CAR T-Cell Therapy Clinical Trials for R/R DLBCL

Matthew J. Frigault, MD: There are 3 main studies of CAR [chimeric antigen receptor] T cells in DLBCL [diffuse large B-cell lymphoma]. The first, ZUMA-1, was published in 2017 in the New England Journal of Medicine. This was published by Sattva Neelapu from MD Anderson Cancer Center. It was 1 of the first studies published. It was also what led to the first approval for CAR T cells in lymphoma for axicabtagene ciloleucel.

The exciting component of this clinical trial is that they had about 111 patients who were treated. They had a 99% manufacturing success rate. Then they had an objective response of about 82%, with a complete response [CR] rate of about 54% at a median follow-up of 15.4 months. They subsequently updated those data, which have shown that these responses are durable. The most common adverse effects were cytokine release syndrome and neurotoxicity, which were seen in a high percentage of patients, but were manageable on this study. All participants received inpatient treatment.

That led to the FDA approval with an accompanying REMS [Risk Evaluation and Mitigation Strategy] program, which many of us are trained in as we prescribe. That was the first clinical trial that was published. It led to the first approval for CAR T cells in lymphoma. The first CAR T-cell approval overall was tisagenlecleucel for pediatric and young adult B ALL [B-cell acute lymphoblastic leukemia]. However, that same tisagenlecleucel product was then subsequently approved a few months later based on the JULIET study. That was published in the New England Journal of Medicine in 2019. The lead author is Steve Schuster. That study comprised 93 patients who had a median follow-up of about 14 months.

The best overall response rate was around 52%, with a 40% CR rate. That is lower than what we saw in ZUMA-1. However, when you go out 12 to 24 months, the progression-free survival [PFS]—meaning the percentage of responders who end up relapsing—is around 35% to 40% for both studies. They’re likely comparable from a PFS benefit perspective.

We’re currently looking at whether subpopulations benefit more from 1 product or another. But high response rates mean nothing unless those response rates are durable. When I try to critique these studies, I’m looking at the 6-, 12-, and 24-month durability of those responses. They do appear to be comparable across all products. The difference in study design between ZUMA-1 and JULIET was that patients were enrolled at apheresis. Patients were screened, and if they met eligibility and there was a spot available, they were then enrolled. Patients went on to get apheresis and treatment without bridging therapy being available. Patients were very sick, had progressive disease, and did well anyway. For JULIET, patients were enrolled and collected, and then their cells were stored as they were awaiting a manufacturing slot.

Bridging therapy was allowed because there was a much longer delay between enrollment, apheresis, obtaining a manufacturing spot, and subsequent treatment. There’s a bit of selection bias and differences in bridging therapy and management. However, ultimately, these patients were also very sick. They had chemotherapy-refractory disease and they were getting chemotherapy as bridging therapy, which is not felt to be curative. As such, they were very sick. The most recent trial that came out was published from someone at our institution, Dr Jeremy Abramson. That was the TRANSCEND-NHL-001.

This was the largest cohort of patients at multiple dose levels using the product lisocabtagene maraleucel, which is currently being reviewed by the FDA for commercial approval. That included about 344 patients—a large cohort—at multiple dose levels with the recommended target dose of about 100 million cells. We can see here that the overall response rate was around 73% with a complete response rate around 53%. Again, we’re still looking at around 40% progression-free survival going out to 12 months. It’s very similar.

There are some differences, though, in toxicity across these products. When you’re looking at things like axicabtagene ciloleucel in ZUMA-1, they tended to have higher CRS [cytokine release syndrome] and higher neurotoxicity rates. That was using the Lee grading system from [Daniel] Lee, et al in 2014. When you look at JULIET for tisagenlecleucel, it was prior to having a consensus on an appropriate grading system, which led to them using the Penn criteria. I will speak about that shortly, but it tended to overestimate grade 3 toxicities.

Following the JULIET study, there are also some safety differences that we’re seeing in TRANSCEND-NHL-001, which was published in The Lancet in 2020 by Jeremy Abramson. Overall, they had very low CRS and neurotoxicity rates. They’ve seen them in about 40% of patients. There are around 10% and probably 2% of patients who had severe neurotoxicity and severe CRS. Overall, we’re seeing improvements, whether they are product-specific attributes or we’re learning from the prior studies how to manage these patients.

We’re using tocilizumab and steroids earlier, and as such, we’re seeing lower rates of severe CRS in subsequent studies.

However, there has been an evolution in management of these patients. When we go back to 2017, management was typically tocilizumab or steroids for grades 2 and 3 CRS. The same is true in JULIET. We weren’t sure at that point in time whether steroids or tocilizumab inhibited CAR T cells and whether they were detrimental to the overall outcomes.

We now think that tocilizumab is safe to give and does not impact CAR T function. TRANSCEND-NHL-001 benefited from that knowledge by being able to treat earlier and therefore, having less severe CRS and neurotoxicity. Taken together, this is a win across all 3 studies. We have 3 great products that offer PFS of around 40% for all comers. That is effectively curing high-risk patients who are otherwise chemotherapy refractory. It offers new treatment options for all patients with large cell lymphoma.

Transcript Edited for Clarity