FDA Halts Phase 2 Trial Exploring Novel CAR-T Cell Therapy

The U.S. Food and Drug Administratioin (FDA) put a hold on a phase 2 study of the immunotherapy CD19-targeted CAR-T cell therapy JCAR015 for adult patients with relapsed or refractory B cell acute lymphoblastic leukemia (ALL), according to a statement from the developer of the therapy, Juno Therapeutics.

The hold is meant to address three deaths that occurred in the trial from cerebral edema, which might be related to the recent addition of fludarabine to the preconditioning regimen used in the study. Two of the deaths occurred within the past week, leading to the halt. The first death was seen in May, although the cause of this event could not be determined.

In response to the hold, Juno plans to omit fludarabine from the trial, which is known as ROCKET, and will provide a complete response to the FDA this week. As part of the complete response, the FDA requested that Juno submit a revised patient informed consent form, investigator brochure, and trial protocol along with a copy of information presented to the agency. The FDA typically reviews this type of information submission within 30 days.

"The FDA placed the ROCKET trial on clinical hold after the occurrence of two deaths last week, which followed the addition of fludarabine to the preconditioning regimen on the trial," Hans Bishop, CEO at Juno Therapeutics, said during a webcast. "We have proposed to the FDA that we continue the trial using cyclophosphamide only preconditioning. In response, the FDA has asked us to submit four specific documents in complete response to the clinical hold. We will plan to submit them this week and the FDA has told us that they will review these documents on an expedited basis."

Fludarabine has a history of eliciting neurotoxicity, which dates back to the mid-1980's. In fact, investigation into fludarabine was nearly halted altogether for patients with leukemia, following the development of delayed central nervous system toxicity. Development was continued only after a maximum tolerated dose was found that compensated for fatal neurotoxicity.

In a trial for patients with chronic lymphocytic leukemia from the mid-1990's, 14 percent of patients treated with fludarabine experienced neurotoxicity, which was primarily grade 1 motor dysfunction. Additionally, in 2011, an analysis of 1,596 patients identified severe leukoencephalopathy in 2.4 percent of those treated with fludarabine prior to hematopoietic stem cell transplantation (HSCT). More recently, in a 2015 case report, a 55-year-old patient with myelofibrosis experienced severe, irreversible neurotoxicity following treatment with a fludarabine conditioning regimen and HSCT.

The phase 2 ROCKET trial was designed to assess JCAR015 following lymphodepleting chemotherapy in 90 adult patients with ALL. The preconditioning regimen consisted of cyclophosphamide at 30 to 60 mg/kg plus 25 to 30 mg/m² of fludarabine over three days.

The majority of patients treated in the trial received preconditioning with cyclophosphamide alone, according to Bishop. There have been no treatment related deaths in those receiving cyclophosphamide alone, and neurotoxicity is within the range of expectations, he noted.

"If the FDA is satisfied with the materials and lifts the clinical hold, we will continue the trial with the aim to make JCAR015 broadly available to adult patients with relapsed or refractory ALL," said Bishop. "The early data with cyclophosphamide precondition alone in the ROCKET trial encourage us that using this approach will be comparable to the phase 1 experience; however, the hold will likely impact our ability to achieve our goal of getting JCAR015 approved as early as 2017."

Prior to this development, JCAR015 had been granted an FDA breakthrough therapy designation for patients with relapsed or refractory B-cell ALL. This designation was based upon early findings showing a high complete response (CR) rate with the CAR T-cell therapy.

In a phase 1 study that included 51 adult patients with ALL, the CR rate with JCAR015 was 77 percent in those with morphologic disease (31 patients) and 90 percent in those with minimal disease (20 patients). At the time of this analysis, the median overall survival (OS) was not yet reached for those in the minimal disease arm. The median OS was nine months for those with morphologic disease.

The first 42 patients enrolled in this study receive cyclophosphamide alone. The remaining nine patients enroll received both fludarabine and cyclophosphamide as preconditioning. Overall, severe cytokine release syndrome (sCRS) was observed in 27 percent of patients and grade 3 or higher neurotoxicity was observed in 29 percent of patients.

More patients in the morphologic disease cohort experienced CRS compared with those in the minimal disease group (42 percent vs 5 percent). When CRS did occur, 12 patients received tocilizumab, 11 received steroids and 10 received both. Three patients with morphologic disease died due to toxicity versus none in the minimal disease cohort. Events leading to death consisted of sepsis/multi-organ failure (2 patients) and seizure with an unknown cause of death (1 patient).

"We remain encourage by the ability of JCAR015 to treat ALL," said Bishop. "The [ASCO] results show a marked improvement over traditional cytotoxic chemotherapy for adult ALL patients."

Given the currently available data, the added toxicity associated with fludarabine appears to be confined to trials exploring JCAR015, according to Bishop. Other trials utilizing the fludarabine/cyclophosphamide preconditioning regimen for Juno’s other CD19-directed CAR T-cell therapies will continue unchanged.

In support of these claims, Bishop announced previously unreported findings from an ongoing phase 1 study looking at the CD19-directed CAR T-cell therapy JCAR017 plus fludarabine/cyclophosphamide preconditioning for adult patients with non-Hodgkin's lymphoma (NHL). In this study, 13 patients were evaluable for safety, with severe neurotoxicity seen in 15 percent of those receiving treatment. Overall, no patients have experienced CRS. In those assessable for efficacy, the objective response rate was 80 percent, which consisted of a 70 percent CR rate.

"Our JCAR017 plans remain on track. We expect to be able to have our first approval in the United States as early as 2018," said Bishop. "Although we have experienced a setback in the ROCKET trial with the clinical hold, based on what we know today, if the FDA is satisfied with the materials we submit in support of our proposal to continue the trial, and the trial comes off hold, we are optimistic that we will achieve results comparable to those in the phase 1 trial. These results should provide the basis for an accelerated approval."

Clinical trials continue to assess JCAR017 and JCAR014, which also targets CD19. In non-Hodgkin lymphoma, a similar cyclophosphamide/fludarabine regimen is being utilized. For pediatric patients with ALL, a lower intensity dose of the preconditioning regimen is under exploration. This lower dose includes cyclophosphamide at 900 to 1000 mg/m² plus fludarabine at 25 to 30 mg/m² over three days.