Overcoming CAR T Manufacturing Failure in LBCL

Out-of-specification (OOS) autologous chimeric antigen receptor (CAR) T-cell therapy was safely administered in patients in the UK who had relapsed/refractory large B-cell lymphoma (LBCL), potentially widening the pool of available treatments for these patients, despite initial manufacturing failure.¹

Results recently published in Blood Cancer Journal also show that remanufacturing led to successful infusion of a product made to specifications in approximately 50% of patients, meaning this process, too, could be an option for patients for whom an OOS CAR T is unavailable.

There is an especially high rate of CAR T manufacturing failure (MF) among people who have non-Hodgkin lymphoma, of which LBCL is a common subtype. The rate is just about 25% of cases in this patient population, compared with 1% to 13% among everyone qualified to receive CAR T.^2-5 Less is known, however, about outcomes from an OOS-administered product and if there are risk factors that contribute to initial treatment MF.

To learn more, the study investigators used data on patients approved to receive axicabtagene ciloleucel (axi-cel [Yescarta]; Kite; n = 805) or tisagenlecleucel (tisa-cel [Kymriah]; Novartis) by the UK national CAR T Clinical panel (NCCP) between January 2019 and January 2023. The CAR Ts were approved in 2023 and 2019, respectively, by the UK’s National Institute for Health and Care Excellence.^6-7

Results Following Manufacturing Failure

Overall, 3.9% of the patients—38, or 28 set to receive axi-cel and 10 set to receive tisa-cel—saw at least 1 CAR T MF; broken down, this was 3.5% of the axi-cel group and 5.7% of the tisa-cel group, with 18 of the failures being OOS products and the remaining 20 failures not producing any usable product. Most of these patients were male and had de novo diffuse LBCL, comorbidities, 1 to 2 prior lines of therapy, prior bendamustine use, prior high-dose cytarabine, and an Eastern Cooperative Oncology Group status of 1.

Between the 2 periods for which they compared outcomes, January 2019 to January 2021 and February 2021 to January 2023, there was minimal difference in MF rate. Period 1 saw an MF frequency of 3.6% vs 4.0% in period 2 (P = .74). Manufacturing attempts totaled 59 among the 38 patients, and for those requiring remanufacturing, either new apheresis material was used of or repeated on the same material.

Twenty of the remanufacturing attempts led to OOS products, and 13 of these were infused; 13 produced in-specification products after remanufacturing attempts, and 11 were infused; and the product was not available after 26 of the attempts. Thirteen patients received an OOS CAR T, 11 had a delayed infusion with an in-specification CAR T, and 14 were not infused.

Data incorporated for this study were collected from 9 centers in the United Kingdom focused on third-line and beyond chimeric antigen receptor T-cell administration in patients with relapsed/refractory large B-cell lymphoma.

| Image Credit: © Funny-stock.adobe.com

Reasons for MF did not differ significantly between the 2 periods, but the top reasons for MF were low cell viability and T-cell purity (8 each) for the attempts that produced an OOS or not viable product and low cell viability and low interferon-gamma (5 each) when just an OOS product was manufactured. However, bendamustine was implicated as a risk factor for MF in 28.9% of the 38 failures (P = .026), with the risk primarily concentrated among patients who received the chemotherapy agent within 6 months of apheresis (P = .0029).

Overall and complete response rates did not differ significantly at 1 (P = .47) and 3 (P = .26) months after infusion between the treatment and control groups:

• 53.9% and 46.2% for both at both time points among those infused with an OOS product
• 54.6% and 27.3% and 45.5% and 27.3%, respectively, among those who underwent a delayed infusion with an in-specification CAR T
• 71.4% and 57.1% and 32.1% and 39.3% among the control group

There also were no significant differences for 1-year overall survival among these 3 groups:

• 52.8% (95% CI, 23.4%-75.5%), OOS infusion
• 46.8% (95% CI, 14.8%-73.9%), delayed infusion
• 68.4% (95% CI, 48.0%-82.1%), control group

The authors noted, in particular, a lack of a significant difference between the patients who underwent an OOS infusion and the control group (P = .34) and between those who had a delayed infusion and the control group (P = .81).

Safety Outcomes

Any-grade cytokine release syndrome (CKS) was seen in 83.3% of the OOS-infused group, 90.9% of the delayed-infusion group, and 93.1% of the control group; grade 3/4 CKS was seen in 15.4%, 0%, and 6.9%, respectively. Again, these results were deemed not significant (P = .81 for any grade; P = .50 for grade 3/4). Corresponding rates for immune effector cell–associated neurotoxicity syndrome were 38.5%, 30.0%, and 34.5%, and 7.7%, 18.2%, and 10.3%, and not significant (P > .99; P = .72).

Four patients died: 1 patient from the OOS group died from unknown causes, but after a myelodyplastic syndrome diagnosis; 2 patients from the control group died from SARS-CoV-2 infection; and 1 patient from the control group died from an unknown cause.

Limitations

Limitations on these findings are that these were patients approved for third-line and beyond CAR T, so the authors hypothesize the risk of MF may be less in less heavily pretreated patients. They also were only able to evaluate possible effects on the CAR T manufacturing process, not variables within the process itself, and had a small sample size.

Conclusion

“Similar to the findings reported from previous studies, we did not find any significant difference in toxicity between infusion with an OOS or in-specification products,” the authors concluded. “Our results suggest encouraging outcomes for patients infused with an OOS product comparable to control patients without.”

References

1. Dulobdas V, Kirkwood AA, Serpenti F, et al. Risk factors for CAR T-cell manufacturing failure and patient outcomes in large B-cell lymphoma: a report from the UK National CAR T Panel. Blood Cancer J. 2025;15(1):30. doi:10.1038/s41408-025-01225-9

2. Bersenev A. CAR-T cell manufacturing: time to put it in gear. 2017;57(5):1104-1106. doi:10.1111/trf.14110

3. Bersenev A, Kili S. Management of ‘out of specification’ commercial autologous CAR-T cell products. Cell Gene Ther Insights. 2018; 4(11), 1051-1058. doi:10.18609/cgti.2018.105

4. Baguet C, Larghero J, Mebarki M. Early predictive factors of failure in autologous CAR T-cell manufacturing and/or efficacy in hematologic malignancies. Blood Adv. 2024;8(2):337-342. doi:10.1182/bloodadvances.2023011992

5. Wang X, Borquez-Ojeda O, Stefanski J, et al. Depletion of high-content CD14+ cells from apheresis products is critical for successful transduction and expansion of CAR T cells during large-scale cGMP manufacturing. Mol Ther Methods Clin Dev. 2021:22:377-387. doi:10.1016/j.omtm.2021.06.014

6. Kite’s Yescarta is first CAR T-cell therapy recommended for routine use in England, United Kingdom. News release. Kite; January 26, 2023. Accessed March 20, 2025. https://www.kitepharma.com/news/company-statements/kite-yescarta-is-first-car-t-cell-therapy-recommended-for-routine-use--in-england-united-kingdom#:~:text=%E2%80%93%20January%2026%2C%202023%20%E2%80%93%20Kite,with%20certain%20forms%20of%20lymphoma

7. Keown A. Novartis’ CAR-T treatment Kymriah snags second approval in the UK. BioSpace. February 1, 2019. Accessed March 20, 2025. https://www.biospace.com/novartis-car-t-treatment-kymriah-snags-second-approval-in-the-u-k