Video
Use of CAR T-Cell Therapies in Earlier Lines of DLBCL
Author(s):
Panelists discuss data on the use of CAR T-cell therapies in earlier lines and implications of these data on the DLBCL treatment landscape going forward.
Transcript:
Kami Maddocks, MD: The last thing that I want to touch on is the new data we have looking at the CAR [chimeric antigen receptor] T product that we talked about in the second-line setting. Maybe we could have a little summary of those presentations and then how you all view the data. Do you see this as something that will move it forward? Does it apply to certain populations? How are you interpreting this?
David Miklos, MD, PhD: Thank you. I’m going to ask my colleagues to help me because we’re still digesting this data, and it’s quite complex. We used to complain that we didn’t have clinical trials. I almost feel like I’m going to complain that we have clinical trials now because of the obligation on us to present them in a fair manner. Let me take a swing at this. Let’s start out by identifying what we’re discussing.
We’re discussing whether the 3 commercially approved therapies in the United States for third-line treatment of diffuse large B-cell lymphoma (DLBCL) after refractory disease and receiving 2 lines of therapy, including anthracycline rituximab, could provide benefit by moving forward in the application of therapies to follow in patients who have induction failure to the initial therapy, frequently R-CHOP (rituximab, cyclophosphamide, doxorubicin, vincristine, and prednisone) or R-EPOCH (rituximab, etoposide, prednisone, vincristine, cyclophosphamide, doxorubicin)—, and maybe that's going to change—or those who have relapsed within 12 months of initiating or ending the induction cycle of 6 rounds of chemotherapy. The names of these studies are ZUMA-7 (NCT03391466) for axi-cel (axicabtagene ciloleucel), TRANSFORM (NCT03575351) for liso-cel (lisocabtagene maraleucel), and BELINDA (NCT03570892) for tisa-cel (tisagenlecleucel).
They were all randomized controlled trials, which is fantastic. They’re obviously not blinded. The differences in the treatment, for an autologous transplant, they were required to give salvage chemotherapy, demonstrate chemotherapy benefit, and then move on to high-dose chemotherapy with stem cell rescue, which was already outlined by Rich (Richard T. Maziarz, MD). That wasn’t going to be blinded. In the other arm, how would you make this fair if you could immediately go to a therapy of apheresis? But then you still had to manufacture and take into account the amount of time that would be necessary for manufacturing, which varies between the 3 products and, historically, the utilization of the products. Some of the products have been more used to being incorporated with bridging therapy. Bridging is after you’ve collected the peripheral blood mononuclear cells for CAR T production. That’s the therapy you gave after you’ve committed to making the therapy.
Those are 3 randomized studies. Two of the randomized studies—TRANSFORM and BELINDA— apheresed everybody who enrolled. Those 2 studies included the concept of crossover as part of the trial. As the advocate for the patient, I want to say that that was a favorable way to attract patients to the study. The axi-cel product that didn’t include apheresis of everybody basically stated that the crossover was already FDA approved, and so that would be something that would be covered in the United States by the payors in the first place.
The end point is the critical aspect. What would be a useful end point? Historically, we always talk about complete response rate, partial response rate, duration of response, and overall survival. But that wouldn’t work in this study if the change in therapy wasn’t going to be focused on progression, but by the lack of 50% response, or what we call a partial response, to salvage chemotherapy. All 3 studies used event-free survival as the primary end point for analysis of benefit. But they all recognized that long-term overall survival was going to be an end point that we all wanted to hear about. Crossover would confuse this, but we wanted to incorporate enough power in the study to be able to get to a 5-year survival benefit. Two of the studies are looking at 5-year overall survival benefit, and that’s what they powered for.
They powered for events. It’s like COVID-19 vaccine challenges these days. The trials were designed anticipating 70%, 85%, and 90% events. When you hit that number of events, the study would unblind, be analyzed, and then be reported. Three of these studies were then analyzed because the number of events had occurred. Two of the studies are showing benefit for event-free survival. The significance of the event-free survival was very large in the 2 studies. The axi-cel ZUMA-7 and the TRANSFORM study of liso-cel showed hazard ratios of 0.34, or a 3:1 or better difference in the event-free recognition or median event-free time. And because the primary end point is event free, the time to the events was 24 months in the ZUMA-7 study, and the event-free survival was 40% vs 16%. In the TRANSFORM study, at 12 months, it was 44.5% vs 23%.
The median event-free survival was a factor of 4 or more in both of those studies, so that wasn’t a small difference. Many will discuss that this study showed what Rich already discussed with the CORAL study (NCT01379014), that platinum-based salvage chemotherapy isn’t effective in patients who have the highest-risk induction failure of not responding to chemotherapy up-front or progression of their disease within the first 12 months. None of these studies are addressing late relapse or alternative reasons why patients come to us.
