‘We were shocked’: CAR-T rapidly reduces size of brain tumors in phase 1 trial
Click Here to Manage Email Alerts
Key takeaways:
- Novel CAR-T targeting both EGFR and IL13Ra2 reduced tumor sizes within 48 hours in patients with glioblastoma.
- No grade 4 or 5 neurotoxicity occurred during the study.
A novel bivalent-targeting, intrathecally delivered chimeric antigen receptor T-cell therapy decreased tumor size in patients with recurrent glioblastoma, according to results from a phase 1 study published in Nature Medicine.
All six patients treated in the trial had their tumors shrink within two days of receiving the CAR-T — which targets both epidermal growth factor receptor (EGFR) and interleukin-13 receptor alpha-2 (IL13Ra2) tumor-associated antigens — and one who has maintained stable disease for 7 months and counting.
Despite the positive signals, none of the patients treated achieved an objective radiographic response based on the study’s criteria, according to researchers.
“These results exceeded our expectations,” Stephen J. Bagley, MD, MSCE, section chief of neuro-oncology at Penn Medicine, told Healio. “We’ve done two trials previously and we didn’t see tumors shrink in any of the patients — not a single one,” he added. “Here, the very first patient we treated [had] an MRI scan 24 hours after the injection because the[y] experience[ed] eurotoxicity. The MRI was actually obtained for safety monitoring reasons, and we were shocked to see that the tumor had almost completely regressed in 24 hours.”
Trial and error
The impetus Bagley and colleagues had to develop this dual-target CAR-T came from the unsuccessful trials Bagley referenced.
Median OS for adults with recurrent glioblastoma is less than a year, creating an urgent need for this patient population, according to background information provided by researchers.
They attempted to alleviate that burden in their previous trials, which targeted EGFRvIII. They did not achieve efficacy but did produce tissue to analyze.
“One of the key reasons we think [the trials did not work] was antigen escape,” Bagley said. “In our first trial, we were able to compare pre- vs. post-CAR T cell tumor tissue in all seven patients. In all cases, EGFRvIII expression went down post-CAR-T, indicating that we had reduced the tumor that was expressing that antigen, but EGFR amplification levels actually stayed the same or went up in some cases. We were kind of reducing target, but the tumor was still growing out antigen-negative cells.”
Since EGFR alterations occur in 50% to 60% of patients with glioblastoma, and IL13Ra2 appears in roughly 75%, attacking both could give the cancer care team a fighting chance, Bagley said.
“You’re statistically increasing the likelihood that a given patient is going to have at least one of our targets, if not both,” he added.
Phase 1 results
Researchers tested the CAR-T on six patients aged 18 years or older with recurrent, isocitrate dehydrogenase wild-type glioblastoma and positive for wild-type EGFR.
Additionally, they administered the CAR-T intrathecally directly into the spinal fluid, using an Ommaya reservoir. They had given IV CAR-T in their previous trials.
“When we delivered CAR-T cells into the blood, we’ve showed that some of those cells do actually make their way from the blood into the brain tissue, but it’s probably just a subset of the cells,” Bagley said. “With intrathecal delivery, you’re essentially delivering cells directly into the central nervous system. You are making it much easier for the CAR T cells to encounter tumor cells more readily and more quickly.”
The dual-targeting and a new delivery system produced rapid results.
Patient 1 maintained tumor regression for 33 days. Patient 2 died, but not from treatment, rather hydrocephalus, which they did not want treatment for. They had 2 months of stable disease before that occurred.
Patient 3 is at 7 months of no progression, Bagley said.
Patient 4 had 3 months of tumor stability and patients 5 and 6 had 1 month at the writing of the study.
“We’re getting a hint of durability in some of our patients, but it’s early days,” Bagley said. “I think all we can say so far is that there is clearly a signal that we can see.”
Every study participant developed neurotoxicity within 72 hours of treatment, sometimes as early as 8 hours, including encephalopathy. However, these results encouraged Bagley about the treatment’s efficacy.
“We don’t want patients to be sick, obviously,” he said. “We don’t want them to experience symptoms, but the wealth of years of experience giving CAR T cells to patients with hematologic malignancies like leukemia and lymphoma have shown us that when these cells are active and they’re actually exerting meaningful antitumor effects, it often comes with substantial toxicity.”
Every participant developed immune effector cell-associated neurotoxicity syndrome (ICANS), although researchers had to create an adapted ICANS grading system because patients with glioblastoma can already have issues such as speech problems and motor deficits.
Researchers did not observe any grade 4 or grade 5 toxicities. Grade 3 adverse events included fatigue, decreased lymphocyte count, CAR neurotoxicity, skin ulceration, anorexia, muscle weakness and hypoxia.
“The neurotoxicity was manageable and reversible in all six patients,” Bagley said.
Future implications
The current protocol has approval for 18 participants, and researchers are hoping to get more answers on durability of response and how that affects OS, as well as why the CAR-T works better in some patients more than others.
They are gathering answers to these questions rapidly thanks to the intrathecal delivery.
“We can easily pull out spinal fluid for biomarker and correlative studies,” Bagley said. “We can do this serially, on a very frequent basis, which is what we’ve been doing. We can track the signal of the CAR T cells by [polymerase chain reaction] in the spinal fluid over time. We see these beautiful peaks in every single patient we’ve treated — getting really high-level expansion in that first week or so — and we’re finding that the cells persist in the spinal fluid out to 28 days post CAR-T, and in some cases much longer.
Bagley hopes this trial can benefit the medical community in multiple ways. He believes multivalent targeting and local delivery could help researchers investigating treatments of other solid tumors.
And for glioblastoma, this trial may ultimately lead to more effective therapies early in the process.
“There are a lot of challenges with treating patients when the tumor has already grown back,” he said. “Their clinical condition isn’t as good. They’re often more symptomatic. Time is of the essence, and there’s a lot to do in these CAR-T studies. You have to collect the cells by leukapheresis, manufacture the cells and administer them. If we were able to do all of that earlier in the disease course, when things are not as urgent for the patient and the patient is in better clinical condition, I think there’s a higher likelihood for long-term success.”
For more information:
Stephen J. Bagley, MD, MSCE, can be reached at sbagley@pennmedicine.upenn.edu.
Collapse
Click Here to Manage Email Alerts