David Maloney, MD, PhD: Despite all these really outstanding and excellent results, we’re not curing everybody with CAR [chimeric antigen receptor] T-cells. There are patients who basically don’t respond, and there are patients who actually, unfortunately, relapse despite having a response. Caron, do you want to lead us through some of the recent findings?

Caron Jacobson, MD: One big reason is T-cell fitness. We’re all starting to understand more about what type of T cells within the product predict for efficacy and thinking about how to sequence therapies beforehand, or how to influence the T-cell phenotype of a product in order to enhance the types of T cells that lead to improved outcomes.

The other is obviously antigen loss. I’m not a leukemia doctor, but I think it’s actually as high as 50% antigen loss in the ALL [acute lymphoblastic leukemia] population upon relapse. We think this is the reason why CD19 was picked as an important tumor target. It’s essential for the malignant B cell to survive.

It’s astounding. How does the B cell lose the antigen and stay alive? It’s not that they’re actually losing the antigen. It’s that the gene is being alternatively spliced into the epitope that the CAR is binding to. And also, it’s the epitope that many of our immunohistochemical stains and antibodies bind to for the purpose of flow. That’s an important mechanism of escape.

David Maloney, MD, PhD: Max, how do we fix that?

Max Topp, MD: One way is maybe just to go after other antigens that are commonly expressed, like CD22. That may come in with clinical trials when we use the bicistronic approach, targeting CD19 and CD22 and applying more pressure to leukemia in that situation.

David Maloney, MD, PhD: We’ve never used just 1 antibiotic for a serious infection. We usually use 2. I think it’s the same kind of approach.

Caron Jacobson, MD: HIV is really the paradigm for this, right? We see that if we target more than 1 pathway, the virus can’t keep up, or you decrease the selected pressure for any of those mutations.

Jason Westin, MD: Frontline chemotherapy for aggressive malignancies is often combination chemotherapy. The same idea: having to pay a cost to escape 1 mechanism, but can’t pay the same cost multiple times over.

Max Topp, MD: With our novel technologies like NGS [next-generation sequencing], we actually also trace clones. It’s not just 1 clone. So if we have a better understanding of the clones’ architectures, then we can predict patients who are at risk.

Frederick Locke, MD: It’s also important to recognize that although we do see CD19 antigen or epitope loss of where the CAR binds, this is probably not the only mechanism of treatment resistance. There are data in ALL patients treated with 4-1BB constructs that actually reveals that loss of a persistence of the construct may correspond to relapsing disease. But we also have to be careful to say that’s in a specific CAR with a specific construct and a specific disease, and that may not hold the same for lymphoma and multiple myeloma. But we really don’t fully understand the reasons why patients may not respond or have durable responses to CAR T-cell therapy, so we really need to do additional translational discovery.

Max Topp, MD: There are resistance mechanisms and checkpoint pathways. And there’s something else that we may dip into: that we can give the checkpoint inhibitors to actually help the CAR T cells to actually fight off resistance mechanisms. I think it’s a very interesting field of research, clinically, that is very useful too, potentially...

David Maloney, MD, PhD: We’re beginning to see some of the data pop up with CD22 and CD19 CARs, with either 2 CARs, or bicistronic CARs. I think that’s going to maybe decrease some of this. But I want to come back to the T-cell fitness. Right now, the patients who are really being treated are relapsed/refractory. They’ve been through multiple therapies. Obviously, your T cells can’t be very healthy after all these treatments. Is there a role for collecting T cells from high-risk patients, maybe early on in their disease before they go through all this therapy?

Frederick Locke, MD: Scientifically, it absolutely makes sense. There are strong data to suggest that patients who had more lines of chemotherapy or have less fit T cells at the time of collection for manufacture of CAR T-cell therapy have an end T-cell product that is not as strong or fit; and they’re less likely to have responses or durable responses. It does introduce some complexity when we get beyond the scientific aspects of it. Can we be collecting cells? Can we be holding them in freezers? Who’s going to do that? And what are the payment mechanisms?

David Maloney, MD, PhD: Excellent points.

Caron Jacobson, MD: Not all the products are produced from frozen products, right? This concept is limited to certain products that are only produced from frozen cells.

Transcript Edited for Clarity