Jason Westin, MD: We struggle with this decision of, which product do you choose? If there’s a third product, how will you select for that third product? We’ve treated patients north of age 80 with all these products, and I think it’s a select group. This is not the typical 80-year-old patient who comes in the clinic in the wheelchair, who’s having trouble getting up and caring for themselves. These are all very fit, older patients. So the data set that show older patients may do better—it’s possible that might be a little bit of selection of the right older patients. It could be some of the biology mentioned, but it may be other factors.

I think 1 of the take-home points we like to stress is that CAR [chimeric antigen receptor] T-cell eligibility is different from transplant eligibility, and there are a number of older patients who might not have gone for autologous transplants because of concerns about who may be able to tolerate CAR T-cell therapy.

David Maloney, MD, PhD: But there’s also an important difference. When we talk about transplant, we won’t transplant patients unless they’re essentially in a complete remission, right?

Jason Westin, MD: Correct.

David Maloney, MD, PhD: This is really 1 of the big dilemmas on the cellular immunotherapy service. You’re getting patients who are actively trying to basically die because of their disease.

Jason Westin, MD: It’s a higher bar.

David Maloney, MD, PhD: Often, you’re trying to just keep them alive; to get a product back to be able to treat them. So it’s way different from the typical transplant arena, where these people are responding to their therapy to be eligible for transplant. But we’ve really seen encouraging results from the real-world data with both of these agents. We’re seeing treatment of patients who would not have qualified for the clinical trials. Thus far, I’ve always been struck by any major differences that are predicting outcome in these real-world patients.

Caron Jacobson, MD: In our series, patients who fell into the ineligible category actually didn’t do as well as the patients who fell into the eligible category, in terms of long-term progression-free survival and overall survival. But they still did way better than what was predicted by SCHOLAR-1, so it’s still the best thing we have for those patients. The other things that came out of our analyses were things that we’ve talked about before; but tumor volume, in terms of response rate, and then also high-pretreatment CRP [C-reactive protein]. A patient who had a CRP that was greater than 30 mg/L had incredibly discrepant durability of response, progression-free survival, and overall survival compared with a patient who had a lower CRP at the time of T-cell infusion.

David Maloney, MD, PhD: And they had greater CRS [cytokine release syndrome] as well. Fred, do you have to have CRS to get a response?

Frederick Locke, MD: No. There’s an interesting abstract here, again from our real-world data set, standard-of-care-treated patients. Dr Miriam Jacobs presented that abstract, and what we looked at was patients who did not develop cytokine release syndrome. Of course, the way cytokine release syndrome is defined, if you have a fever, you have grade 1 cytokine release syndrome. In axicabtagene ciloleucel—treated patients, that’s between 80% and 90% of patients who will develop at least a fever and cytokine release syndrome. We found 25 patients of the almost 300-patient data set who did not develop a fever after axicabtagene ciloleucel treatment. In fact, their outcomes were not different from patients who did develop CRS. The notion that including both response rate and progression-free survival and overall survival—the notion that a patient who hasn’t had a fever after treatment with axicabtagene ciloleucel is somehow doomed to have a relapse is actually not the case based on our real-world data set.

The other thing I’d add to complement what Caron said about the real-world data or the standard-of-care data is, there are features that do seem to predict for worse outcomes. We analyzed the data not by eligibility of ZUMA-1, but we actually broke down into all the individual characteristics that would have made them ineligible for the ZUMA-1 trial. We found very similar findings. But again, ECOG performance status of 2 or higher was highly predictive of a worse outcome. A high LDH [lactate dehydrogenase]—so a marker for worse disease and disease burden. These are the features that seemed to associate with worse progression-free survival and higher toxicity rates.

Jason Westin, MD: This abstract is very important, because we have patients who get very nervous if they’re not getting a fever. “Is it working?” We sometimes root for a little bit of fever because that says that there’s some expansion of the cells. There’s something happening. These data are very important to say that if you don’t get a CRS, that’s OK. Don’t panic.

David Maloney, MD, PhD: I have a practical question about that. We generally administer axicabtagene ciloleucel in the hospital, and we keep people hospitalized for about 7 days. If you have somebody who does not have any fever over 7 days, do you still keep them in the hospital for the whole time? We usually discharge them around day 8, and then I’ve had people readmitted on day 9 with fevers. What do you do?

Frederick Locke, MD: We admit all our patients who are getting axicabtagene ciloleucel, and we keep them in the hospital for at least 5 days. We’ve treated over 300 patients with CAR T-cell therapy. We’ve had rare cases in which a patient was discharged at 5 or 6 days and then came in with a fever or, more commonly, neurological toxicities, which often happen later. But in general, axicabtagene ciloleucel leads to high rates of cytokine release syndrome—80% to 90%—and these are patients with fever and are generally neutropenic because they’ve received fludarabine and cyclophosphamide. And so that’s a patient for whom you need to rule out sepsis and other issues. And they don’t feel well when you’ve got a 41°C fever. They really need aggressive supportive care. The average day of onset of the cytokine release syndrome with axicabtagene ciloleucel in lymphoma is basically within a day, day and a half.

Caron Jacobson, MD: But I would say that even though the rates of cytokine release syndrome are really quite high, many of the patients just have a fever. They just have a fever that lasts, and those may also be patients for whom you could consider treating as an outpatient. I think what we need to do is figure out a composite of biomarkers to follow for these patients that goes into some pretreatment characteristics. That may mean that companies can figure out point-of-care testing where we can get real-time information about CAR T-cell expansion in our commercial patients as well as the rate of increase in some of our key cytokine levels. That may come into play. I’m not saying right now. We do something different from you probably. We give all our patients G-CSF, so their window of neutropenia is actually quite small.

Frederick Locke, MD: Another area of potential controversy.

Jason Westin, MD: I agree with Caron. This may also be product specific. The idea for outpatient administration, even outpatient management of fever, is something that’s controversial but could potentially be done. Especially with different constructs, it may be more common as we go along.

Transcript Edited for Clarity