More Juno CAR-T Deaths

By Derek Lowe November 23, 2016

Reminisce back in the summer when Juno Therapeutics disclosed three patient deaths in one of their CAR-T oncology trials and the trial was put on hold? And then the hold was lifted about ten minutes later? OK, three days, but by clinical death standards, that’s pretty prompt. Perhaps too quick, because now there have been more deaths in the same trial. They’ve only treated twelve more patients, and two of them have died of cerebral edema.

Adam Feuerstein has been wondering on Twitter whether the FDA would now shut down Juno’s entire clinical program until someone has a better idea of what’s going on, and I can’t say that he doesn’t have a case there. One could lightly ask, if that doesn’t happen, just what it does take to halt clinical research these days. This particular trial (ROCKET) is halted, of course, and you’d have to be pretty damned optimistic to imagine it ever beginning up again at this point. But what makes Juno’s other trials different from this one, and how do they know that they’ve identified what the problems are and which patients might be at risk? Assuming that such identification is even possible, that is. You’re going to have to have a pretty wooing argument to response such questions, and it’s going to take time and data to put one together. Not good.

Inbetween this and Lilly’s announcement this morning, it has not been a particularly good day for the biopharma industry, or the patients that we’re attempting to help. Things have to look up from here; some good news would be very welcome.

nine comments on «More Juno CAR-T Deaths»

I am wondering how come Juno has these death issues but Kite Pharma seems to be free of them? Reminds of the IL-17 antibodies – Novartis and Eli Lily had no issues with suicidal ideation but Amgen/AZ brodalumab did and that led to both companies abandoning the drug and selling it to Valeant.

That is a very good question, something Juno and the FDA should have asked four months ago. I can tell you Kite has been looking at CFS in some subjects, but as of yet nothing has come up on a treatment related issue. They are wrapping up one trial and ramping up others. I wouldn’t be astonished to see this come out in Q1 next year. It looks good for them as Novartis may not get their application in before Kite.

Ironic how most of their top catches sight of are packed by MDs. The kings of handwaiving.

Theyll most likely just say it was a blip because of a conmed or this was an anomaly and that if they treat another one hundred they wont see another death. They might even say some patients were doing so poorly prior to getting dosed this was an outcome tied to happen. The FDA certainly should shut it down. Look at Mirna, those guys voluntarily shut down their program over infusion reactions and dropped their market cap to the cash they have in the bank. Juno could surely get through if they do the right thing here.

So let me get this straight, the VX-661-107 cohort gets axed because of a lack of efficacy but with no safety issues. But deaths aren’t enough to shut down a trial? WTF.

Also, sorry for bringing up Vertex stuff, just pissed that cohort was cancelled.

This means that the fludaribine story was false; it’s not the conditioning, it’s the therapy. There are a duo of possibilities. One is that it is the particular antibody they used as the basis of the CAR-T; perhaps it cross-reacts with some other epitope that is involved in the toxicity. If that’s the case then Kite’s product may not have the same cross-reactivity. The other possibility is that it’s a class effect and there’s something about Juno’s patient population vs Kite’s that resulted in more patients being affected in Juno’s trial. That would mean you’d begin to see the same effect with Kite as their patient pool expands. I guess only time will tell.

Pretty much all the effective CD19 CARs are causing neurotox in a substantial fraction of patients. I don’t think it’s clear yet why.

At the same time, Juno’s JCAR015 seems to be having more problems than most, including Juno’s other CD19 CAR product!

Keep in mind these different CARs vary in more than just the CD19-specific scFv they use. They also have different costimulatory domains (CD28, 4-1BB), different transmembrane domains, etc. Not hard to imagine that such differences could exacerbate an intrinsic safety risk, or even be a direct contributor that’s unrelated to any possible class effect.

Related movie: