<p>Among the high-grade neuroendocrine neoplasms, an important distinction is made between well-differentiated grade 3 neuroendocrine tumors (G3 NETs) and poorly differentiated neuroendocrine carcinomas (NECs). Treatment strategies for advanced G3 gastroenteropancreatic NETs (GEP-NETs) are still being refined, and data are limited. Guidelines suggest that extrapolation from lower-grade (grade 1/2 [G1/G2]) GEP-NETs to well-differentiated G3 GEP-NETs may be appropriate in some circumstances.</p>

Grade is one of the factors that we consider as we develop treatment strategies for patients. Although we also consider a number of other factors, including primary tumor site, disease extent, symptoms, and patient comorbidities, as we think about the tumor biology, grade can often give us a sense of how the disease may behave in the future. Higher-grade tumors tend to have a more aggressive biology and can grow and spread more quickly than lower-grade tumors. One may need to take a more aggressive approach to treatment in patients who have extensive, high-grade disease.

 

One of the most important distinctions to make as we think about high-grade neuroendocrine neoplasms is that of NET vs NEC. The entity of high-grade, well-differentiated NETs (eg, G3 GEP-NETs) is a relatively new diagnostic category, and we are refining our treatment strategies for this group of patients. I think that many of the options that are on the table for patients with lower-grade NETs are still on the table for those with G3 NETs, but we do sometimes think about starting with a more aggressive upfront treatment strategy. If there is more extensive, higher-grade disease, we might think about using a chemotherapy approach, particularly if a patient has a pancreatic NET. In addition, there may be a role for upfront PRRT, especially now that we have data from the NETTER-2 trial. In NETTER-2, patients receiving first-line ¹⁷⁷Lu-dotatate plus SSA therapy had an improved progression-free survival benefit compared with those on high-dose SSA therapy alone. There was also a higher response rate to ¹⁷⁷Lu-dotatate.

 

As previously noted, other factors in addition to grade need to be considered. Further, if the patient has relatively low-volume disease and is asymptomatic, an initial trial of an SSA could be considered with close monitoring and recognition that there may be earlier progression that would warrant an escalation of therapy.

As we think about selecting therapies in patients with more aggressive disease, grade can be viewed as one of many variables that may be predictive and/or prognostic, with the latter having importance in how well patients do, regardless of the decision about which therapy we should give them. The 2 concepts can be spliced out from each other, and I do find that, sometimes (particularly in historical documents), prognostic biomarkers such as grade may be mistakenly interpreted as predictive ones due to the lack of a better basis for a medical decision.

 

So, I think that it is important to consider those areas where we have clear evidence of a higher grade or a higher Ki-67 as a predictive or selective biomarker vs a prognostic biomarker. In the poorly differentiated space, which, of course, is a subset of higher-grade disease, we can think about Ki-67 being both predictive and prognostic in the classic NORDIC NEC study. The higher the Ki-67, the higher the likelihood of an objective response to platinum-based chemotherapy and the worse patients did overall in terms of overall survival. Both are true. I think that it illustrates the importance of being clear about the clinical context at hand and whether we are considering grade and Ki-67 as predictive, prognostic, or both, as well as what we really mean when we talk about using it for treatment selection.

As Dr Chan noted, well-differentiated G3 NETs is a new category in the World Health Organization (WHO) classification, and we do not have high-level evidence to guide treatment selection and whether we should treat these patients differently. In my own personal experience, the behavior of G3 NETs can be quite variable, from behaving more indolently like a G1 or G2 NET to behaving more aggressively like a poorly differentiated NEC. If the clinical behavior is discordant with what the biopsy is showing me, or if I am concerned and/or need additional data, I will sometimes consider rebiopsy. We do have data showing that there can be grade evolution over time in NETs, so I may sometimes consider rebiopsy, and this can sometimes guide our treatment decisions. It might push me in one direction or another.

 

We can also consider using imaging to help us think about a patient’s tumor biology. We know that SSTRs tend to be expressed on more well-differentiated tumor cells and less so on poorly differentiated ones. For those patients with well-differentiated G3 NETs, we might consider getting both SSTR imaging, either ⁶⁸Ga-dotatate positron emission tomography/computed tomography (PET/CT) or ⁶⁴Cu-dotatate PET/CT, and fluorodeoxyglucose (FDG) PET/CT imaging. Doing so may also teach us about the biology. Hopefully, we will learn more and get more prospective data to guide us, but, presently, I think that G3 NETs are challenging to treat.