CAR T-Cell Therapy Response Rates

Andre Goy, MD: I alluded to the limits and practical issues to get access to the CAR [chimeric antigen-receptor T] T-cell therapy. But what do we do to try to make it better? With this ASCO [American Society of Clinical Oncology] 2020 Annual Meeting data, there was a small experience on re-treatment. Can you re-treat those patients with CAR T? And the reasoning behind this is that there is some evidence that these T cells will not stick around long enough. There was some discussion that on the need or not to have precise CAR T-cells. We had about one-third of the long responders in the ZUMA-1 trial who did not have resistant, long-term precision CAR T cell.

With the recovery of the B cells and even immunoglobulin, after a year we can see the CAR T cells disappear. But there’s also the idea that we can re-treat those patients who relapse. This was a small number of patients who were from the axi-cel setting, and over half the patients achieved a response: 4 CR [complete response], 3 PR [partial response]. What was interesting is that some of these responses can be durable: close to 3 months, up to 225 days for the longest. There was a rate of CRS [cytokine release syndrome] at re-treatment, but it seems there was less neurotoxicity and maybe also less CAR T expansion. We don’t know exactly how to read this, but it’s feasible.

There is an interesting question being explored at some other trials: Can we do repeat infusion of CAR T? That’s 1 strategy. The other 1 is that can we improve by bringing the CAR T earlier. That’s the question about the timing of the CAR T if we do it earlier in the second line. There’s the ZUMA-7 trial that has completed, but the results are pending. This other study is with liso-cel and tisa-cel also at a second line with a crossover, so in second-line therapy, randomizing against salvation transplant.

There was an interesting study looking at liso-cel in patients who are not transplant eligible. It is an interesting study because liso-cel potentially has an initial toxicity profile, so it might be easier to do. It’s a slightly different product when the CD4 and CD8 subtype of T cells are prepared separately and infused separately. They get the same for the patient. What was impressive in that study was not a huge number of patients but that the overall response rate was 89% and the CR rate was 56% with very low toxicity, and 12 of 29 patients we have done as an outpatient. This is really important because that offer is potentially an option for these nontransplant-eligible patients. It’s for durability, but this is something to take into account.

As for the results and data on the dual CAR T anti-CD19 and -CD22 in combination with pembrolizumab, there’s a very interesting study called the ALEXANDER study. The rationale for this is there is a subset of patients in large cell lymphoma that become resistant to CAR T because they lose CD19 expression. It’s something we see more frequently in ALL [acute lymphoblastic leukemia] as a mechanism of resistance than in large cell lymphoma, but it still happens. And so we try the dual CAR T CD19,-CD22 with PEMBRO [pembrolizumab] to try to see if we can improve. Knowing that a checkpoint blockade is also 1 of the mechanisms of resistance, can we improve the responses seen? This is dose escalation, not a large number of patients. But the response rate was very high: 10 of the 11 CRs. A small number are ongoing. There was no grade 3 CRS and 1 grade 3 neurotoxicity. The recommended phase 2 is 150 to 450 million, and we did want PEMBRO [pembrolizumab] selected.

The last study at ASCO that is worth mentioning was the ALLO CAR T-cell off the shelf. That is a product called ALLO-501, a gene-editing-based technology to remove the T-cell receptor and CD52. This is also used as a dose escalation, presented by Dr Sattva Neelapu. It shows a response rate of 63% with half these patients in complete remission, and 75% of the patients who responded were still in response at the time of cutoff.

That’s also interesting because there was no GvHD [graft-vs-host disease], and there was allogeneic healthy T cell. There was no dose-limiting toxicity, 1 grade 3 CRS, and no neurotoxicity. That’s very appealing because it would be off-the-shelf, ready-to-go T-cells. Potentially that’s very important because 1 of the factors of failure of CAR T is the T-cell fitness. Patients who are heavily pretreated have a lot of immunosuppressive adverse effects with the active disease of ongoing lymphoma. These patients don’t have the best activity. If we have normal allogeneic normal T cells, that might address the problem.

The last aspect that is also important was from ASH [American Society of Hematology Annual Meeting], and this is the BiTE [bispecific T-cell engager] antibody because we know that patients still can fail quite a bit of CAR T cells, and it was an anti-CD20 bispecific antibody, a BiTE antibody, presented in a plenary session at ASH. It was very impressive because this was a response rate of 39%, 22% CR. But there were some responses, durable and a CR durable in continued remission, even in patients who had stopped therapy after 16 months of treatment. This is also being explored in the frontline setting in combination with chemotherapy, CHOP [cyclophosphamide, hydroxydaunorubicin hydrochloride, vincristine, prednisone]–POLA [polatuzumab] versus R-CHOP [rituximab, cyclophosphamide, hydroxydaunorubicin hydrochloride, vincristine, prednisone]. A number of initiatives are trying to address the new unmet need that we have in CAR T cells. Now we know how to manage more toxicity. Can we improve on efficacy and look at easier ways to use those CAR T cells?

Transcript Edited for Clarity