CR [complete response] rates using therapy in the second line instead of the third line were a little better. The axi-cel CR rate was 65%. The TRANSFORM liso-cel CR rate was 66%. That might be due to fresher T cells, more fit immune systems, and less exposure to prior chemotherapy—arguing again that these therapies, which are good in the third line, might be even better in the second line. And in fact, when we look at the publications, we’re all struck by how if you got to the autologous transplant and if you got the CAR T cell, there was a remarkable plateau in the patients who received both therapies. Unfortunately, none of [the authors of] these 3 studies are here today to discuss what the benefit and persistent benefit of an autologous vs CAR T therapy is in those who received the therapy.
Jason Westin, MD: I agree with you, Dr Miklos. These are paradigm-shifting studies, specifically ZUMA-7 and TRANSFORM. ZUMA-7 is the primary analysis with 24 months of median follow-up, and the TRANSFORM is an interim analysis with 6 months of median follow-up, so we have to caution the interpretation here because we certainly need more follow-up. But so far, so good. It looks wonderful.
Regarding your comments about what’s really being compared here, whether it’s CAR T-cell therapy vs transplant or CAR T-cell therapy vs platinum chemotherapy, the idea here is the treatment strategy. What we do in practice right now is give platinum chemotherapy, and those who are chosen go on to transplant. You don’t know ahead of time which proportion of patients or which patients are going to get to transplant. However, we know it’s not a majority. Platinum chemotherapy doesn’t work on refractory patients, and these trials all had a significant majority of patients who were either refractory or early relapsing.
On these trials, almost everybody got CAR T-cell therapy in the CAR T-cell arm. Approximately a third of patients got transplant in the transplant arm. Of those who crossed over on the transplant arms, more patients got CAR T cells than transplant. To me, that’s game over. For these refractory patients, CAR T-cell therapy should be the de facto new standard, and transplant should be reserved for a discussion about patients who are less refractory, who have a later relapse, or who have already had chemotherapy and are shown to be sensitive to that chemotherapy.
David Miklos, MD, PhD: In the BELINDA study, patients were enrolled and apheresed, and products were made for all patients. Then they received 2 cycles of a platinum-based chemotherapy. If they achieved a PR (partial response) or CR, they went on to the autologous transplant if that’s where they were randomized to. If they were randomized to the CAR T arm, it was the choice of the treating physician, but 84% of the patients received bridging therapy with platinum-based chemotherapy, and greater than 50% received 2 cycles of bridging therapy, which essentially puts you in the autologous arm, not on the CAR T arm.
For those 54% who got 2 cycles of salvage and still had measurable disease, that’s actually third-line therapy, which is what we’ve been doing for the last 4 years, so that’s going to be a criticism. In the autologous arm, if the patients who received salvage therapy didn’t achieve a PR, they gave you another chance to get a PR and offered you the ability to cross over or get another platinum-based chemotherapy for 2 more cycles. But the real bugaboo in the analysis is the predetermined time to event-free survival detection. I want to point out that the time to event-free survival was going to be—
Jason Westin, MD: Twelve weeks.
David Miklos, MD, PhD: It’s different in every study. In ZUMA, it was 50 days and 100 days, which is 7 weeks or double that: 14 weeks. On TRANSFORM, it was 9 weeks, or 56 days and 112 days. Those are very similar. But it was 12 weeks in the BELINDA study. Being 12 weeks instead of 7 weeks allowed a whole bunch of things to happen. You could have a good response that got worse. You could have a bad response that got salvaged with a second line of platinum. There were a lot of things that could’ve happened where we don’t know what happened. I’m sure we’ll refine that by going through the data management and trying to figure that out in the BELINDA study. But that late 12-week assessment of event-free survival is the bugaboo of the comparison of the studies. It’s not the rest of the design.
Jason Westin, MD: That all comes down to the time to manufacture the product. We know that the product manufacturing for tisagenlecleucel has historically been a little slower, as seen in the JULIET study (NCT02445248). On the BELINDA trial, the median time to infusion was 52 days. That comes down to that late time point and the need to get 2 different types of platinum chemotherapy in the standard-of-care arm. Everybody was apheresed, but the products weren’t actually made until you had shown what your response was to the first type of platinum chemotherapy, hence the need to give another 1, because the long time frame to make the product started after the 6 weeks of chemotherapy.
On that trial, it’s confounded by the use of bridging therapy or bridging therapies. Twelve percent of patients in the CAR T-cell arm got 2 different types of bridging therapy. This is a very complicated trial design. Unfortunately, it’s difficult to know how much of that confounded the results vs the efficacy of the product.
Transcript edited for clarity.
