Addressing Real-World Use of Liso-Cel vs Axi-Cel in DLBCL

CAR T-cell therapies such as liso-cel and axi-cel in DLBCL were the focal
points during an Around the Practice program at the 2025 Tandem Meeting.

CAR T-cell therapies such as lisocabtagene maraleucel (liso-cel; Breyanzi) and axicabtagene ciloleucel (axi-cel; Yescarta) in diffuse large B-cell lymphoma (DLBCL) were the focal points during an Around the Practice® program that took place during the 2025 Tandem Meeting. Highlights from the conversation included a discussion about the results from the phase 3 TRANSFORM (NCT03575351) and ZUMA-7 (NCT03391466) trials.1,2

The discussion was led by Krish Patel, MD, director of Lymphoma Research and executive chair of the Lymphoma Research Executive Committee at Sarah Cannon Research Institute. He was joined by Saurabh Dahiya, MD, FACP, associate professor of medicine at Stanford Medicine; Nathan Denlinger, DO, MS, assistant professor in the Division of Hematology at The Ohio State University College of Medicine; Mohamed M. Hegazi, MD, associate professor of medicine, and section chief of the Blood Cancers, Cellular Therapeutics, and Transplant Program at the University of Louisville School of Medicine; and Samuel Yamshon, MD, assistant professor of medicine at Weill Cornell Medicine/New York Presbyterian.

Overall, the colleagues focused on which populations liso-cel and axi-cel were better suited for. Additionally, they discussed the real-world use of these therapies and whether that differed from what was observed in the clinical trials.

Liso-Cel vs Axi-Cel Trial Results

Patel / The TRANSFORM trial led to the approval of liso-cel in the second-line setting for early or relapsed DLBCL, [and] ZUMA-7 led to the approval of axi-cel [for patients with LBCL who are refractory to the first-line chemoimmunotherapy or relapses within 12 months of first-line chemoimmunotherapy].3,4 These trials are slightly different, first and foremost, when we think about the control arm, both use platinum-based chemotherapy to deliver high-dose chemotherapy with an autologous transplant. How patients got there was a bit different. For liso-cel, patients were allowed to undergo bridging chemotherapy…[similar to] ZUMA-7, but slightly different. Maybe we’ll call it a minimal or bridging-light approach. Corticosteroids were allowed, but no chemotherapies. Perhaps a bit of a difference there in terms of the biology of the patients in the 2 trials.

Crossover was built into TRANSFORM, where patients ultimately underwent collection and manufacture of the CAR T before randomization. That’s a very different approach than what we saw in ZUMA-7, where there wasn’t a protocol-specified crossover. [Still,] for patients who are not responding on the standard-of-care arm, we know that a significant proportion of them went on to get commercial CAR T-cell therapy, but again, a bit different, maybe introducing some delays. You have to go through the commercial pathway rather than [go] right away, so that’s going to impact how we interpret these trials.

The median age is similar. We recognize large cell lymphoma is mostly a disease of older adults, and when we look at those other key biological factors outlined, including being refractory to initial therapy, with early [laboratory results] showing similar proportions there. The secondary International Prognostic Index is an interesting factor. It’s something we measure in trials. I’m not sure I use it so much in the clinic, but there may be some small differences here. We hear a lot about vein-to-vein time. That time from when we collected the patient’s [T cells] to when [the CAR T cells are infused back into the patient], and we’re in a similar ballpark, slightly longer for liso-cel, with a median of 34 days compared with 29 days. These were global trials, so there were some variations in the global markets where the studies were done.

Thinking about efficacy, the most important thing we look at here is that both trials show us good long-term event-free survival [EFS]. The median EFS in TRANSFORM was 29.5 months, and [in] ZUMA-7 was 8.3 months. Here is where the shape of the curves tells the important story we see. When we’re looking at long-term EFS, in that 40% to 45% range for both trials, and [it] is slightly higher for TRANSFORM. Overall response rates are very high for both studies in the CAR T arm specifically, and the complete response rate, which is what we want to achieve in DLBCL, was numerically higher for TRANSFORM. Looking at progression-free survival [PFS], which is what we’re thinking about in terms of that potential of a durable cure, we see the 3-year PFS rate in the TRANSFORM study [is] just under 51% with slightly longer follow-up for ZUMA-7 at 4 years. [ZUMA-7] PFS is just slightly under 42%, and that translates into overall survival. [These trial results] support the idea that we need to get patients to CAR T-cell therapy when that curative outcome is the goal, and the patient is medically able to undergo that treatment.

Just diving a little bit here into the safety data, this is where we can start to see a distinction in the CAR T products themselves. This mirrors the experience we’ve had in other settings as well. We’re very fortunate to see that with the use of liso-cel, cytokine release syndrome [CRS] is not an event that happens in everybody. We see a 49% overall CRS rate, the rate of high-grade CRS, essentially, those patients needing intensive care unit–level care, is very low, and similarly, neurologic events are much less common with patients treated with liso-cel.

