received lymphodepleting chemotherapy (cyclophosphamide and fludarabine) before treatment with LN-145 and an interleukin-2 (IL-2); the IL-2 promotes TIL expansion.
Two out of 13 evaluable patients achieved durable complete responses (CRs), both lasting for longer than a year. Four patients have confirmed or unconfirmed partial responses (PRs), if the fourth PR is confirmed, the overall response rate will be 33 percent.
The most common adverse events were cytopenias related to lymphodepleting chemotherapy; Creelan said there was almost no on-target, off-tumor toxicity.
Gal Markel of Sheba Medical Center in Israel, the discussant for Creelan’s presentation, noted several unanswered questions, including whether or not the four doses of Opdivo administered before LN-145 contributed to patients’ responses and how this treatment regimen would work in older, sicker individuals – patients with cardiovascular co-morbidities were excluded from the study.
However, Jefferies analyst Biren Amin said in a 27 April note based on the AACR abstract for Creelan’s presentation that a 10 percent CR rate would have been clinically meaningful, so the 15 percent CR rate seen in the study was impressive. Amin pointed out that the CR rate for Opdivo alone is 4 percent in second-line metastatic NSCLC and 1 percent in the third-line setting.
PD-1 DIDN’T IMPROVE YESCARTA EFFICACY, BUT STUDIES CONTINUE
Data for Kite’s CD19-targeting CAR-T therapy Yescarta in combination with Roche’s PD-L1 inhibitor Tecentriq (atezolizumab) in the treatment of refractory diffuse large B-cell lymphoma (DLBCL) did not show an improvement in Yescarta’s efficacy. Caron Jacobson from Dana-Farber Cancer Institute in Boston presented the results, which came from the Phase I/II open-label study ZUMA-6.
The AACR presentation covered data from 28 patients, most of whom were enrolled in the Phase II portion of the study. Some of the patients were enrolled in Cohort 3 of the Phase I portion of the study, which determined the Yescarta/Tecentriq dosing for Phase II.
Immune checkpoints, including PD-1 and PD-L1 have been shown to be upregulated in tumors and after CAR-T cell infusion, the ZUMA-6 investigators hypothesized that PD-L1 inhibition would augment Yescarta activity.
However, the overall response rate in Kite’s ZUMA-1 study, which supported approval of Yescarta monotherapy, was 83 percent and the CR rate was 58 percent. But in the ZUMA-6, the ORR was 75 percent and the CR rate was 46 percent in the 28 patients receiving combination therapy, showing that Tecentriq did not improve Yescarta’s efficacy. Tecentriq did not improve expansion of CAR-T cells in ZUMA-6, Jacobson reported.
In terms of safety, 27 patients in the ZUMA-6 data experienced CRS. All but one of the CRS events were Grade 1 or 2; the other was Grade 3. There were 19 neurological events, including eight Grade 3 or 4 events, but no Grade 5 events. The safety profile in this study was similar to prior Yescarta trials.
Jacobson said further analysis of ZUMA-6 will consider whether certain patient groups are more likely to benefit from combination PD-L1/CAR-T therapy, but more studies are needed to better understand the interaction of CAR-T cells and PD-1/PD-L1 inhibition.
Discussant Markel noted that preclinical studies justified testing this combination in patients and said the results overall were encouraging in terms of justifying additional testing in patients.
A Gilead spokesman told Scrip that “we are performing further analysis of ZUMA-6 data to better understand whether there are identifiable patient populations who may derive benefit from this combination approach, and this will inform our decision on how best to proceed.”
NCI CONTINUES TO EVALUATE BISPECIFIC CAR-T
Haneen Shalabi, from the National Cancer Institute Pediatric Oncology Branch, reported results from the first 13 patients treated with an NCI-developed autologous CAR-T cell construct targeting both CD19 and CD22 for the treatment of children and young adults with relapsed or refractory B-cell ALL.
This bispecific therapy was designed to overcome the challenge of antigen escape, which happens in ALL when patients who respond to CD19-targeting CAR-T therapy but have a relapse when there are no more CD19-expressing cancer cells for the CAR-T cells to attack. By targeting CD19 and CD22, the NCI researchers hypothesized that their bispecific treatment would continue to eliminate leukemia cells because the CAR-T cells were engineered to seek out two different antigens.
Six patients (46%) had CRS and two (15.4%) had Grade 3 or higher CRS, while one patient (7.7%) experienced Grade 3 neurotoxicity. All of these adverse events were reversible.
Five out of 12 evaluable patients experienced complete remission – all of those patients were treated at the middle of three doses tested and all five responders were MRD-negative. Two patients relapsed with CD19/CD22-positive ALL at 265 days and 123 days after infusion with the bispecific CAR-T cells. The other three responders remain in remission at one to eight months post infusion (median of seven months).
While the CD19/CD22 CAR-T cells showed clinical activity with limited and reversible neurotoxicity, Shalabi noted that there was a discrepancy between responses observed in bone marrow and in extramedullary (EM) disease, which suggests limited delivery of CAR-T cells to EM sites. The NCI investigators determined that higher doses may be needed to address this issue, but close monitoring and longer-term follow-up are needed to understand the bispecific cells’ impact on EM disease.
The ongoing study may incorporate a higher dose and patients may be separated by disease burden, since patients with higher disease burden had better responses. Also, the lymphodepleting chemotherapy used prior to CD19/CD22 CAR-T administration is being intensified for patients who previously received CAR-T therapy to overcome immune-mediated resistance. A new manufacturing process, evaluation of CAR-T products to understand T-cell exhaustion, and co-administration of a checkpoint inhibitor to boost efficacy in EM disease also are being considered.
Discussant Chen from UCLA said CD22 is a confirmed target for the treatment of B-cell ALL, but patients treated in studies of CD22-targeting CAR-T therapies have been more likely to stop responding to treatment due to antigen escape. The CD19 functioning was dominant in the bispecific CD19/CD22 CAR-T cells tested by NCI, she noted.
However, Chen also pointed out that T-cell exhaustion or functional issues appear to have been the problem in this study rather than antigen escape, so the NCI needs to assess its manufacturing process to improve bispecific CAR-T cell persistence and function going forward.
[Editor’s Note: This article was updated on 30 April to correct the name of Dana-Farber investigator Caron Jacobson, which previously was reported as Caron Johnson. Scrip regrets the error.]
This article was first published on April 29 in Scrip.
Image: iStock/Dr_Microbe