Navigating the journey to develop safer CARs
The development of chimeric antigen receptor T-cell therapy has revolutionized treatment for patients with advanced blood cancers.
However, currently available therapies have significant toxicities that require nearly all patients to remain as inpatients at the treating center, typically for 2 weeks or more. In addition, FDA post-monitoring requires that patients stay within a certain proximity of the treating center for up to 4 weeks after infusion.
Given these restrictions, researchers and clinicians continue to search for strategies to improve the safety of CAR T-cell therapy without reducing efficacy.
Cell Therapy Next compiled the latest CAR-T research from ASCO20 Virtual Scientific Program and asked experts in the field what is being done to improve the safety of CAR T-cell therapy, both in therapy development and modifications to post-infusion care.
Generations of CARs: a primer
The first clinically effective CARs were demonstrated by Kalos and colleagues and are commonly referred to as second-generation CARs, Caron A. Jacobson, MD, assistant professor of medicine at Harvard Medical School and medical director of the immune effector cell therapy program at Dana-Farber Cancer Institute, told Cell Therapy Next.
First-generation CAR T cells contained a CD3-zeta co-stimulatory domain but were ineffective in clinical practice, she said. FDA-approved CAR-Ts are second generation and have a second co-stimulatory domain.
Modifications being applied to newer-generation CAR T cells to improve safety involve exploring different ways to promote or control CAR T-cell activation. These can be considered next-generation CARs, Jacobson said.
Among the modifications soon to be or currently being tested, Jacobson highlighted those requiring two activation signals (antigens) before a CAR responds. Other groups are exploring ways to modulate the CAR response, much like using an on/off switch.
Another new construct experiments with signaling and activation and uses the patient’s native T-cell receptor for signaling and attachment to the T cell. The antibody fragment still binds to the tumor antigen to achieve tumor specificity, but it then binds to the T-cell receptor and activates T-cell growth through the receptor, similar to what would happen when human T cells launch a normal response against an infection.
Some investigators are trying to control where the CAR gene gets inserted in the genome, Jacobson said.
“They are trying to put it in the place of the T-cell receptor gene so that the CAR gene is under the control of the T-cell receptor promoter, which is regulated based on the activity of the T cells,” she said. “This allows researchers to modulate the number of CAR gene copies based on the activity level of the T cell.”
Next-generation constructs seek to “control the CAR and its proliferation,” according to Sattva S. Neelapu, MD, professor and deputy chair of the department of lymphoma/myeloma in the division of cancer medicine at The University of Texas MD Anderson Cancer Center.
“There is no shortage of strategies being attempted by many groups to improve on the overall safety of CAR T-cell therapy,” Neelapu told Cell Therapy Next.
In addition to the next-generation modifications Jacobson outlined, different approaches to post-infusion clinical care are being attempted to further enhance the safety of CAR T-cell therapy, Neelapu said. These include combinational strategies with other drugs, such as the interleukin-1 receptor antagonist anakinra (Kineret, Sobi), the anti-GM-CSF antibody lenzilumab (Humanigen), and prophylactic steroids to prevent severe cytokine release syndrome (CRS).
When asked which studies at ASCO stood out to him from a safety standpoint, Neelapu pointed to ZUMA-5, for which Jacobson was one of the primary investigators.
Interim study results showed axicabtagene ciloleucel (Yescarta, Kite/Gilead) induced a 93% response rate among patients with relapsed or refractory indolent non-Hodgkin lymphoma — which includes follicular lymphoma and marginal zone lymphoma. The responses appeared durable, and the treatment had an acceptable safety profile.
“The efficacy of CAR T-cell therapy in other types of CD19-positive diseases has been transformative,” Jacobson said. “It makes sense to expand the indications to other CD19-positive B-cell malignancies.”
This year’s ASCO program marked the first time data from the ZUMA-5 trial had been presented.
The efficacy analysis included 96 patients (mean age 63 years; range, 34-79; 51% women) with median follow-up of 15.3 months (range, 1.9-28.8). Eighty of the patients had follicular lymphoma and 16 had marginal zone lymphoma.
All patients had a median three (range, 2-9) prior lines of therapy, and approximately half (52%) had stage IV disease. Nearly three-quarters (73%) of patients were refractory to their previous line of treatment.
