New CAR T-Cell Therapy Approvals Diversify Treatment Options in Large B-Cell Lymphoma

The potential for outpatient treatment and how quickly a patient can undergo apheresis and subsequent dosing with chimeric antigen receptor (CAR) T-cell therapy may inform treatment selection for axicabtagene ciloleucel (axi-cel; Yescarta), tisagenlecleucel (tisa-cel; Kymriah), and lisocabtagene maraleucel (liso-cel; Breyanzi) in relapsed/ refractory large B-cell lymphoma (LBCL), in addition to efficacy, toxicity, and real-world data, said Matthew Lunning, DO.

“The field is exploding with cellular therapies in hematologic malignancies, namely lymphomas, because the field is so mature,” said Lunning.

Additionally, a key research venture that is under way in LBCL is to shorten the time from when a patient is determined eligible and ready for CAR T-cell therapy to when they receive the infusion. Lunning, who coins this time period brain-to-vein, said, “In the commercial environment, the brain-to-vein time is starting often weeks before we are even getting to apheresis or starting that vein-to-vein clock. [It is important to] try to shorten that time frame from the brain-to-vein, where every day matters.”

In an interview with OncLive^® during an Institutional Perspectives in Cancer webinar on hematologic malignancies, Lunning, an assistant professor of internal medicine in the Division of Oncology and Hematology, College of Medicine, University of Nebraska Medical Center, discussed the findings from pivotal CAR T-cell therapy trials in LBCL, the potential to utilize these agents in the outpatient setting, how off-the-shelf allogeneic products could further transform outcomes, and the importance of shortening brain-to-vein time for this patient population.

OncLive®: What stood out to you from the 4-year analysis of the ZUMA-1 trial? 

Lunning: [The investigators of ZUMA-1 (NCT02348216)] reported 4-year follow-up data at the 2020 American Society of Hematology Annual Meeting and Exposition. What was interesting about that [data set] was that [we saw an] ongoing tail for the overall survival [OS] curve [with axi-cel]. The 4-year OS [rate] was 44% [with axi-cel]. No new signals of long-term toxicity [were observed] in the cohort with relapsed/refractory diffuse large B-cell lymphoma [DLBCL] that remains in remission. We have to [continue to] watch these interesting data mature to see whether the [benefit is sustained with more follow-up].

In [ZUMA-1], the other [intriguing factor] was that B-cell recovery was being [observed in] patients who are longterm survivors.

How have the data from ZUMA-5 added to existing knowledge of antigen escape mechanisms in patients with LBCL?

The ZUMA-5 study [NCT03105336] evaluated axi-cel in relapsed/refractory follicular lymphoma and marginal zone lymphoma [MZL].

When we look at ORRs in the population who was [deriving] durability, there seemed to be a striking difference between the follicular lymphoma cohort and the MZL cohort. Although [some people have said] the follow-up hasn’t been long enough in the MZL cohort [to comment on the agent’s activity in that subset, a lot of patients experienced] early progression events that were out of character for what I would expect from patients with MZL, even though the OS curves continued to look [good] in that population. Patients could respond to other lines of therapy post CAR T-cell therapy, where it has become difficult to treat patients with MZL or DLBCL. 

In my opinion, this brings to light whether there is a separate escape mechanism in patients with MZL that we need to look into. What is it about MZL that patients aren’t responding to, at least with regard to CD19-directed therapies and axi-cel? I am not aware of data in MZL with other CAR T-cell constructs, such as tisa-cel or liso-cel.

Do you believe the initial responses that were seen with brexucabtagene autoleucel (Tecartus) in the ZUMA-2 trial will translate into durable remissions in patients with mantle cell lymphoma (MCL)?

[ZUMA-2 (NCT02601313) evaluated] patients with 1 to 5 prior lines of therapy who had previously been exposed to a BTK inhibitor. These patients could be bridged with steroids or a BTK inhibitor. Patients with relapsed/refractory MCL who have failed BTK inhibitors [and receive] single-agent venetoclax [Venclexta] or lenalidomide [Revlimid] may have a response, but those responses are not always durable. The early results we saw from the ZUMA-2 trial [showed] a very high overall response rate [ORR], as well as a respectable complete response rate around 50%. That was unique in this patient population.

