December 5, 2020 - Axicabtagene ciloleucel demonstrated high rates of durable responses in patients with indolent non-Hodgkin lymphoma.
Axicabtagene ciloleucel (axi-cel; Yescarta) demonstrated high rates of durable responses in patients with indolent non-Hodgkin lymphoma (NHL), according to extended follow-up results of the ZUMA-5 trial that were presented during the 2020 ASH Annual Meeting.1
Results showed that, at a median follow-up of 17.5 months, the objective response rate (ORR) by independent radiology review committee (IRRC) was 92% (95% CI, 85%-97%), along with a complete response (CR) rate of 76% (95% CI, 67%-84%). The partial response (PR) rate was 16%; the stable disease (SD) was 3% and 5% of patients were undefined.
In patients with follicular lymphoma, the ORR was 94%, with a CR rate of 80%, and a PR rate of 14%. Four percent of patients had SD and 2% were undefined. In those with marginal zone lymphoma (MZL), the ORR was 85%, with a CR rate of 60%, and a PR rate of 25%; 0 patients had SD and 15% of patients were undefined.
“Axi-cel may be a highly promising therapeutic approach for patients with relapsed/refractory indolent NHL,” said lead study author Caron A. Jacobson, MD, MMSc, medical director of the Immune Effector Cell Therapy Program, and a senior physician and assistant professor of medicine, Harvard Medical School, Dana-Farber Cancer Institute. “Given the long natural history of these diseases, safety is of paramount importance. The safety profile was manageable and reversible and appeared to be at least similar to that of axi-cel in aggressive lymphomas. This may have implications for the possibility of outpatient therapy, and evaluation of this is planned.”
Patients with advanced-stage indolent NHL, including follicular lymphoma and MZL, frequently relapse on standard treatment; remission durations for these patients shorten with subsequent lines of therapy.
Axi-cel is an autologous anti-CD19 CAR T-cell therapy that is currently approved for the treatment of adult patients with relapsed/refractory large B-cell lymphoma following 2 or more lines of systemic treatment. The regulatory decision was based on findings from the ZUMA-1 trial, in which axi-cel elicited an ORR of 82% in the modified intent-to-treat population, along with a 58% CR rate.2
In the multicenter, single-arm, phase 2 ZUMA-5 trial (NCT03105336), investigators sought to evaluate axi-cel in 151 patients with relapsed/refractory indolent NHL, either follicular lymphoma or MZL, following 2 or more lines of therapy. Five patients could not be treated due to transformation to diffuse large B-cell lymphoma (DLBCL) via pretreatment biopsy (n = 1), ineligibility (n = 3), and death prior to infusion (n = 1). Therefore, a total 146 patients were treated as of the data cutoff and had either follicular lymphoma (n = 124) or MZL (n = 22).
To be eligible for enrollment, patients needed to have relapsed/refractory follicular lymphoma in grades 1 to 3A or MZL that was nodal or extranodal. Patients must have received 2 or more prior lines of treatment that included an anti-CD20 monoclonal antibody plus an alkylating agent.
Conditioning regimens included 30 mg/m2 of intravenous (IV) fludarabine and 500 mg/m2 of IV cyclophosphamide on days –5, –4, and –3. Axi-cel was administered at 2 x 106 CAR+ cells/kg.
The primary end point of the trial was ORR, which was IRRC-assessed per Lugano classification. Secondary end points included IRRC-assessed CR rate, investigator-assessed ORR, duration of response (DOR), progression-free survival (PFS), overall survival (OS), adverse effects (AEs), and CAR T-cell and cytokine levels. The MZL cohort was exploratory, and all analyses for this cohort were descriptive in nature, Jacobson noted.
The efficacy analyses included 104 patients: those with follicular lymphoma who had at least 1 year of follow-up (n = 84), and patients with MZL with at least 4 weeks of follow-up (n = 20). The safety analysis included all 146 treated patients.
As of March 12, 2020, the median follow-up for the efficacy analysis was 17.5 months (range, 1.4-31.6). The median follow-up for the safety analysis was 15.1 months (range, 0.5-31.6). Axi-cel was manufactured for all 151 enrolled patients and was delivered to the study site a median 17 days following leukapheresis.
Overall, the median age was 61 years (range, 34-79), and 35% of patients were 65 years or older. Fifty-seven patients were male, and 38% had an ECOG performance status of 1. Most patients (86%) had stage III to IV disease, but less than half (47%) had a Follicular Lymphoma International Prognostic Index (FLIPI) score of 3 or lower, and 49% had high tumor bulk based on Groupe d'Etude des Lymphomes Folliculaire criteria.
The median number of prior therapies was 3 (range, 1-10), but 64% received 3 or more treatments; 29% had previously received a PI3K inhibitor. Sixty-eight percent of patients had refractory disease, and 55% had disease progression for more than 2 years from the first anti-CD20 monoclonal antibody–containing therapy. Twenty-three percent of patients underwent prior stem cell transplantation.
Results also showed that the median time to first response was 1 month (range, 0.8-3.1). Additionally, of 25 patients with follicular lymphoma who initially achieved a PR, 52% subsequently converted to a CR after a median 2.2 months (range, 1.9-11.2).
The response rate was consistent across key subgroups analyzed on the study, including age, sex, ECOG performance status, FLIPI score, high tumor burden, number of prior therapies, prior PI3K inhibition, prior autologous SCT, relapsed/refractory disease, progression of follicular lymphoma within 24 months from initiating first anti-CD20 monoclonal antibody–containing therapy, CD19 status, and corticosteroid and/or tocilizumab (Actemra) use.
At a median follow-up of 17.5 months, the estimated median DOR was not reached in the overall population, and the 12-month DOR rate was 71.7% (95% CI, 60.7-80.1).
In the follicular lymphoma cohort, which had a median follow-up of 12.1 months, the median DOR was not estimated and the 1-year DOR rate was 77.0%. Moreover, 64% percent of patients with follicular lymphoma had an ongoing response at data cutoff. Additionally, responses were ongoing in 78% of patients who achieved a CR and in 17% of patients with a PR.
In the MZL cohort, the median DOR was 10.6 months and the 1-year DOR rate was not estimated.
When the median DOR was calculated by best response, patients with follicular lymphoma who had a CR and PR had a median DOR of not estimated (NE) and 2.8 months, respectively; the 1-year DOR rates were 87.0% and 13.6%, respectively.
In the patients with MZL who achieved a CR and PR, the median DOR was 10.6 months and 8.1 months, respectively; the 1-year DOR rates were NE and 0%, respectively. However, Jacobson noted that the follow-up for the MZL cohort is much more limited, making these DOR data immature.
Moreover, further findings showed that the median PFS was NE (95% CI, 23.5–NE), and the 1-year PFS rate was 73.7%. In patients with follicular lymphoma, the median PFS was NE and the 12-month PS rate was 77.5%; in patients with MZL, the median PFS was 11.8 months and the 1-year PFS rate was 45.1%.
The median OS in all patients was NE, and the 1 year-OS rate was 92.9%. The median OS was NE and the 1-year OS rate was 92.8% in patients with follicular lymphoma; in patients with MZL, the median OS was NE and the 1-year OS rate was 92.9%.
Regarding safety, grade 3 or higher adverse effects occurred in 86% of patients, the most common of which were cytopenias (70%) and infections (16%). Three deaths were reported: 1 was related to axi-cel due to multisystem organ failure in the context of cytokine release syndrome (CRS; n = 1, day 7), while the other 2 were unrelated to study treatment and were due to aortic dissection (n = 1, day 399) and coccidioidomycosis (n = 1, day 327).