Efficacy results showed an objective response rate of 93% (95% CI, 86-97), with a complete response rate of 80% (95% CI%, 71-88).
Median duration of response was 20.8 months for all patients. Sixty-eight percent of patients with follicular lymphoma had an ongoing response to therapy at data cutoff.
Median PFS was 23.5 months (95% CI, 22.8 to not reached) for all patients, and median OS was not reached. One-year OS for all patients was 94.3% (95% CI, 86.8-97.6).
The safety analysis (n = 140) showed nearly all patients (99%) had at least one treatment-related adverse event. Eighty-five percent of patients experienced grade 3 or higher adverse events, the most common of which were neutropenia (34%), decreased neutrophil count (34%) and anemia (22%).
Two patients died during the study; one had CRS-related multisystem organ failure that was determined to be related to axicabtagene ciloleucel treatment.
Most patients (79%) had some form of CRS; however, only 8% of patients had grade 3 or higher CRS. Median time to onset was 4 days (range, 1-15), with a median duration of 6 days (range, 1-27).
Fifty-eight percent of patients experienced neurotoxicity, with 17% of cases being grade 3 or higher. Median time to onset was 7 days (range, 1-177), with a median duration of 14 days (range, 1-452).
Jacobson highlighted the relatively low rate of CRS in the study, especially serious cases. She told Cell Therapy Next that the safety profile “could potentially allow for outpatient dosing in the future.”
“Yes, it can be done as an outpatient, especially in patients who have a lower tumor burden,” he told Cell Therapy Next. “These patients can be given CAR-T as an outpatient and admitted only if ... they develop more severe symptoms of CRS or neurotoxicity.”
Jacobsen said she is confident the FDA will approve axicabtagene ciloleucel for the indications examined in ZUMA-5.
“The data we have to date on this would surpass almost any other therapy we have available in the third line or beyond for these patients,” Jacobson said.
Patients at high risk, such as those in ZUMA-5, typically lack effective treatment options, which makes the response rates in this study “very impressive,” according to Samantha M. Jaglowski, MD, MPH, associate professor of hematology and medical director of the Blood and Marrow Transplant Program at The Ohio State University Comprehensive Cancer Center.
Jaglowski pointed to the safety profile of axicabtagene ciloleucel, also known as axi-cel, in this patient population as especially promising, including the low incidence of grade 3 CRS and neurological events. She said both were “relatively low compared with what has been reported for both large B-cell and mantle cell lymphoma.”
Jaglowski said axi-cel likely will receive FDA approval for this indication based on its high response rates plus prolonged durability in a relatively refractory patient population.
“Combine this with its relatively good safety profile and track record with previously approved indications, and I think it is a matter of time before axicabtagene ciloleucel is approved for follicular and marginal zone lymphoma,” she said.
An allogeneic CAR T-cell therapy plus an anti-CD52 monoclonal antibody appeared safe for certain patients with large B-cell or follicular lymphoma, according to data from the phase 1 ALPHA trial.
The combination therapy showed clinically relevant antitumor activity in the study, for which Neelapu was one of the primary investigators.
For allogeneic CAR T-cell therapies to be successful, they must be engineered to reduce the likelihood of graft-versus-host disease, Neelapu said.
The second-generation, anti-CD19 CAR T-cell product (ALLO-501; Allogene Therapeutics, Servier) used in the ALPHA trial was genetically modified to knock out T-cell receptor (TCR)-alpha gene expression and reduce incidence of GVHD.
Prior to infusion, trial participants received fludarabine dosed at 90 mg/m² and cyclophosphamide dosed at 900 mg/m² for lymphodepletion, followed by ALLO-647 (Allogene Therapeutics, Servier) — an investigational monoclonal antibody that targets CD52 on most T cells.
“By giving ALLO-647, we deplete the hosts’ T cells without affecting the donor T cells in ALLO-501 because it lacks the CD52 gene,” Neelapu told Cell Therapy Next. “This leads to better persistence of the donor T cells, better expansion of the cells, and improved clinical outcomes and durability.”
