Loading
10 slides generated from the cancer page, with a quiz from the open benchmark and speaker notes that cite the sources. Arrow keys move between slides; Print gives one slide per page.
Grade 3 well-differentiated neuroendocrine tumours divide fast enough to be called grade 3 yet still look and behave like their slower relatives rather than like neuroendocrine carcinoma. Recognised as separate since 2017, they keep the somatostatin receptor, respond less well to platinum chemotherapy, and in the NETTER-2 trial were among the first treated with lutetium-177 dotatate up front.
Until 2017 every neuroendocrine neoplasm with a Ki-67 above 20 percent was called neuroendocrine carcinoma and treated like small-cell lung cancer. Pathologists noticed that some of these tumours kept the organoid architecture, uniform nuclei and somatostatin receptor expression of well-differentiated tumours, and that their patients lived far longer than those with carcinoma. Multicentre series, notably Heetfeld and colleagues (2015), showed that these tumours, usually pancreatic and usually with a Ki-67 between 20 and 55 percent, responded poorly to platinum-etoposide but survived longer, and the NORDIC NEC series (2013) had already found that a Ki-67 below 55 percent predicted the same pattern. The WHO classified pancreatic NET G3 as a distinct entity in 2017 and extended it to the whole digestive system in 2019; molecularly these tumours carry the MEN1, DAXX and ATRX changes of neuroendocrine tumours and retain p53 and Rb, whereas carcinoma loses them, which is why p53 and Rb immunohistochemistry is now used when morphology is ambiguous.
Treatment evidence is thin because the entity is new and small. Capecitabine with temozolomide is the most used chemotherapy, on the basis of pancreatic tumour data from E2211 and retrospective grade 3 series, and everolimus and sunitinib are used with less evidence. Somatostatin receptor PET is usually positive, often with FDG avidity as well, and this dual pattern makes radioligand therapy plausible: NETTER-2 (Lancet 2024) was designed to include grade 3 tumours with a Ki-67 up to 55 percent alongside higher grade 2 tumours, and first-line lutetium-177 dotatate lengthened progression-free survival from 8.5 to 22.8 months across the 226 patients, the first randomised evidence in this group. COMPOSE randomises well-differentiated aggressive grade 2 and grade 3 gastroenteropancreatic tumours between 177Lu-edotreotide and CAPTEM, everolimus or FOLFOX, and is due to report in 2027.
| Setting | Approach | Guideline |
|---|---|---|
| Diagnosis | Morphology, Ki-67 and p53 or Rb immunohistochemistry to separate grade 3 tumour from carcinoma; somatostatin receptor and FDG PET together. | not mapped |
| Advanced, somatostatin receptor-positive | Lutetium-177 dotatate first line for Ki-67 up to 55 percent (NETTER-2); somatostatin analogue alongside. | not mapped |
| Advanced, shrinkage needed or receptor-negative | Capecitabine with temozolomide; everolimus or sunitinib for pancreatic tumours; platinum-etoposide for carcinoma-like tempo or Ki-67 near 55 percent. | not mapped |
| Trials | COMPOSE: 177Lu-edotreotide against CAPTEM, everolimus or FOLFOX in aggressive grade 2 and grade 3 gastroenteropancreatic tumours. | not mapped |
| Limited disease | Resection and liver-directed therapy as for other well-differentiated tumours. | not mapped |