Juno Therapeutics seems optimistic that a clinical hold on its lead CAR-T candidate, following several patient deaths, will delay the project rather than scuppering it altogether. This did not stop investors in the group, and other CAR-T companies, getting the jitters.
If the safety issue turns out to be a class effect it would not just be the end for JCAR015 but also the hot area of CAR-T and its major players Juno and Kite Pharma, as well as being a setback for the likes of Novartis and Celgene, which have placed big bets in this space. Juno was down 32% in early trading today, and all these groups and their shareholders will be watching the story with JCAR015 unfold.
For its part, Juno does not think the clinical hold, spurred by three fatal cases of cerebral oedema in the pivotal Rocket trial in acute lymphoblastic leukaemia, will have a knock-on effect on its pipeline. It pointed out that the side-effect was only seen in those receiving a recently-introduced pre-conditioning regimen of cyclophosphamide plus fludarabine.
Pre-conditioning with chemotherapy is used to reduce the number of immunosuppressive cells, which is intended to improve outcomes with CAR-T.
Patients in Rocket had previously been receiving pre-conditioning treatment of cyclophosphamide alone, and it is to this regimen that Juno hopes to return. But first the FDA needs to approve Juno’s protocol changes and allow it to restart the trial – which the company believes should not take too long.
It plans to submit documents requested by the agency this week; the FDA then has 30 days to review them. The delay will likely push approval of JCAR015 from 2017 to 2018 even if all goes smoothly from now.
Juno is using the cyclophosphamide plus fludarabine pre-conditioning treatment in other trials, notably with its fellow anti-CD19 CARs JCAR014 and JCAR017 – but chief executive Hans Bishop stressed on a conference call that this regimen has been used in over 90 patients to date with no CNS adverse events.
Mr Bishop believes that it is not necessarily fludarabine that is the problem, but the overall pre-conditioning regimen plus JCAR015 at the particular dose studied. JCAR014 and JCAR017 might not have the same issue because, among other differences, the products use different co-stimulatory domains: JCAR015 employs CD28 while JCAR014 and JCAR017 use 4-1BB.
Juno plans to continue its studies of JCAR014 and JCAR017 – which are being trialled in blood cancers including ALL, chronic lymphocytic leukaemia and non-Hodgkin lymphoma – with no modifications.
If the co-stimulatory domain is to blame it might be bad news for Kite: its lead project, KTE-C19, also employs a CD28 co-stimulatory domain and its pivotal Zuma-1 study uses cyclophosphamide and fludarabine pre-treatment.
The fact that Zuma-1 is still ongoing suggests that similar adverse events have not been seen, so perhaps the problem stems from something unique to JCAR015.
But other factors could be at play, including the disease being treated, Juno’s Mr Bishop said. He pointed out that ALL patients have the highest incidence of neurotoxicity, followed by CLL, and NHL patients have the lowest. Zuma-1 is in NHL, but Kite is also studying KTE-C19 in adult ALL in the Zuma-3 trial.
Kite is at the front of the CAR-T race, hoping to file KTE-C19 this year for US approval in 2017. Making up the trio of the most advanced CAR-T players is Novartis, whose CTL019 could be filed in 2017 – later than previously anticipated. The product uses a 4-1BB co-stimulatory element (Therapy focus – How do you solve a problem like CAR-T relapse?, December 22, 2015).
Kite’s stock was down 7% in trading this morning.
It will be some time before it becomes apparent whether the problem is with JCAR015 alone, something linked with the CD28 co-stimulatory domain or an issue with CAR-T in general. Until then, there is not much that Juno and Kite can do but sit tight and wait.
EP Vantage has published a broad overview of the current opportunities and risks in the CAR-T space. A free copy of the report is available by download.