Dr Ip on Real-World Data Informing CAR T-Cell Therapy Selection in LBCL

“For many of us in the community who are referring [patients to] or treating [patients] with CAR T-cell therapy, we're trying to give them the quickest option, even though we realize there might be some more risks of toxicity. That is something [we] should obviously take into account, and [these] data confirm that.”

Andrew Ip, MD, medical oncologist, Division of Lymphoma, Hematologic Malignancy, John Theurer Cancer Center, Hackensack University Medical Center, discusses the importance of making informed clinical decisions when selecting CAR T-cell therapy for older and frailer patients vs younger patients with large B-cell lymphoma (LBCL).

Findings from the real-world retrospective multicenter cohort study evaluating CAR T-cell therapy in patients with LBCL were presented at the 2024 ASH Annual Meeting.

Clinical trial data have previously indicated that the manufacturing for axicabtagene ciloleucel (axi-cel; Yescarta) can be more consistent and does not have as many out-of-specification failures, Ip details. This means that axi-cel could be available for administration before other CAR T-cell therapies, Ip adds.

In the real-world study, Ip says that there were more patients treated with axi-cel who had primary refractory disease and had an overall poorer performance status. Based on this finding, he emphasizes that this means patients in the real-world setting could be sicker and would require CAR T-cell therapy sooner. An important takeaway he notes is considering the risks of toxicity when referring or treating patients with CAR T-cell therapy, especially when it’s the quickest treatment option.

The study included 501 patients with LBCL from 4 centers in the U.S. and Israel between April 2016 and January 2024. Of note, 96 patients were treated with lisocabtagene maraleucel (liso-cel; Breyanzi), 133 with tisagenlecleucel (tisa-cel; Kymriah), and 272 with axi-cel. At a median follow-up of 19 months in the overall cohort, the complete response rates with liso-cel, tisa-cel, and axi-cel were 73%, 49%, and 61%, respectively (P = .001). Partial response rates the same arms were 18%, 15% and 20%, respectively (P = .001). Median durations of CR were not reached (NR) for liso-cel (95% CI, NR-NR), 18 months for tisa-cel (95% CI, 7.4-NR), and 27 months for axi-cel (95% CI, 19-NR).

Regarding safety, lower rates of any-grade cytokine release syndrome (CRS) were observed in patients treated with liso-cel (55.1%) compared with tisa-cel (69.1%) and axi-cel (88.8%; P < .001). Liso-cel was also associated with lower rates of grade 2 or higher CRS (16.1% liso-cel; 31.3% tisa-cel; 42.8% axi-cel), any-grade immune effector cell-associated neurotoxicity syndrome (ICANS; 13.8%, 17.5%, and 37.9%; P < 0.001), and grade 2 or higher ICANS (6.2%; 9.9%, and 25.9%; P < .001) than tisa-cel and axi-cel.

Overall, differences in efficacy outcomes with liso-cel vs axi-cel were not statistically significant, although their safety profiles differed. However, he concludes that there is still selection bias, which could be inherent in real-world studies across the board.