Gamma Secretase Inhibition May Boost CAR-T Effectiveness in Some Myeloma

Hi, my name is Dr. Andrew Portugues. I'm an assistant professor at the Fred Hutchinson Cancer Center in Seattle. I see patients in my myeloma clinic and also on our stem cell transplant service. I'm presenting a poster at ASH this year on how gamma-secretase inhibition may mitigate the impact of low target BCMA level on the surface of myeloma cells. The broad strokes of this project was to look at the impact of a gamma-secretase inhibitor on patients who received a BCMA-directed CAR T-cell therapy. Gamma-secretase is an enzyme that cleaves BCMA on the surface of myeloma cells. It regulates the amount of BCMA that's present on myeloma cells in patients. It's something that reduces the amount of BCMA in a general sense for all patients. It can also lead to increased levels of soluble BCMA when the BCMA is shed from the surface of the myeloma cells. You can kind of think of BCMA as something, it's almost like hair that gets cut by the scissor that is gamma-secretase. When you're going after a target in myeloma, you want as much target on the surface of the myeloma as possible, and you don't want target that's just floating around as a distraction or decoy. The idea with using a gamma-secretase inhibitor in a general sense is to prevent this cleavage of BCMA on the surface of the myeloma cells to increase the target antigen density and the amount of this soluble BCMA that's been shed that could act as a distraction or decoy to the targeting. Gamma-secretase inhibitors do not. There have been a couple of gamma-secretase inhibitors that have been developed over the years and they were initially developed actually as a treatment for certain cognitive disorders like Alzheimer's and it didn't really work for that. They sort of sat on a shelf and they've been repurposed now that we're using BCMA targeting in multiple myeloma. In this project, basically we analyzed results of two separate phase one clinical trials conducted at Fred Hutch that were undertaken contemporaneously. One was Fred Hutch 9952 and in this trial all patients received the gamma-secretase inhibitor Crenogastat. They received three doses leading up to their CAR T cell infusion and then they received three doses after the CAR T cell infusion. In all patients, the amount of BCMA in the surface of the myeloma cells increased in some cases fairly dramatically. Then patients received the CAR T cell therapy. The specific BCMA direct CAR T therapy that we used was something called FCARH143 which is a product developed at our center. On the other clinical trial, which was Fred Hutch 9762, nobody received a gamma-secretase inhibitor. This wasn't a phase three trial where patients either received it or didn't in a randomized fashion. On one trial, the patients all received it and there were 18 patients in 9952 who received the gamma-secretase inhibitor and then in 9762 there were 25 patients who did not. We can do a secondary analysis where we look at the impact of the GSI by analyzing the patients from both trials. The characteristics of the patients on the two trials was similar. One thing that was not significantly different, but there were more patients in the 9952 trial where patients received the gamma-secretase inhibitor who had received prior BCMA therapy. There was a higher proportion of BCMA naive patients in 9762 where there was no gamma-secretase inhibitor. When we look at the impact of prior BCMA exposure, we find that the progression-free survival and the overall survival basically is a third in the BCMA exposed population compared to the BCMA naive. There was a huge impact of having prior BCMA exposure. When we wanted to compare apples to apples, we decided to focus on the patients who were BCMA naive in order to make the cross-trial comparison because otherwise the characteristics were similar. Patients received similar doses of the CAR-T product, the ECOG performance status, the age, the sex, the high risk proportion was very similar across the trials. EMD amount was similar across the trials. The baseline characteristics once we excluded the patients with prior BCMA exposure looked very similar. Essentially what we found was most importantly, there was a survival benefit to the group that received the gamma-secretase inhibitor. I think that's the top-line finding. We did see a trend towards increased acute toxicities post CAR-T treatments such as cytokine release syndrome and neurotoxicity. Again it wasn't statistically significant, but for instance there was one case of a grade 4 CRS and one case of a grade 4 ICANs that was seen in the GSI trial and we didn't see any such high-grade toxicities in 9762 which was the non-GSI trial. So it's possible that by increasing the amount of surface BCMA density, you can increase the proportion of patients who experience these toxicities and potentially put people at higher risk for higher grade toxicities. However, I think this is the first analysis to my knowledge that has shown something more than a theoretical benefit of up-regulating BCMA. We conducted some additional analysis to try to understand where there are some patients who really benefited from this. When we modeled this, what we found was that the greatest impact on survival seemed to be in the patients who had relatively low BCMA surface level. I think that that makes sense that probably for the patient who has average or high BCMA surface of their tumor cells, increasing that just goes from a high level to a really high level and it doesn't help. But for patients who have a low BCMA level, they may be the patients who really drive a benefit from a gamma-secretase inhibitor whereby you're increasing to a more moderate level. It is not a standard practice to measure the tumor BCMA level ahead of a BCMA-directed therapy. To my knowledge, there is not a commercial assay to do this. It is not something that's routinely done. As of right now, there isn't really a prognostic use for doing such a test. It would be super helpful if we had a test and then we could say, okay, if you're going to get a BCMA-directed CAR-T therapy and you have really, really low BCMA, this probably isn't going to work as well for you or might not work as well for you. We don't have a way of doing that. I think that it's sort of been debated for a while how low is too low and is there a level where it starts to impact outcomes? There's a level where it starts to impact outcomes. I think right now we still don't have a great sense of exactly where that is, but at least in our analysis, we do see that the low patients seem to benefit from the gamma-secretase inhibitor. I think it's sort of maybe providing some impetus to look at this more carefully. Clearly, if somebody has no target, then what are we doing? There theoretically has to be a level where too low is too low and the CAR-Ts are not functioning. Exactly where that is has yet to be determined. Why aren't we using gamma-secretase inhibitors all over the place? Why hasn't this really taken off? It's because there are some toxicities with gamma-secretase inhibitors. Namely, they can cause GI toxicities. A gamma-secretase inhibitor probably isn't a medication that people want to be on long term. I think that the strategy of using a gamma-secretase inhibitor to boost the target might not be an ideal strategy for a therapy that's given in a continuous fashion like a bi-specific antibody. I think that for a strategy like a CAR-T cell therapy where you can time the maximal upregulation of the BSCMA to the peak expansion of the CAR-T cell therapy, you might be able to get more bang for your buck given the toxicities that you might incur with a gamma-secretase inhibitor.