Safety and Efficacy of Cilta-Cel in RRMM: Real-World Experience | Surbhi Sidana, MD | IMS 2024

Hi, I'm Surabhi Sadana from Stanford University, and we presented data on real world experience with Silted Capsule gene auto-new cell CAR-T or Silted Cell CAR-T. This was a multi-center study done at 16 sites across the US Multiple Myeloma Immunotherapy Consortium. So as the audience might know, Silted Cell was FDA approved in 2022. It's a BCMA-directed autologous CAR-T that was approved based on the phenomenal results seen in the CAR-T1 clinical trial, with response rates of over 98%, CR rates of 83%, and a PFS achieving three years. What we also know is that clinical trial patients can differ significantly from patients we see in a real clinical practice, where patients have more comorbidities, are not as fit. So that's why it's very important to elucidate the experience of patients being treated with standard of care therapy, especially for CAR-T. So we evaluated patients. This was 255 patients who underwent aphoresis with the intention to manufacture standard of care CAR-T. Out of this, 236 patients, or 92.5%, ended up receiving Silted Cell CAR-T. 19 patients did not, primarily due to disease progression or death. Out of these 236 patients, 192 patients received a conforming CAR-T product. There were about 19% of patients where the CAR-T product did not meet the release criteria that were set by the FDA. These patients were still treated, but were treated on an expanded access protocol. And I define these two cohorts separately because the efficacy can differ here. So when we look at the baseline characteristics of the entire 236 patients, this was a heavily pre-treated group of patients. Median of six prior lines of therapy, 30% were penta-refractory, 14% had prior BCMA therapy, something that was not allowed in the CAR-T1 study. This was also a high-risk cohort of patients. 39% of the patients had high-risk cytogenetics defined as deletion 17p, translocation 414, and translocation 1416. This was higher numerically than seen in CAR-T1. 26% of patients had extra medullary disease defined as disease that was non-contiguous with bone. This was double than what was seen in CAR-T1. So this was a heavily pre-treated population with high-risk features. So how did they do? Well, and how were they in terms of comorbidities? That's the first question. As we expected, over half of the patients did not meet trial eligibility criteria due to low blood counts, poor performance status, organ dysfunction, or treatment history. This is not surprising. In terms of toxicity, despite half of the patients not meeting criteria for eligibility due to comorbidities or other things, the toxicity profile we saw in the real world was consistent with the clinical trial data. Most people did get cytokine release syndrome, but the risk of grade 3 or higher cytokine release syndrome was only 5%. ICANNs of immune cell-associated neurotoxicity syndrome was seen in 14%, but grade 3 or higher events were seen in 4% of patients. Now, delayed neurotoxicity is a concern with Cytosel more commonly than with other CAR-Ts. We saw a 10% incidence of delayed neurotoxicities. Only 2% of these were Parkinsonian neurotoxicities of five cases. The majority were cranial nerve calcys. About half of them had recovered, but only one out of the five patients had recovered with Parkinsonism. In terms of efficacy, so the median follow-up that we have is 13 months of follow-up. And in terms of response rates, we observed very high response rate, 89% in the whole cohort, 70% CRA. When we look at the patients who got conforming Cytosel, these response rates are much higher, achieving near 95%, and CRA rates also go higher. So getting close to the CAR-T1 population. When we look at progression-free survival, the median has not been reached. The one-year estimated PFS in the entire cohort was 68%. When you look at conforming Cytosel, it's higher, it's near 75%. And when you look at the data that was initially presented with CAR-T1, the initial publication in Lancet, the estimated one-year PFS at that time was 77%. So this is within range of what we were seeing in the trial, maybe numerically slightly less. So when we look at what happens after prior BCMA-directed therapy, a sequencing of therapies is such an important question in myeloma with BCMA-ADC, BCMA-bispecific, and CAR-T being available. As a reminder, 14% of our patients had received prior BCMA-directed therapy, most commonly ADC, but there were several patients with CAR-T and prior bispecific antibody. And as expected, we observed lower overall response rates and complete response rate. The overall response rate here was 70%, the CR rate was 42%, and the median PFS was 13.6 months. But does that mean should we not do sequencing of BCMA-directed therapy one after the other? We also saw another signal in our study that the timing of sequencing may matter, and that actually informs clinical practice. So less than half, about half of the patients, 45% of the patients in our study, had received BCMA-directed therapy within six months of cell-to-cell infusion. And those patients did particularly had bad outcomes in terms of response rates and PFS. The ones who had received prior BCMA exposure more than six months before cell-to-cell had much better outcomes in terms of response rates and progression-free survival. In fact, the response rates in those who had received prior BCMA-directed therapy more than six months prior were 94% compared to only 54% those within six months. The PFS for more than six months prior was almost 17 months compared to only six months for those within six months of cell-to-cell infusion. So if someone's gotten prior BCMA-directed therapy, if there is an option, it's a good idea to wait at least six months before giving them cell-to-cell or other BCMA-directed CAR-2 cell therapy. We then investigated, well, what are some other factors that can impact the response impact outcome of patients? Do certain patients do better? Do certain patients do worse? In a multivariable analysis adjusting for everything, for progression-free survival, prior BCMA-directed therapy again was significant borderline zone for inferior PFS. But two things that we already know from prior data, extra medullary disease and high-risk cytogenetics resulted in inferior progression-free survival and overall survival. But a new signal that is coming out with the CAR-T literature and that also came out with our study is if you have an elevated baseline ferritin level more than 400 prior to start of lympho depletion, those patients had inferior progression-free survival and overall survival independent of everything else. Now the biological rationale for that is not very clear, and I think this needs to be investigated. When we also look at other toxicity issues like non-relapse mortality, we saw a 10% non-relapse mortality in our patients primarily due to infections. And this tells us that we still have room for improvement in infection prevention and improving non-relapse mortality. Second primary malignancies are also a concern with CAR-T. With a median follow-up of 13 months, our rate here in the real world cohort is 8.5%. When you exclude non-melanoma skin cancers, it's 5.5%. And when you look at just myeloid neoplasms or leukemia, it's 1.3%. So to summarize, despite over half of the patients in the real world cohort being not eligible for the pivotal CAR-T-1 trial, we saw deep responses, we saw high response rates, and these responses were durable. So CAR-T is feasible in the real world with Celtosel. The toxicity profile was consistent with trial data, but I think we do need to take more measures so that we can decrease non-relapse mortality. And we need to follow patients closely to see what the risk of second primary malignancies will be over the long term, as these are delayed toxicities. So thank you very much for listening. I hope this was helpful.