Transcript: Leo Gordon, MD: The patient population of the CAR-T trials are quite similar, although there were some differences. Juno’s TRANSCEND study initially included patients with performance status [PS] as high as 2. It included patients who’d had a prior allogeneic stem-cell transplant, which allowed bridging therapy. Because of the long wait to get a slot for these studies, and the long wait that was associated with manufacturing the product, patients sometimes had anywhere from 3 to 8 weeks, or maybe even longer, from the time that it was determined they needed another treatment to the time they could receive it.
The allowance of bridging in the Juno trial versus, say, the ZUMA trial, where bridging was not allowed, led to some differences in patient population. You know there is an inherent selection bias in clinical trials and cancer chemotherapy in general. It takes a long time to get a patient on a clinical trial, not even with CARs, but any cancer clinical trial, because of the regulatory process involved and the consent process. One has to take the data from clinical trials with that into account, because, when you have a selection bias that delays your treatment, you’re selecting out a group of patients who can afford that delay that may have an inherently better PS and not as bad of a disease. That’s certainly one of the criticisms against all of the CAR trials—there’s an inherent selection bias because it takes a long time to get patients on studies.
When these drugs are approved—they’re being approved now and the time shortens—we may not see data quite as good as that from the initial clinical trials. The other difference is, in the TRANSCEND study, that you had a group of patients who were allowed who might have been sicker but still recognized, when referring to earlier data. What we did was select out a so-called core group of patients who had a good performance status, who didn’t have prior allogenic stem-cell transplant, and who didn’t have transformation from marginal ZUMA lymphoma or chronic lymphocytic leukemia, but rather, only had a transformation from follicular lymphoma. The initial differences were there, and then when the TRANSCEND study started looking at the core group of patients, that limited the patients to a better group and was more consistent with what was seen in the JULIET study and ZUMA study.
Matthew Lunning, DO: The conditioning regimens are interesting also because they are different across constructs. The axi-cel [axicabtagene ciloleucel] uses cyclophosphamide. Its construct uses Cytoxan and fludarabine, and the Cytoxan is given for 3 days at 500 mg per meter squared, where fludarabine is given at 30 mg per meter squared for 3 days also. Whereas liso-cel [lisocabtagene maraleucel] is a slightly lower dose of the Cytoxan at 300 mg per meter squared. And then the fludarabine is given at 30 mg per meter squared—at least in the current TRANSCEND lymphodepletion clinical trial.
I chose to discuss tisa-cel [tisagenlecleucel] last because they are different than the other 2 in that they have a Cy/Flu [cyclophosphamide/fludarabine] lymphodepleting arm, which, again, is a little bit lower than the axi-cel and the liso-cel dosages currently, as it is Cytoxan 250 mg per meter squared, and then fludarabine at 25 mg per meter squared.
There is the option in those patients who received prior Cytoxan or were refractory to it, to use a bendamustine-containing regimen or lymphodepletion, at 90 mg per meter squared given on days 1 or 2. A third option is to not give any lymphodepleting therapy at all if within 1 week of infusion of the planned CAR-T cells, the white blood cell count is less than 1000.