Does the use of bridging therapy impact patient outcomes whilst awaiting axi-cel manufacturing?

Axicabtagene ciloleucel (axi-cel) is an anti-CD19 chimeric antigen receptor (CAR) T-cell therapy approved for the treatment of patients with relapsed/refractory (R/R) diffuse large B-cell lymphoma (DLBCL) in the United States (US) and Europe.^1,2 Bridging therapy, defined as lymphoma therapy given between apheresis and the start of lymphodepleting chemotherapy, may be given to control a patient’s disease whilst awaiting manufacturing of the CAR T product. The impact of bridging therapy on subsequent patient outcomes is unknown. Bridging therapy was not permitted in the original ZUMA-1 study, which the axi-cel approvals are based on.³

During the 61^st American Society of Hematology (ASH) Annual Meeting & Exposition, Michael D Jain, Moffitt Cancer Center, Tampa, US, presented results from a study evaluating the impact of bridging therapy on patient outcome while awaiting axi-cel manufacturing, on behalf of the U.S. Lymphoma CAR-T Consortium, and co-first-author Miriam T Jacobs, Washington University School of Medicine in St Louis, St Louis, US.³

The U.S. Lymphoma CAR-T Consortium comprises of 17 academic centers in the US that treat patients with R/R DLBCL, primary mediastinal B-cell lymphoma (PMBCL), and transformed follicular lymphoma (TFL) with CAR T therapies. In this study, 298 patients were leukapheresed across all centers, with 23 not proceeding to CAR T infusion, predominantly due to lymphoma progression or death (n= 20). The axi-cel infusion was administered to 275 patients.

Patient characteristics (n= 298)

• Age > 60 years: 154 (52%)
• Eastern Cooperative Oncology Group (ECOG) performance status 2–4: 58 (20%)
• Advanced-stage disease (III or IV): 244 (82%)
• Diagnosis of DLBCL: 203 (68%); PMBCL: 19 (6%); TFL: 76 (26%)
• Double/triple hit: 64 (23%)
• Received ≥ 3 prior lines of therapy: 222 (75%)
• Did not meet inclusion criteria for original ZUMA-1 study: 129 (43%)

Efficacy

• Median follow-up: 13.8 months from leukapheresis
• Median progression-free survival (PFS): 7.16 months (95% CI, 5.65–12.4)
• Median overall survival (OS): not reached

To determine if any covariates were associated with a poor OS, multivariate analysis was conducted. This showed several high-risk groups as demonstrated in Table 1, one of which was the use of bridging therapy (p= 0.03).

Table 1. Statistically significant factors associated with poor OS in multivariate analysis

HR, hazard ratio; LDH, lactate dehydrogenase

High-risk group

HR

95% CI

p value

Male sex

1.7

1.0–2.7

0.04

ECOG score 2–4

1.8

1.1–3.0

0.02

Primary refractory disease status

1.9

1–3.6

0.04

Refractory disease status

1.7

0.9–3.3

0.08

Total bilirubin > 1.5g/dl

5.1

1.8–14.5

0.002

LDH level > upper limit of normal prior to conditioning

3

1.7–5.4

0.0001

Bridging therapy given

1.8

1.0–2.7

0.03

The impact of bridging therapy

In this study, bridging therapy was given to 53% of patients (n= 158), most commonly chemotherapy (54%) followed by steroids only (23%), radiotherapy (12%) and targeted agents (10%). Of the patients receiving bridging therapy, 89% (n= 141) went on to receive the axi-cel infusion. This is in comparison to the 47% of patients who did not receive bridging therapy (n= 140), of whom 96% received the axi-cel infusion (n= 134).

Covariate analysis

Some baseline covariates of patients receiving bridging therapy were significantly different to those not receiving bridging therapy, as shown in Table 2.

Table 2. Covariates that are statistically different between patients receiving, and not receiving, bridging therapy

IPI, international prognostic index; LDH, lactate dehydrogenase

Covariate

No bridging, %

Bridging, %

p value

ECOG score 2–4

8

30

< 0.001

Stage III/IV disease

75

89

< 0.001

IPI score 3–5

38

69

< 0.001

Bulky disease

14

31

< 0.001

Did not meet ZUMA-1 eligibility criteria

34

52

0.002

LDH < 200 IU/L at conditioning

26

12

0.03

Toxicity and causes of death

Given as no bridging vs bridging

Looking at toxicity and cause of death following axi-cel infusion, more patients who received bridging therapy had grade ≥ 3 cytokine release syndrome (5% vs 9%), more grade ≥ 3 neurotoxicity (28% vs 34%), higher non-relapse mortality rate (1.5% vs 7.1%) and increased death after lymphoma relapse at 12-months post-infusion (17% vs 37%).

Outcomes by bridging therapy status

• PFS (no bridging vs bridging): HR= 1.61, p= 0.002
• OS (no bridging vs bridging): HR= 2.45, p= 0.001
• The decrease in PFS and OS was not associated with one specific type of bridging therapy (chemotherapy vs steroids vs radiotherapy vs targeted agents)

Propensity score matching

To evaluate the differences in baseline characteristics between patients who did not receive bridging therapy, and those who did, patients were matched by propensity score. In total, 104 patients were identified and matched in each group. Following the matching, baseline characteristics were found to be similar, with no statistically significant differences. When analyzing PFS following propensity score matching, there was no longer a difference between patients who did not receive bridging, and those who did (HR= 1.19, p= 0.339). However, there was still a difference in OS, with patients receiving bridging having a worse OS compared to the no bridging group: HR= 1.68, p= 0.023.

Conclusion

The use of bridging therapy was associated with a worse OS in multivariate and propensity score matched analysis. It is possible that bridging therapy may identify a sub-group of patients with a different biology to the rest of the cohort, or alternatively that bridging therapy has an effect on the host or tumor microenvironment, ultimately affecting the efficacy of CAR T therapy. Therefore, the use of bridging therapy should be carefully considered within each specific scenario, though it is not currently recommended for patients awaiting axi-cel manufacturing.