During a Case-Based Roundtable® event, Matthew A. Lunning, DO, discussed the updated trial data for 2 chimeric antigen receptor T-cell therapies in patients with diffuse large B-cell lymphoma.
CASE SUMMARY
Peers & Perspectives in Oncology: What was the design of the phase 3 ZUMA-7 trial (NCT03391466) investigating axicabtagene ciloleucel (axi-cel; Yescarta)?
Lunning: [In] the ZUMA-7 trial, the key eligibility for the patients was [being] relapsed or refractory to primary therapy within 1 year of completion. They had to declare that they intended to proceed to an autologous stem cell transplant [ASCT]. There was stratification [by response to first-line therapy and second-line agent-adjusted IPI].
One of the main differences between ZUMA-7 and TRANSFORM [NCT03575351] was that there was no bridging allowed as part of this therapy, and so you could only get steroids if you were randomly assigned to the axi-cel arm. Once T cells were apheresed, axi-cel was returned, and fludarabine/cyclophosphamide lymphodepleting chemotherapy was given, then [patients were] infused with axi-cel. In the standard-of-care arm, it could be a platinum-based chemotherapy at the investigator’s discretion for up to 3 cycles. If you had initial disease response, you went on to an ASCT in the control arm. Nonresponders went off therapy and could receive chimeric antigen receptor [CAR] T-cell therapy [next], but there was no crossover built into the trial, and those patients could receive CAR T cell off trial, an investigational CAR T-cell therapy, or defer CAR T-cell therapy and move on to next line of therapy. The primary end point of the study was event-free survival [EFS] by blinded central review. The secondary end points [included progression-free survival (PFS), safety, and patient-reported outcomes].
What are the updated outcomes of this trial?
PFS now with 4-years’ follow-up continues to be statistically significant, with an HR of 0.51 [95% CI, 0.38-0.67].1 The primary end point of EFS was also met with 4-years’ follow-up, showing a statistically significant advantage for the axi-cel arm with an HR of 0.42 [95% CI, 0.33-0.55]. Tails on these [Kaplan-Meier] curves are occurring [after approximately 12 months].
Initially, there was not overall survival [(OS) benefit], but what came out [in 2023] was that with longer follow-up, there was an OS [benefit] seen comparative to the standard-of-care arm with axi-cel with a hazard ratio of 0.73 [95% CI, 0.54-0.98; stratified P = .03]. Again, this was with almost 4 years’ median follow-up.
What was the safety profile of axi-cel in this trial?
In the ZUMA-7 trial, cytokine release syndrome [CRS] was seen in all grades in 92% of the individuals, but only 6% of them experienced grade 3 or greater CRS. With regard to neurologic events, the any-grade [rate] was at 61%, with 21% [having] neurologic events of grade 3 or higher. [In the comparator arm] with ASCT, some neurologic events could happen [20% any grade], but grade 3 or higher events [occurred] in only 1%.
How is this therapy used in patients ineligible for ASCT?
Axi-cel was done as second-line therapy in those patients with relapsed/ refractory large B-cell lymphoma who were felt to be ineligible for an ASCT. There was both investigator-assessed as well as central review panel-assessed [objective response]. The primary end point of the phase 2 ALYCANTE study [NCT04531046] in this transplant-ineligible population was complete response rate, and there was a 71% complete response rate, with an objective response rate [ORR] of 75%, so the majority of the patients were achieving a complete remission.2
CRS was seen previously in ZUMA-7 in 92%.1 In this population with advanced age who were transplant ineligible, it was 86% grade 1/2, with 94% any-grade CRS.2 If you look at ICANS [immune effector cell–associated neurotoxicity syndrome], the any-grade [rate] was at approximately 50%. Grade 1/2 ICANS was 37%, and grade 3/4 was 15%. The median time of onset of CRS was 1.5 days vs median onset of ICANS at 6 days. There were some findings of prolonged grade 3 or higher cytopenias, and in this population, intensive care unit transfers due to CAR T-cell therapy occurred in approximately one-fourth of the patient population.
What positive data are there for lisocabtagene maraleucel [liso-cel; Breyanzi] in patients with DLBCL?
The other study that had a primary positive end point is the TRANSFORM trial. This is with liso-cel. Axi-cel has a CD28 [costimulation], whereas liso-cel has a 4-1BB [costimulation]. Liso-cel has a 1:1 ratio of CD8 and CD4 [components]. In the TRANSFORM trial, you can see the age population had an upper bound of 75, whereas ZUMA-7 had a patient aged 83 in the trial. These patients also had to either have primary refractory disease or relapse within 12 months of completion of their anthracycline-based therapy. Here, there was secondary central nervous system lymphoma allowed. You also had to be deemed eligible for a transplant. Left ventricular ejection fraction [LVEF] of 40% or greater was allowed, and there was no minimum absolute lymphocyte count. The primary end point was EFS, but here the onboarding of the patients was slightly different. Patients went through screening and everybody was leukapheresed and then randomization occurred. If you were sent to the standard-of-care arm, you then received a platinum-based therapy for up to 3 cycles and…if the response was not amenable to going to an ASCT, you could have the option to have your cells that were leukapheresed and manufactured to make liso-cel and then remain on trial and get fludarabine/cyclophosphamide and liso-cel within the study design.
