended for difficult-to-biopsy lesions or for cases that involve re-do tissue sampling | 2 2 | 81% 91% |
Statement 3: When obtaining tissue, tissue sampling of an abnormal lymph node in lieu of the pancreatic mass is acceptable if deemed to be lower risk and in-room cytology is present to confirm the diagnosis | 2 | 91% |
Statement 4: A patient with a hypervascular pancreatic mass on multiphase CT and a history of another malignancy with the potential for hypervascular metastases (ex. renal cell carcinoma, hepatocellular carcinoma) should undergo tissue sampling of the pancreatic lesion to distinguish primary PNET from metastasis | 1 | 94% |
Surgical candidates with > 2 cm nonfunctional PNET |
• Utilize SSTR PET imaging for workup and staging • Proceed to surgical resection without preoperative tissue sampling if the mass is SSTR PET-avid, lacks aggressive features, and shows no metastasis • Consider tissue sampling only if there are imaging discrepancies or suspicion of aggressive or high-grade radiological features • If metastases are detected on SSTR PET, sample the most aggressive-appearing lesions with the lowest risk of complications |
Non-surgical candidates with > 2 cm nonfunctional PNET |
• Utilize SSTR PET to evaluate disease extent • Perform tissue sampling for definitive diagnosis and grading before initiating medical treatment |
Asymptomatic patients with ≤ 2 cm nonfunctional PNET |
• Options include short-term observation, SSTR PET imaging, or selective tissue sampling based on lesion characteristics (size and clinical features), patient discussion, and available resources • If SSTR-avid without aggressive features or metastasis, use multiphase CT or MRI for follow-up instead of immediate tissue sampling • Consider surgical resection for larger (> 1.5 cm) SSTR-avid lesions, based on surgical risk/tumor location and discussion with the patient • Refer for surgical and multidisciplinary evaluation if growth is detected during follow-up. Perform tissue sampling to exclude high-grade (G3) tumors and use SSTR PET to assess occult metastatic disease |
Surgical candidates with functional PNET |
• Utilize SSTR PET imaging for workup and staging • Consider tissue sampling before surgical intervention in the presence of aggressive or high-grade radiological features |
Surgical candidates with multifocal PNETs |
• Utilize SSTR PET imaging for workup and staging • Consider tissue sampling before surgical resection in the presence of aggressive or high-grade radiological features • Obtain germline genetic testing to evaluate for MEN1 or other syndromes |
Imaging |
• Prefer SSTR PET and multiphase CT/MRI for primary imaging • Use prompt tissue sampling and possibly FDG PET/CT for suspected high-grade PNETs (e.g., low SUV max or poor correlation with multiphase CT) • Perform abdominal MRI or EUS (with or without tissue sampling) for tail lesions suspicious for accessory spleen or uncinate process lesions with unclear etiology or uptake location • Consider EUS without tissue sampling prior to resection when multifocality is suspected to assess for additional lesions not seen on cross-sectional imaging |
Tissue Sampling |
• Opt for fine needle biopsy over aspiration to minimize sampling error • Utilize in-room cytology, if available, particularly for difficult-to-biopsy lesions or re-dos to ensure adequate sampling • Consider lymph node sampling (if abnormal) instead of mass sampling if in-room cytology is present • Perform tissue sampling for hypervascular pancreatic masses or history of other malignancies to distinguish primary PNETs from metastases |