The open-label, dose-escalation ALPHA trial enrolled 22 adults (median age 63 years; range, 34-73; 77% men), with relapsed or refractory large B-cell or follicular lymphoma who had received at least two previous lines of therapy. Patients who received previous anti-CD19 cell therapies were eligible for the trial.
After lymphodepletion, patients received ALLO-647 dosed at 39 mg or 90 mg. They then received an IV infusion of ALLO-501 at one of three dose levels — 40 × 106 CAR T cells/kg, 120 × 106 CAR T cells/kg or 360 × 106 CAR T cells/kg — 5 to 7 days after lymphodepletion.
Safety and dose-limiting toxicity of the regimen served as the study’s primary endpoints.
Median follow-up was 3.8 months (range, 0.7-6.1).
Neelapu presented data on 19 patients in the efficacy analysis and 22 patients in the safety analysis.
The safety analysis showed the most common treatment-related adverse event was neutropenia (68% overall, 32% grade 3, 32% grade 4). Seven patients (32%) experienced CRS, including one grade 3 case.
No patients showed evidence of GVHD or symptoms resembling immune effector cell-associated neurotoxicity syndrome (ICANS).
The rate of CRS in the ALPHA trial is comparable to or lower than that seen with autologous CAR T-cell products, Neelapu said.
Neelapu warned the process of depleting the hosts’ T cells in this trial leaves patients more susceptible to subsequent infections. Half of patients in the safety analysis experienced infections, but these were easily resolved through monitoring and medication Neelapu said.
The efficacy analysis showed a 63% (95% CI, 38-84) ORR and a 37% (95% CI, 16-62) complete response rate. Nine of the 12 patients who responded remained in response at data cutoff May 11.
“The complete and overall response rates we have seen are in the same ballpark as what we have seen with autologous CAR-T cell therapy,” he said. “What we don’t know at this time is whether its durability will be as lasting as autologous CAR T-cell products.”
Jaglowski said that one of the drawbacks of CAR T-cell therapy is the time it takes to produce the product — potentially several weeks — during which time a patient’s disease may progress.
“They may no longer have the time to benefit from the CAR T cells once they are ready,” she said.
The study by Neelapu and colleagues — although in its early stage and having treated just a few patients — is “very exciting,” Jaglowski added.
“It is encouraging to see this treatment regimen is relatively safe and induces a good response rate compared with historical numbers,” she said.
The rate of grade 4 or higher CRS was lower in the ALPHA trial than typically seen with autologous CARs, Jaglowski said. The same was true for the rate of neurotoxicity.
“With response rates comparable to currently available treatments, the value of this approach remains to be determined based on the accumulation of more efficacy data,” she said.
JNJ-4528 (Janssen Pharmaceuticals) is a CAR T-cell therapy that targets the B-cell maturation antigen. Results of the phase 1b/phase 2 CARTITUDE-1 trial showed that JNJ-4528 induced early, deep and durable responses among patients with relapsed or refractory multiple myeloma.
In addition, the low rate of high-grade CRS observed in the study could encourage use of this investigational autologous cell therapy in an outpatient setting, according to the researchers.
“The results are pretty impressive,” Jesus G. Berdeja, MD, director of myeloma research at Sarah Cannon Research Institute, told Cell Therapy Next.
“These results are strong enough that they make JNJ-4528 a very strong candidate to be one of the first BCMA-directed CAR-T products approved by the FDA for multiple myeloma,” Berdeja said. “I think it will eventually revolutionize the treatment of patients with relapsed or refractory disease.”
The CARTITUDE-1 trial included 29 adults (median age, 60 years; range, 50-75; 52% women) with multiple myeloma per International Myeloma Working Group criteria. All patients had received at least three previous lines of therapy (median, 5; range, 3-18), and 25 (86%) were triple-refractory to a proteasome inhibitor, immunomodulatory drug and anti-CD38 antibody.
Updated results, with median follow-up of 11.5 months (range, 3-17), showed an overall response rate of 100%. Twenty-five patients (86%) had stringent complete responses, three patients had very good partial responses and one patient had a partial response to therapy.
“The reason there is so much excitement around CAR-T therapies is that we are seeing response rates that are similar to front-line therapies, but the question remains as to how durable these responses will be,” Berdeja said.
