Researchers have reported “encouraging outcomes” in blood cancer patients treated with CAR T cell therapy that did not meet the required specification for infusion.
A UK team found that LBCL patients infused with out-of-specification (OOS) product following CAR T cell manufacturing failure had comparable outcomes to those who received treatment that fully complied with infusion criteria.
“A 12-month overall survival of 52.8% and progression free survival of 46.2% for OOS-infused patients is reassuring and supports infusing patients with an OOS product where one is available,” they said, in a paper published in Blood Cancer Journal [link here].
CAR T cell manufacturing failure has been reported in up to 13% of cases, either because the process failed to yield the minimum required cell dose, or the product did not meet the required specification for factors such as viability, transduction efficacy, sterility and potency.
In the UK, regulatory processes allow administration of an OOS CAR T-cell product if the potential benefit outweighs the potential risks, on approval from the NHS England OOS CAR T Panel.
Other options for patients after manufacturing failure include a delayed infusion with a product that meets specification after remanufacturing, or foregoing CAR T cell infusion altogether.
The research team, which included haematology clinicians, researchers and pharmacists, sought to address the lack of evidence on these options by analysing outcomes for each. They also explored whether there were any factors increased the risk of CAR T cell manufacturing failure.
The retrospective, multicentre study collected data from nine CAR T centres in the UK and from the OOS CAR T panel, focusing on patients with relapsed or refractory LBCL approved for 3rd line or beyond CAR T-cell therapy with axicabtagene ciloleucel (axi-cel) or tisagenlecleucel (tisa-cel).
Among the studied cohort, 11 patients received delayed infusion with a product in-specification, 13 received an OOS product, and 14 were not infused. The control group included 38 LBCL patients without manufacturing failure, 29 of whom received a CAR T cell infusion.
Notably, the analysis did not reveal any significant differences in response, overall survival (OS) or progression free survival (PFS) between patients who received an OOS infusion, a delayed infusion, and controls.
One-year OS was 52.8%, 46.8% and 68.4%, respectively, while PFS was 46.2%, 24.2% and 41.4%.
Best ORR rates were 53.9%, 54.6% and 71.5% for OOS-infused, delayed-infused and controls-infused, respectively, according to the paper.
Regarding safety, the researchers did not observe any significant differences between the cohorts on rates of CRS, ICANS and cytopenias.
Increased risk of manufacturing failure with bendamustine
In addition, the only baseline variable linked with a significantly higher risk of manufacturing failure was bendamustine, with prior exposure in just 7.9% of controls versus 28.9% of patients with product failing to meet specifications (p=0.026).
Most of this excess risk was attributable to timing of therapy, with 0% of controls versus 23.7% in the manufacturing failure group (p=0.0029 having received the treatment within six months of apheresis, the authors noted.
Overall the findings “suggest encouraging outcomes for patients infused with an OOS product comparable to control patients without manufacturing failure infused with a product in-specification”, the authors said.
However, they also noted that outcomes may be different depending on the reason for the CAR T cell products being OOS, as they were unable to analyse this.
They also highlighted that all bar one OOS-infused patient met the required amount of viable CAR T cell dose, so it remains unknown whether the results would be replicated in those for whom this level is too low.
Also of note, “remanufacturing led to successful infusion with an inspecification product in only around 50% of attempts in our study, suggesting that the delayed-infused cohort is a select group of patients,” the authors said.
“Our results therefore support proceeding to an infusion where a suitable OOS product is available. Delaying infusion and attempting remanufacturing may be an option reserved for selected patients where an OOS product is not available,” they concluded.