This is one of the exciting aspects of having 2 CAR T-cell therapies—to think about how that safety profile shapes our choices. I know we don’t have direct comparisons of liso-cel and axi-cel. We’re not likely to have that ever in clinical trials, but in practice, we have to come to a decision about which CAR T-cell [therapy] we’re taking forward. I would like to hear from all of you about how you approach that.

Dahiya / The real world or clinical experience that we have with both the products is in alignment with what was seen in the registrational trial. There are several real-world studies for both liso-cel and axi-cel, and the numbers in terms of toxicity and safety are quite comparable with what was seen. The most interesting thing is that in most real-world studies, there’s a clear demonstration that the majority of the patients would not be eligible, and yet, you would see a good and comparable outcome in the real-world setting.

The earlier consortium studies showed that this observation would prove that this is not a boutique therapy and can be delivered at scale, even though you’re working outside the auspices of a clinical trial. You can have these effective remissions that are long-lasting in terms of safety data for both [trials]. Liso-cel is quite comparable to what we see, but there is much less neurotoxicity with liso-cel compared with axi-cel. There’s a high-grade CRS risk as well between the products. It’s a patient-centric decision again, and you go through the data in a very detailed manner, and you make the decision [with them]. Does this represent truly the shared decision-making model that we have all learned in our training and are now practicing?

Patel / You highlighted so nicely [how] the patient has to be at the center of that decision-making, and so our job is to frame that data. Coming to the point you raised, where our experience in the real world represents a broader patient population. Sam, is that your experience as well, where maybe we’re able to deliver these therapies to a much larger population of patients [who] maybe wouldn’t have made it into these trials?

Yamshon / My experience completely mirrors Saurabh’s. Both the efficacy and safety outcomes that we’ve presented in the slides are uncannily similar to what we see so often in these trials, and then you bring it to the clinic, and it just does not pan out that way. They are very similar to what we see in the real world. What I would say is the one area where what we saw in the trials doesn’t mirror the real world is in the vein-to-vein time. We see a significant difference in the vein-to-vein time for axi-cel vs liso-cel. For axi-cel these days, our average is about a 14-day turnaround. I had a patient recently [who had] a 10-day turnaround. That does sometimes change your thinking in terms of a patient who has [aggressive] lymphoma and just needs to get their cells as soon as possible. Even if there’s someone who I might be a bit more worried about CRS neurotoxicity, I might still choose axi-cel for them because I know that I’m going to get it much quicker.

On the flip side, of course, as we’ve noted, there is now data from the phase 2 ALYCANTE pilot study (NCT04531046) assessing liso-cel and axi-cel but for the transplant-ineligible population.5 We know that we can deliver CAR T cells, both liso-cel and axi-cel, to patients who are older and frailer, but [for] those 85- to 88-year-olds, I’m more likely to choose liso-cel, just because I know that risk of neurotoxicity or CRS—we can generally handle [it]. The neurotoxicity is the scary part, and so I tend to be a little more conservative. Even if someone has a bit more aggressive lymphoma, I’m more likely to choose liso-cel. I present this to the patient, and I’ll tell them that these are the trade-offs that we’re making. This one we’re going to get faster. This one might be a little less toxic.

Patel / When we think about the applicability or the generalizability of that trial data in the real world, are there any areas that you’ve noticed where you think maybe things are different in one direction vs other populations [for which] you think this seems to be diverging from what we’ve seen in clinical trials?

Denlinger / In terms of patient selection, I feel like we’re treating much sicker patients than are on the clinical trials. For example, in these clinical trials, they had to be eligible for autologous stem cell transplant, have certain organ criteria, which is standard across the board with an [ECOG] performance status of 0 or 1, have good heart function, not just adequate, and good renal function, not just adequate. I would say our own data are even more striking in the fact that we are seeing the same efficacy outcomes in terms of PFS and response rates with bad organ function, heart failure, or renal failure. Patients who have poor performance status, are much older than [those in] the 74 to 75 range. That’s probably the biggest divergence we’ve seen in the real-world data. We’ve been able to push it quite far and have not reached what I think is any real, true limit, particularly with the use of liso-cel and how favorable its toxicity profile is.

Characterizing CAR T-Cell Therapy Use

Patel / To unpack a bit about what you mentioned, one of the things that we also may be seeing is our ability to manage toxicity has improved. When we think about when these trials were done and what we knew about how to approach early toxicity, we may have had a different approach, and we see, maybe in some senses, even slightly higher toxicity in the trials vs what we might be seeing in our center. Mohamed, I want to ask you: Has that been your experience as well? When you think about your real-world use of CAR T-cell therapies, are you seeing high-grade toxicities at comparable rates or lower rates?

Hegazi / Nathan said it very nicely: We are treating sicker patients, but I’m not seeing that the toxicity is going above what’s been reported [on clinical trials]. The only part that I would say that is different is what you said, [that]we now are grading it much better than 7 years ago. We have a clear algorithm for the team to treat; it’s not like we’re trying to use our own experience [in] how we’re treating it. It’s standardized and well-established for the team. We’re [getting] better at it, trying to figure out how to prevent [the toxicity], even if we can. In general, you’re grading it in a uniform way. You are treating it in a uniform way.

Now, having said that, when axi-cel [is first used], it’s not an if [CRS happens]. It’s when CRS happens, and it’s an if in liso-cel, because it’s less than 50% of the time that you will deal with any grade CRS. Even when it happens, most of the time, it’s at a lower grade. It’s not something that works and [then you bring] the option to [other] patients who can be treated with this and not have to spend a night in the hospital, meaning an outpatient approach.

Patel / We talked about these 2 trials where the definition of entry was based on transplant. Hearing from all of you, [we realize that] our experience with delivering this to a broader population is much greater. The safety profile of the CAR T cell [therapy] may allow us to think about it. What do you think is the patient profile for somebody who can receive CAR T cells?

Yamshon / This question is a bit different for DLBCL than for other types of lymphoma, where there are other good treatment options. Again, alluding to what I said before, where the goal is a cure, I would say that the patients who are truly CAR T ineligible in 2025 for medical reasons are few and far between. We have treated patients with an ejection fraction of 15%, we’ve treated patients on dialysis, and we’ve treated patients with solid organ transplants. In my discussions with community oncologists, they all have my cell phone number, and I always impress upon them [that] even if [they] think it would be insane to give this person CAR T, just send them [to me].

Patel / You raise an important point. What we thought we knew about who we could deliver this to has changed a lot, and that narrative around who’s CAR T ineligible is changing all the time. Nathan, what are some of the things that maybe you see as differences in how we approach it now? Are there patients [for whom] you think, OK, these clinical factors make it difficult to deliver a CAR T-cell therapy?

Denlinger / I agree with Sam wholeheartedly on every point. We’ve done it all at this point and had very good, successful outcomes. It’s shocking to both patients and the outside hospital community providers what we can achieve compared with even just routine chemotherapy or other things that [patients] may never be able to tolerate. We’ve come a [long] away.

I think about2015 to 2017 when the [treatments were] first approved. It was like the last line of therapy, [patients’] disease was exploding, and we threw some CAR T-cell therapy at them, and things went badly from a toxicity standpoint. Now, with the way that we sequence it, we’re getting patients [treated] much earlier. They’re much healthier. The disease usually is much better controlled, so the toxicity is much better tolerated than before, and we’re much better at treating it so aggressively upfront. We know that tocilizumab [Actemra] is not going to hurt the CAR T cells. We know that a bit of steroids isn’t going to hurt the CAR T outcome.

The key is communication with the outside hospitals and those providers, I don’t think that there’s a good amount of understanding about who is eligible [for CAR T-cell therapy]. It seems to be shifting so rapidly. I mean, we’re growing and pushing the envelope. The one thing that we do, [which] is critical, is a social situation. We do still require caregivers [for the patients] for a 30-day period. I’m always shocked that somebody always comes [and says they don’t have a caregiver]. I say CAR T-cell therapy is the chance for a durable remission, [and] suddenly, family comes to help.

Dahiya / The key message that we’ve been trying to communicate with our community colleagues is that if you’re transplant-ineligible, that doesn’t mean you are currently ineligible. The transplant eligibility criteria have been practiced for a long time. The toxins are nonoverlapping. [There are] very different toxins that are there. That’s a key message that we’re trying to convey to our community colleagues. Second, there is no real contraindication. It’s mostly optimization. We’re going to optimize our patients quite well before they can get this therapy. Optimize and debulk them for their lymphoma so they can go through it safely. Those 2 points are important to convey, and it’s a collaborative effort. It’s the collaborative care model that we have to be prepared to champion within our community.

References

1. Abramson JS, Solomon SR, Arnason J, et al. Lisocabtagene maraleucel as second-line therapy for large B-cell lymphoma: primary analysis of the phase 3 TRANSFORM study. Blood. 2023;141(14):1675-1684. doi.10.1182/blood.2022018730
2. Westin JR, Oluwole OO, Kersten MJ; ZUMA-7 Investigators; Kite Members; et al. Survival with axicabtagene ciloleucel in large B-cell lymphoma. N Engl J Med. 2023;389(2):148-157. doi:10.1056/NEJMoa2301665
3. FDA approves lisocatagene maraleucel for second-line treatment of large B-cell lymphoma. News release. FDA. June 24, 2022. Accessed April 8, 2025. https://tinyurl.com/5a93fer4
4. FDA approves axicabtagene ciloleucel for second-line treatment of large B-cell lymphoma. News release. FDA. April 1, 2022. Accessed April 8, 2025. https://tinyurl.com/297rsh4p
5. Charton E, Anota A, Bachy E, et al. Health-related quality of life after axi-cel as a second-line therapy in patients with high-risk relapsed/refractory large B-cell lymphoma who are ineligible for autologous stem cell transplantation: results of the Alycante phase II trial. Blood. 2024;144(suppl 1):4505. doi.10.1182/blood-2024-200614