ZUMA-12 primary analysis: Axi-cel for the frontline treatment for high-risk LBCL

The CAR T-cell product axicabtagene ciloleucel (axi-cel) has become established as a successful treatment for patients with relapsed/refractory (R/R) large B-cell lymphoma (LBCL). The efficacy and safety of axi-cel as a frontline treatment for high-risk LBCL is being investigated in the ZUMA-12 phase II clinical trial (NCT03761056). Interim study results have been previously reported by The Lymphoma Hub. Here, we summarize results from the primary analysis, published by Neelapu, et al., in Nature Medicine.¹

Study design

The study is a multicenter, open-label, single-arm, phase II trial in adult patients with high-grade confirmed LBCL. High grade was defined as the presence of MYC and BCL2 and/or BCL6 translocations, or an International Prognostic Index ≥3 at any time prior to enrolment. Further information on the study design can be found here.

Results

• In total, 42 patients were enrolled between February 6, 2019, and October 22, 2020, of which 40 patients received an axi-cel infusion. The safety analysis included 40 patients (median follow-up, 17.4 months; range, 6.0–26.7 months), with 37 patients in the efficacy analysis (median follow-up, 15.9 months; range, 6.0–26.7 months).

• Accrual is now complete; patients were recruited from the USA (n = 33, centers), Australia (n = 7, single center), and France (n = 2, single center).
• Key patient baseline characteristics are summarized in Table 1.

Table 1. Key patient baseline characteristics in ZUMA-12*

DA-EPOCH-R, etoposide phosphate, prednisone, vincristine sulphate (Oncovin), cyclophosphamide, doxorubicin hydrochloride (hydroxydaunorubicin), and rituximab; ECOG, Eastern Cooperative Oncology Group; FISH, fluorescence in situ hybridization; IPI, International Prognostic Index; R-CHOP, rituximab, cyclophosphamide, hydroxydaunorubicin hydrochloride (doxorubicin hydrochloride), vincristine (Oncovin) and prednisone. *Data from Neelapu, et al.1
Baseline characteristic, n (%) (unless otherwise stated)	N = 40
Median age (range), years	61 (23–86)
≥65 years	15 (38)
Male	27 (68)
Disease stage III–IV	38 (95)
ECOG Performance Status ≥1	25 (63)
One prior systemic line	32 (100)
Double- or triple-hit only	6 (15)
Double- or triple-hit and IPI ≥3	4 (10)
IPI ≥3 only	20 (50)
Deauville score
4	19 (48)
5	21 (53)
Alterations by FISH
MYC	19 (48)
BCL2	15 (38)
BCL6	10 (25)
Previous systemic therapy
R-CHOP	19 (48)
DA-EPOCH-R	18 (45)
Neither R-CHOP nor DA-EPOCH-R	6 (15)

• Efficacy outcomes were reported in 37 patients, these included:
  • Complete response rate: 78% (95% confidence interval [CI], 62–90)
  • Objective response rate: 89% (95% CI, 75–97)
  • Median time to first complete response was 30 days (range, 27–207 days)
  • Median time to first objective response was 29 days (range, 27–207 days)
  • Partial response was 11% (n = 4)
  • Stable disease was 8% (n = 3)

• At median follow-up of 15.9 months:
  • Objective response was 73%

• • Duration of response, event-free survival, and progression-free survival were not reached
• With regards to safety, of 40 patients included in the analysis:
  • Overall, 34 patients experienced ‘any adverse event’ (85%)
  • The most common Grade ≥3 events are summarized in Table 2
  • No Grade 5 events occurred
• In terms of neurological events:
  • Grade ≥3 cytokine release syndrome occurred in three patients (8%)
  • Grade ≥3 neurological events occurred in nine patients (23%)
  • No patients died from neurological events
• In total, six patients died (18%):
  • Four patients died from progressive disease
  • Two patients died from COVID-19

Table 2. Most common Grade ≥3 adverse events in patients receiving axi-cel*

*Data from Neelapu, et al.1
Adverse event, n (%)	N = 40
Neutropenia	5 (13)
Reduced neutrophil count	21 (53)
Reduced white cell count	17 (43)
Anemia	12 (30)
Encephalopathy	6 (15)
Thrombocytopenia	6 (15)
Hypoxia	5 (12)

Considering biomarkers and CAR T-cell levels:

• CAR T-cell expansion was seen in all 40 patients

• Median time to peak circulating CAR T-cells was 8 days (range, 8–37 days)
• Median peak CAR T-cell level was 36.27 cells µl^-1

No statistical significance was found between median peak levels of CAR T-cells in patients who relapsed/did not respond, and those who had an ongoing response to axi-cel. Several cytokines were found to be at least doubled in patients who experienced Grade ≥3 neurological events, including interleukin (IL)-5, macrophage inflammatory protein-1α (MIP-1α), interferon (IFN)-γ, granulocyte-macrophage colony-stimulating factor (GM-CSF), ferritin, tumor necrosis factor (TNF)-α, IL-10, IL-8, and programmed death-ligand 1 (PDL1) (p < 0.05).

Conclusion

The primary analysis of the ZUMA-12 phase II trial suggests that axi-cel has potential as a safe and effective option for the frontline treatment for high-grade LBCL. In the cohort of 37 efficacy-evaluable patients, objective response rate and complete response rate were 89% and 78%, respectively, and CAR T-cell expansion was observed in all participants. Safety data is comparable to other studies using axi-cel, with manageable adverse events and toxicities. Larger, randomized control studies are needed to replicate these results in large study populations and further compare efficacy and safety data with current standard-of-care treatment.