Overall, 82% of patients experienced any-grade CRS, and 7% had grade 3 or higher CRS. Ninety-six percent of patients experienced pyrexia and 41% had hypotension. Moreover, approximately half (49%), had their AEs managed with tocilizumab and 17% received corticosteroids. Grade 4/5 CRS occurred in 1 patient each, and no patients had ongoing CRS as of the cutoff date.
The median time to onset of CRS was 4 days (range, 1-15) and the median duration of events was 6 days (range, 1-27). Ninety-nine percent of patients, however, had resolved events.
The overall incidence of any-grade and grade 3 and higher neurologic events were 60% and 19%, respectively. Tremor and confusional state occurred in 52% and 40% of patients, respectively. To manage this AE, 36% of patients received corticosteroids and 6% were given tocilizumab. The median time to onset of neurologic events was 7 days, and the median duration of events was 14 days. Ninety-three percent of patients had resolved events.
When stratified by cohort, patients with MZL had higher rates of grade 3 or higher neurologic events versus those with follicular lymphoma, at 41% vs 15%, respectively.
Grade 4 neurologic events occurred in 3 patients, and no grade 5 events were reported. At the data cutoff date, events were ongoing in 6 patients, which included grade 1 memory impairment (n = 2), and grade 1 attention disturbance, intermittent paresthesia, and tremor (n = 1), and grade 2 facial paresthesia (n = 1).
Furthermore, the median time to peak of anti-CD19 CAR T-cell levels after the infusion of axi-cel was 9 days (range, 8-371). The CAR T-cell expansion by peak and area under the curve trended higher in patients with MZL. Seventy-eight percent of patients with evaluable samples had low levels of detectable CAR gene-marked cells at 1 year.
There was a trend toward higher CAR T-cell levels in patients with follicular lymphoma who had an ongoing response at 12 months, Jacobson added. Moreover, peak CAR T-cell levels were associated with grade 3 or higher CRS and neurologic events in patients with FL. Similar trends linked between CAR T-cell expansion and outcomes were seen in patients with MZL.
In an interim analysis of ZUMA-5, in which 96 patients were evaluable for efficacy, the ORR was 93% (95% CI, 86-97) and the CR rate was 80% (95% CI, 71-88); the median time to first response was 1 month (range, 0.8-3.1).3 In patients with follicular lymphoma (n = 80), the ORR was 95% and the CR rate was 81%. In those with MZL (n = 16), the ORR was 81% and the CR rate was 75%.
In a press briefing during the meeting, Jacobson suggested that the biology of indolent NHL could be a reason for the higher efficacy rates and more tolerable safety profile with axi-cel compared with data in more aggressive lymphoma subtypes.
“It is the disease biology itself; we know that each of these lymphomas have unique microenvironments that may or may not make them more susceptible to immunologic killing,” Jacobson explained. “We know follicular lymphoma is a disease that can be cured by allogeneic stem cell transplant to a larger degree than what is seen with aggressive B-cell non-Hodgkin lymphomas. That may, in part, explain the efficacy.”
Additionally, Jacobson cited that the patients with MZL had slightly worse toxicity, replicative of patients with DLBCL.
“Their pretreatment inflammatory markers were elevated, or similarly, to DLBCL,” said Jacobson. “I think the disease itself is probably what dictates the toxicity profiles that we're seeing.”
Catherine Bollard, MD, MBChB, director of the Center for Cancer, and Immunology Director for the Program for Cell Enhancement and Technologies for Immunotherapy, Children’s National Research Institute, and moderator of the press briefing, commented on the longer ZUMA-5 findings.
“The work of Dr. Jacobson and colleagues is really very laudable. The longer the follow-up, the more robust the data, [we are then able] to really be able to draw some very effective comparisons than with historic [progression-free survival] and [disease-free survival] data, which will only really happen with the test of time,” said Bollard. “It's really exciting to see these data in the [indolent] lymphoma setting, and I actually would like to see it moved further up for the treatment of patients earlier in their disease process.”