The question comes down to this: Are we going to see the durability that we are seeing in the ZUMA-1 patient population in the ZUMA-2 population? I acknowledge that [brexucabtagene autoleucel utilizes] a slightly different manufacturing construct [compared with axi-cel]. In ZUMA-2, there were toxicities that we’ve seen with the CD28 costimulatory construct with regard to cytokine release syndrome [CRS] and neurotoxicity.

Do the data from the ELARA trial provide a clear picture of where tisa-cel could be sequenced in treatment for patients with relapsed/refractory follicular lymphoma?

The ELARA study [NCT03568461] demonstrated efficacy with single-agent CD19-directed tisa-cel using a 4-1BB costimulatory molecule. With tisa-cel, a lower dose of cyclophosphamide and fludarabine with this construct was part of lymphodepleting chemotherapy. There was efficacy seen, but also, they saw potentially less CRS, less neurotoxicity, and less high-grade CRS and neurotoxicity. [Treatment selection is] going to come [down] to the durability [of response] in this patient population and the selection of whom we are going to [treat with tisa-cel] in the relapsed/refractory follicular lymphoma population.

Right now, if axi-cel* and tisa-cel were to get labels [in follicular lymphoma], they would sit in the third-line setting. Lenalidomide and rituximab [Rituxan] or lenalidomide and a CD20[-directed] monoclonal antibody really captured the second-line setting after chemoimmunotherapy in the f irst-line setting. It is going to be interesting to see whether CAR T-cell therapy will challenge that space, [but] again, it is going to depend upon what the labels state, if [the products] get labels.

[Then we need to still sort out] what [patient] population to offer [CAR T-cell therapy to]. We also still have transplant. Some [physicians] would like to take patients in second remission to an autologous stem cell transplant, so [treatment sequencing] is still an open question.

Do the findings from the TRANSCEND-OUTREACH-007 trial support the use of CAR T-cell therapy in the outpatient setting?

[TRANSCEND-OUTREACH-007 (NCT03744676) evaluated] liso-cel, which is our third CAR T-cell construct in the late stages of development. That [study was conducted] at a nonuniversity site, but these were fairly large institutions or hospitals that were doing phase 1 studies or had autologous transplant programs. Although [the product] was unique [to test in] a nonuniversity setting, there were caveats. A significant proportion of patients were treated as outpatients, which is something that we have seen data on with liso-cel in the TRANSCEND-NHL-001 trial [NCT02631044)] and I believe [with] tisa-cel. Those 2 constructs, which are both 4-1BB costimulatory [molecules], appeared to have the potential to be delivered as outpatient [therapy].

The data from the TRANSCEND-OUTREACH-007 study [answered the question] “What percentage of patients could remain outpatient for the entirety of their [treatment with] CAR T-cell therapy?” It was a significant proportion. If patients were coming [into the hospital], it was for adverse effects related to CAR T-cell therapy around day 4 after the CAR T-cell therapy [infusion. However, about] 1% of the [entire population] required intensive care unit [ICU]–level care.

That is an important caveat if we are going to be doing outpatient CAR T-cell therapy. If patients require [hospital] admissions, how quickly do they get sick? Do we have time to see them in the outpatient clinic or in an outpatient treatment center? Do we have to [bring them into] an inpatient setting to treat their CRS or neurologic event? Are they going straight from the outpatient setting to the ICU? If a high percentage [of patients were going] from the outpatient [setting] to the ICU, that would be concerning. 

The data continue to support the outpatient [treatment] capabilities of some of our CAR T-cell therapies, namely liso-cel and tisa-cel.

The pending biologics license application for liso-cel in LBCL was put on hold by the FDA. Is that cause for concern?*

Nobody knows what is going on with regard to the COVID-19 [coronavirus disease 2019] environment and how the FDA is determining when to approve something.

It would be nice to have another construct in the market to see the real-world data and how [liso-cel] will behave. Having a third construct in the market means that in some patients we may choose [liso-cel] over [axi-cel or tisa-cel]. It adds another wrinkle to the discussion, but I’m not concerned about the delay in that regard.

Regarding real-world data, how helpful are those data sets in determining which patient populations these products are best served in?

