Recommendations for treatment do not differ greatly based on whether a patient has grade 1 or grade 2 disease. If a patient has a Ki-67 proliferative index of greater than 10%, you might want to increase your monitoring frequency. The factors that ultimately determine treatment include how the disease develops and manifests over time, not necessarily the patient’s baseline grade, which reflects just one biopsy or one surgical specimen at a particular point in time. The stability of the patient’s tumor in the recent past is likely a better guide than the tumor grade of 1 vs 2 when it comes to determining the best monitoring frequency. For example, if the tumor has been stable for more than 6 months, you can consider gradually increasing the monitoring interval. 

One of the first questions that relate to individualization is whether the patient is a candidate for surgery. Resection of the primary tumor may be advisable in certain scenarios, and sometimes resections are performed prophylactically. For instance, resection of the primary tumor is certainly indicated if it is causing the patient to be symptomatic. Debulking may be possible, and decisions about debulking depend on how much disease is in the liver, the pattern of liver involvement, and how much disease is outside of the liver. There can be a very large tumor involving an entire hepatic lobe. By removing that lobe, you can achieve significant debulking. In contrast, in a patient with a similar hepatic tumor burden but a more scattered pattern of involvement (ie, many small tumors in both lobes), surgical debulking becomes much less feasible. 

The standard treatment for metastatic unresectable NETs with carcinoid syndrome is a somatostatin analogue (SSA), for both control of the syndrome and stabilization of tumor growth. For patients who progress, you can try increasing the dosage or dosing frequency of the SSA for better control. 

The standard second-line therapy for progressive, metastatic, small bowel NETs consists of peptide receptor radionuclide therapy (PRRT) in patients with somatostatin receptor–expressing tumors. However, we still recommend liver embolization as an alternative second-line therapy in patients with liver-dominant disease if they progress on an SSA. 

The difference between a grade 1 and a grade 2 well-differentiated NET may be an artificial distinction in some respects. That is, if one were to biopsy a couple of different sites in the same patient, the Ki-67 proliferative index might be 2% in one sample and 6% in the other, meaning that the same patient might receive a classification of grade 1 or grade 2, depending on the site sampled. I personally review the patient’s symptoms and the disease burden, and I do not know that a Ki-67 of 2% or 6% would make a big difference in how I assess and treat that individual patient. Once you have a patient whose disease starts to tell its own story and the track record of their disease starts to unfold, then I am a little less concerned about what grade they were assigned at diagnosis.

At the population level, however, grade 1, grade 2, and well-differentiated grade 3 NETs are associated with different long-term prognoses. Additionally, the Ki-67 proliferative index can change over time. As the Ki-67s get higher (eg, 10%, 15%, or 20%), that may influence what we anticipate for the future, and I might follow that patient more closely (if I am following them) or treat them more aggressively (if I am treating them).

Regarding surgery/embolic therapy vs PRRT, I would consider trying PRRT in a patient who had symptoms and a significant extrahepatic metastatic disease burden, outside of their primary and regional tumor. Surgical or embolic techniques are likely more reliable for trying to improve difficult symptomatic situations if they are due to the locoregional disease or significant liver involvement.

I agree with my colleagues that baseline tumor grade is a single factor out of many that should be considered. The patient's care is individualized according to their biology, symptomatology, disease sites, and disease bulk. If you were going to try to match a monitoring frequency to their biology, for grade 1, you might scan them twice per year. For grade 2, you might consider scans every 3 or 4 months. For patients with grade 2 NETs, depending on how things are unfolding, we may start them on a treatment earlier rather than waiting for disease progression.

A small group of patients might de-differentiate into a more aggressive phenotype, and then we would have to adjust their treatment accordingly. The role of chemotherapy has not been well described for patients with grade 3 disease. There are some data to support the use of capecitabine. If the tumor is grade 3 and the patient is not syndromic, we may still use an SSA or we might go straight to chemotherapy. For grades 1 and 2, we would never go straight to chemotherapy; we would only use chemotherapy if the tumor de-differentiated. I have seen a few patients who have poorly differentiated NETs that are nonetheless functional. It is not common, but it does occur.

Normally, for grades 1, 2, and 3, if they are well differentiated, you would generally expect them to express somatostatin receptors, and the DOTATATE positron emission tomography/computed tomography would show uptake. PRRT would be useful for patients with significant tumor bulk and significant bone involvement, along with lymph node and, potentially, liver involvement.