JCAR015 in ALL: A Root-Cause Investigation

Last year, a study of Juno Therapeutics’ investigational CD19-targeting chimeric antigen receptor (CAR) T-cell therapy JCAR015 in adult patients with relapsed or refractory acute lymphoblastic leukemia (ALL) ground to a halt after five cases of fatal neurotoxicity occurred. The company probed possible reasons behind these deaths, and Chief Medical Officer Mark Gilbert, MD, presented a summary of their “root-cause investigation” during the Society for Immunotherapy of Cancer's 2017 annual meeting in National Harbor, MD, in November.

The phase II ROCKET trial was placed on clinical hold by the FDA in July 2016, “when we first saw a shift in the rate and severity of cerebral edema,” Gilbert said. Three patients died, and Juno surmised that a protocol change—adding fludarabine to cyclophosphamide during lymphodepletion, which prepares patients for CAR T-cell therapy—might have been to blame. Fludarabine was removed from the regimen, and the FDA allowed ROCKET to proceed. However, after two more deaths due to cerebral edema occurred, the company elected to shelve JCAR015.

One of the key findings from Juno's retrospective analysis, Gilbert noted, was that all five patients who died experienced rapid, early expansion of their modified CAR-bearing T cells within a week of being infused, rather than the typical time frame of 12 to 14 days. High levels of the CD8+ subtype and, consequently, a sharp spike in cytokines such as IL2 and TNFα produced by these cytotoxic cells, were significantly correlated with fatal brain swelling, Gilbert added.

Autopsy results from two patients showed a complete breakdown of the blood–brain barrier, possibly due to this inflammatory cytokine surge, which—in line with findings from an unrelated study—may have driven their cerebral edema (Cancer Discov 2017;7:1404–19). “It wasn't what I expected,” said ROCKET's lead investigator Daniel DeAngelo, MD, PhD, of Dana-Farber Cancer Institute in Boston, MA. He had thought culprit would be CAR T cells or other immune cells infiltrating the brain.

Gilbert also reported that the patients who died were younger than 30 and had received fewer prior therapies. As well, they had higher baseline levels of IL15, another T-cell growth factor. Going on to look at ROCKET's population as a whole, he noted that “those with Philadelphia chromosome–positive disease seemed to have a much better outcome with regard to the overall risk of neurotoxicity—an interesting signature we're trying to confirm through a larger study.”

However, Gilbert cautioned that findings from Juno's internal investigation shouldn't be considered definitive, because “this wasn't a planned analysis, but exploratory in nature.”

The company is moving ahead with additional CAR T-cell therapies, including evaluating JCAR017 in patients with relapsed or refractory non–Hodgkin lymphoma. Lessons learned from ROCKET have been duly applied: For instance, patients no longer receive infusions of a product with wide variability in its T-cell composition. Instead, JCAR017 consists of a fixed ratio of CD8+ cells as well as the CD4+ “helper” subtype. Its CAR construct also has 4-1BB, rather than CD28, as the costimulatory domain, which may better control the pace of cell proliferation. Eventually, Juno hopes to test this therapy in ALL, Gilbert said.

After JCAR015′s failure, “the question was, ‘Should we even try CAR T-cell therapy in adult ALL?’” DeAngelo remarked. “The reality is that we saw a high, confirmed complete remission rate [47%]. We need to better control the product and also focus on patient characteristics that might predispose to neurotoxicity, but I still think this is the way to go.” –Alissa Poh

For more news on cancer research, visit Cancer Discovery online at http://cancerdiscovery.aacrjournals.org/content/early/by/section.