Real-world data are incredibly important, but we haven’t been able to sort out when to start [recording] real-world data. In my mind, CAR T-cell therapy often has started in an intention-to-treat [population] when patients [undergo apheresis] for their treatment, or this concept called vein-to-vein or apheresis-to-infusion time. In my opinion, [CAR T-cell therapy] starts earlier, in a period I call the brain-to-vein time, which means that [the patient and I agree] that their disease is in the right place and the [benefit outweighs the risk] to proceed with CAR T-cell therapy. The clock starts then. That is where my denominator starts in regard to getting that patient through their CAR T-cell therapy.

[The brain-to-vein time] is varied, [depending on] the diseases and disease states that we are having discussions about with CAR T-cell therapy. We are talking about multiply refractory LBCL, as well as multiply relapsed/ refractory MCL. 

Shaving off brain-to-vein time means days shaved off in the vein-to-vein time with regard to lymphodepleting therapies. Shrinking that all down is important because the disease doesn’t care whether it is a Friday or a holiday or whether we are trying to figure out how to get a patient to their apheresis. That is something that we have been focusing on to generate our own institutional data to see how we can continue to serve our patients better.

How could allogeneic CAR T-cell products further the cellular therapy paradigm and potentially shorten the brain-to-vein time for patients?

A lot of the off-the-shelf products have shown encouraging early data, including CD19/CD20-targeted and CD19/ CD22-targeted products, off-the-shelf allogeneic CAR T-cell therapies, and natural killer CAR-T cell products.

Talking about the off-the-shelf concept, we have to keep in mind what that means from a vein-to-vein time, as well as a brain-to-vein time. If we can shorten that brain-to-vein time with regard to getting approvals [for] CAR T-cell therapy, the vein-to-vein time is much shorter.

In theory, [off-the-shelf products] may work out better from the patient’s standpoint, but we are hopeful that these products have equal efficacy and no excess toxicity; that is yet to be seen. 

The other aspect is we are getting a lot of constructs that are putting out data, but no head-to-head CAR T-cell study has been done in a homogeneous patient population that is randomized to one [product] vs another. It is confusing [to determine] which patients [we should] give which construct. [We are relying on] the [single-construct data], the product that one is most comfortable with, or the product that the institution is most comfortable with. A lot of processes are involved in doing CAR T-cell therapy, so we all get comfortable with [a particular] process that we can do over and over again. I am not going to say it becomes habitual, but it can affect how we think about which [construct] to use while keeping in mind what we feel is best for that individual at that time based on available data. 

It will be interesting to see how these CAR T-cell therapies grow up as they compete in the space. Something that we have to think about is [using] CAR T after CAR T. A lot of trials may exclude [patients who have had] prior CD19-directed CAR T-cell therapy, but a lot of patients are getting commercially available CAR T-cell therapy. Do these CAR T cells have different mechanisms of action or different targets? Can they be efficacious in [the post–CAR T-cell therapy] environment? I hope we will start to see some data emerge in the clinical trial realm [to answer those questions].

Will we ever see a head-to-head CAR T-cell therapy trial in the future?

We have 3 trials ongoing in the second-line CAR T-cell therapy space: BELINDA [NCT03570892], TRANSFORM [NCT03575351], and ZUMA-7 [NCT03391466]. These studies are all going up against transplant.

On the back end of that, if all 3 studies are positive, then we will have 3 CAR T-cell constructs that could potentially be approved in very similar spaces, like what is happening in the [third-line setting] if liso-cel gets a label. We would then have 3 CAR T-cell therapies relatively in the same space within the same patient population. How do we decide which one we should move forward with?

That is going to be [based on] patient particularities and institutional comfort with the construct. I don’t expect that the regulatory body [will approve] a 3[-arm] trial where CAR T-cell therapy A goes against B, which goes against C in the same homogeneous population in the second-line setting of relapsed/ refractory LBCL. It would be quite a large study, but it would be interesting, a winner-take-all kind of approach. As a trial purist, it would be something that would be entertaining to think about, but, in my mind, it’s a thought experiment rather than a reality.

*Editor’s note: This interview was conducted prior to the FDA approvals of axi-cel and liso-cel in relapsed/refractory indolent follicular lymphoma and LBCL, respectively.