The immune landscape in gastroenteropancreatic neuroendocrine tumors (GEP-NETs) is a complicated space. Previous and ongoing work trying to improve the immune response to GEP-NETs includes examining the use of ICIs, VEGF inhibitors, CAR T-cell therapies, bispecific antibodies, and PDCs.

Well-differentiated NETs generally tend to have a fairly cold immune landscape and do not attract a lot of immune attention. Going back several decades, we learned about how there were some targets on the tumor cells that circulating T cells could potentially react to but for some reason did not. That aligns with observations of poor T-cell infiltration in NETs. A few groups have observed that there seems to be a suppressive macrophage-driven environment rather than a particularly active T-cell–driven environment.

A number of studies have evaluated ICIs in neuroendocrine neoplasms. The most prominent studies of single-agent ICIs investigated the use of spartalizumab and pembrolizumab, and they were resoundingly negative. Regarding using ICIs in combination, I think that the issue with this approach is partly that we do not know what a rational combination would be based on the specific immune milieu.

Work combining ICIs with VEGF inhibitors has been somewhat disappointing. Even though there were some patients who responded in clinical trials, it was really hard to tease out how much of the benefits were from the VEGF inhibitor vs actual antitumor immunity. So, not a lot of additional research is currently happening there. A very unusual finding in a trial evaluating an ICI in combination with a VEGF inhibitor in patients with advanced NETs was the remarkably tight association between PD-L1 expression at really any level and response to therapy, which was very different from what was previously observed with single-agent ICIs. Although this was rather provocative, that trial came out a couple of years ago and does not seem to be going anywhere.

There was significant interest 5 to 10 years ago in SSTR-directed CAR T-cell therapy. As the research progressed, this eventually led to research specifically targeting CDH17. There is a registered study, which is not yet open for recruitment, exploring targeting CDH17 using CHM-2101. In my mind, the key with CAR T-cell therapy is its therapeutic index. And with such an incredibly broad distribution of outcomes, depending on grade, stage, primary site, and duration of disease, the real key question is: When is the toxicity risk worthwhile? I think that it will be really important to focus on this as the data come out.

The primary target for bispecific antibodies, ADCs, and PDCs has been the SSTR, which is an established target in the space. Some initial phase 1 data on the SSTR CD3 bispecific antibody XmAb18087 was presented years ago, and then we had PEN-221, a PDC that uses an SSA to deliver the toxin DM1, but patients did not tolerate that well.

Overall, I would say that the immune landscape in GEP-NETs certainly is a complicated space. I think that the answer seems to be that simply focusing on T cells may not be the right maneuver, even though that is where so many of our agents have been developed. As additional therapies enter clinical trials, I do not think that we should completely discard immune agents in the field. I would not give up.