For those patients who were randomly assigned to the experimental arm of liso-cel, bridging therapy was allowed and that bridging therapy could be 1 cycle of a platinum-based therapy of the investigator’s choice. The PET scan still had to be positive after that 1 cycle. They then went on to receive lymphodepleting chemotherapy of fludarabine/cyclophosphamide, and [the dose was] 100 × 106 CAR-positive T cells.
What were the key efficacy outcomes in this trial?
EFS, the primary end point, as well as the PFS, were both positive with an HR of 0.356 [95% CI, 0.243-0.522] and 0.4 [95% CI, 0.261-0.615], respectively, showing a statistically significant advantage for liso-cel in both EFS and PFS.3
In the TRANSFORM trial, there was not an OS advantage. It was in favor of liso-cel [HR, 0.724; 95% CI, 0.443-1.183; P = .0987]. But those individuals who were in the standard-of-care arm who did have an event had the opportunity to have their cells manufactured on study and receive those cells after manufacturing. They remained accounted for from an OS standpoint in the standard-of-care arm. If you do some statistical [analysis to] adjust for that crossover, then based upon this analysis, there would have been a statistically significant HR in favor of liso-cel compared with standard of care at 0.415 [95% CI, 0.251-0.686].
Median EFS was not reached for the liso-cel [whereas] in the standard-of-care arm it was only 2.4 months, with a significant HR [Figure3]. The percentage of complete response in the liso-cel arm was at 74% vs 43% with standard of care, [which was] statistically significant. With regard to secondary end points based upon independent review, there were no surprises here with ORR being more prolonged, a higher percentage of ORR, as well as duration of response.
What was the tolerability seen in this trial?
The toxicity profile here [included] a long list, mostly hematologic toxicities that you would expect with lymphodepleting chemo-therapy with fludarabine/cyclophosphamide, as well as with any platinum-based therapy and chemotherapy, and high-dose chemotherapy and ASCT. The all-grade CRS was noted at 49% but there was nothing too surprising with regard to liso-cel vs standard-of-care toxicities. One of the things I do find interesting is the febrile neutropenia rate in those patients who received platinum-based chemotherapy and ASCT was only 26%. I often see higher febrile neutropenia rates after ASCT, so that’s a fairly interesting thing in a prospective study.
[For] adverse events [AEs] of special interest: looking at CRS, any grade with liso-cel was 49%, with the vast majority being grade 1 or grade 2 at 48%, with only 1 event of grade 3 or higher CRS. In regard to neurologic events, any grade was at 11%, with 8% grade 1/grade 2, and only 4 individuals having grade 3, with no grade 4 or 5 events. The onset is later than with axi-cel numerically with a median of 5 days’ onset of CRS and 11 days of onset with ICANS. Looking at cytopenias, neutropenia can be an issue. Grade 3 or higher neutropenia at study day 64 was seen in 37% of the liso-cel arm vs 2% of the standard-of-care arm.
What were the outcomes of liso-cel in patients who were not transplant eligible?
The PILOT study [NCT03483103] investigated liso-cel for those felt to not be transplant eligible [based on] 1 or more of the following criteria: age over 70, ECOG performance status of 2, impaired organ function as defined by diffusion capacity of the lungs for carbon monoxide less than 60%, LVEF greater than 40%, creatinine clearance greater than 30 mL/min but less than 60 mL/min or aspartate transaminase/alanine transaminase greater than 2 times but less than 5 times their limits of normal. Looking at this patient population, about half of them had primary refractory disease, 21% had relapsed within 1 year of therapy, and only 25% of the patients had relapsed greater than 12 months post–first-line therapy.
Looking at the 74 subjects who were accrued and 61 who were infused, the median age was 74, and approximately one-fourth of them had an ECOG performance status of 2.4 This study had a short follow-up at the time of presentation at 12 months, with an ORR of 80%.
There were AEs in this population, as were seen in the ALYCANTE study. Neutropenia was seen in approximately half of the patients and thrombocytopenia in approximately 20% of the patients. Looking at neurologic AEs of interest with CAR T cells, CRS occurred in 38% with only 1 case of grade 3 or higher CRS, and with regard to neurologic events, there were 31% all-grade with 3 AEs that were grade 3. There were no grade 4 AEs or deaths.