These updated data from CARTITUDE-1 provide further evidence of the durability of responses seen with the CAR construct used in JNJ-4528, he added.
The study’s safety results showed 93% of patients had some form of CRS but only two patients had grade 3 or greater CRS. Only three patients experienced neurotoxicity consistent with ICANS, including one grade 3 or greater case.
One patient died of CRS and another died of acute myeloid leukemia not related to study treatment.
Hematologic adverse events included grade 3 or greater neutropenia, experienced by all 29 patients, and thrombocytopenia, which occurred among 86% of patients. Twenty patients (69%) experienced grade 3 or greater thrombocytopenia.
Compared with many other BCMA-directed CAR T-cell therapies, CRS developed more slowly with JNJ-4528, Berdeja said. Patients treated with typical CAR-T products show symptoms of CRS within a day or two of infusion, he added.
Because of the longer median time to CRS symptoms seen in this study, Berdeja said there is a possibility that this therapy could be given in an outpatient setting, and that only those patients who progress to show symptoms of CRS would need to be hospitalized for post-infusion observation and treatment.
“This could possibly make the treatment more widely accessible for patients and clinicians,” he said.
“The development of CAR-T products will eventually change how we treat myeloma, and I believe that these highly effective treatments will supplant some of the continuous therapies that we are currently giving to our patients,” Berdeja said. “The hope is that someday these will become curative therapies.”
Although all patients in the study developed CRS, most cases were mild. That bodes well for the overall safety profile of JNJ-4528, according to Srinivas Devarakonda, MD, a hematologist and assistant professor at The Ohio State University Comprehensive Cancer Center and Arthur G. James Cancer Hospital and Richard J. Solove Research Institute.
Devarakonda also pointed out that one patient died of CRS in the study so far. Conversely, very few patients developed neurotoxicity in the study, with no high-grade cases.
“Treatment with JNJ-4528 achieved impressive results with early, deep and durable responses in a cohort of patients with heavily treated multiple myeloma,” Devarakonda told Cell Therapy Next. “The therapy showed a relatively safe toxicity profile.”
He noted another positive result from the study: Infectious complications were infrequent, and no deaths were reported secondary to infection.
CAR-T’s upward safety trend
Neelapu said the trend toward safer CAR T-cell therapy will make outpatient dosing more feasible but, for the meantime, centers like his will require that most patients remain in the hospital for at least the first week so that clinical staff can monitor for development of severe side effects.
Generally speaking, he added, MD Anderson looks to discharge these patients once their symptoms have resolved, but patients are still obligated to remain within 2 hours of the center for up to 4 weeks after infusion, per the FDA’s Risk Evaluation and Mitigation Strategies requirements.
Jacobson is aware of some centers doing outpatient CAR-T where feasible, but this is not the case at Dana-Farber, which has requirements similar to those of MD Anderson.
Nevertheless, she said outpatient CAR-T could be feasible for certain patients, such as those in the ZUMA-5 trial, and that overall the field has seen an improvement in safety signals as clinicians become more accustomed to providing earlier intervention against treatment-related toxicities.
“The safety profile is getting better, and clinicians are becoming more comfortable with the toxicities and how to manage them,” Jacobson said. “Although the overall safety of providing CAR-T has improved, we still have not eliminated all the risks, and so safety issues remain part of the treatment that clinicians need to remain vigilant of and prepare for.”
- Kalos M, et al. Sci Transl Med. 2011;doi:10.1126/scitranslmed.3002842.
- The following abstracts were presented at ASCO20 Virtual Scientific Program; May 29-31, 2020:
- Berdeja JG, et al. Abstract 8505.
- Jacobson CA, et al. Abstract 8008.
- Neelapu SS, et al. Abstract 8002.
For more information:
- Jesus G. Berdeja, MD, can be reached at email@example.com.
- Srinivas Devarakonda, MD, can be reached The James Cancer Hospital and Solove Research Institute, 460 West 10th Ave. Columbus, Ohio 43210.
- Caron A. Jacobson, MD, can be reached at firstname.lastname@example.org.
- Samantha M. Jaglowski, MD, MPH, can be reached at email@example.com.
- Sattva S. Neelapu, MD, can be reached at firstname.lastname@example.org.