Skip to Content

Pfizer Bails on PCSK9 antibodies

Well, no sooner do I mention that Pfizer has a PCSK9 antibody coming along in the clinic then they don’t. Not any more. The company has announced that they’re discontinuing bococizumab, which comes as a surprise to most everyone, since they’d been spending substantial time and money on it. The company has concluded that it’s “not likely to provide value to patients, physicians, or shareholders“, because it appears that the immune response to the antibody gradually decreases its effectiveness. They’re also seeing more injection-site problems, which goes along with the first observation. The problems, they say, are definitely worse than the other agents in this class.

And that really leaves them nowhere to go on this one. If you’ve proven that the efficacy is lower and the side effects are higher, versus two direct competitors who are already on the market, the game is most surely over. They had, though, made it through at least four of the planned six (!) Phase III trials, eventually to include 17,000 patients, which should tell you just what a joy it is to work in the cardiovascular field. Always has been one, to be honest about it, and Pfizer knows that as well as anyone.

11 comments on “Pfizer Bails on PCSK9 antibodies”

  1. SedatedFMS says:

    Oh well.

  2. LiqC says:

    Might this have something to do with anti-drug antibodies elicited by humanized Pfizer product (-zumab), as opposed to fully human Amgen and Regeneron products?

    1. johnnyboy says:

      Although logically you might think that, in fact there is no apparent difference in the overall incidence of anti-drug antibody reactions between humanized and fully humanized antibodies on the market. Every antibody is different, and there is still a bit of trial and error involved.

  3. Anon says:

    Better to stop if you know the market is not there anymore than spend a few hundred million to develop another me-too. Good decision.

  4. mallam says:

    Big risk for big reward….or not.

  5. Barry says:

    perhaps Pfizer has a small-molecule moving forward against PCSK9? The Cate paper wouldn’t have seen the light of day if that lead were going forward, but there are other points at which a small molecule could act on PCSK9 (or LDLr).
    With our without a small-molecule in play, a mAb with worse side-effects and less efficacy than what’s already out there has no place in the market, no place in the development budget.

  6. bhip says:

    It was the right decision based on the data. That said, it makes you wonder how/if the elevated antigenicity potential of the ab wasn’t picked up preclinically (best) or in Ph2 trials (better) prior to the bigger spend in Ph3.

  7. Chrispy says:

    I wonder about humanized antibodies. My understanding is that you basically swap the mouse CDRs into the human framework, at which point you typically lose binding altogether and have to re-introduce mouse amino acids to get it back. It is a long negotiation with the molecule, which at the end of the process will have more or less murine sequence in it. I could imagine that some antibodies will not be cooperative and would require considerable sequence reversions to mouse before they bind well enough. The antibodies now in the clinic seem to about 50/50 humanized mouse vs. humanized antibodies, and we should have an idea whether the humanized mice generate less immunogenic antibodies. Is this known?

  8. fluorotone says:

    Replying to Chrispy, bhip, johnnyboy and LiqC (all on same topic)-

    Not sure what mAb platform PFE/Rinat was using, whether it was fully human or not. But they also were messing with the Fc to increase half-life/PK via histidine substitution, so the increased immunogenicity may have been predictable, or testable. I don’t know if the mAb in this paper what what was advanced into the clinic, but it’s a free download on the Fc engineering:

    http://www.jbc.org/content/287/14/11090.short

    bhip – pre-clinical (non-human) studies aren’t going to give a good read on what awaits in humans, for immunogenicity, or even efficacy, in some rare and very expensive, cases. Cyno model is not 100% predictable, because this stuff is really complicated!

    I think in this case, the combination of bad molecule choice and the stormy seas ahead for PCSK9 inhibitors made this one a no-brainer, even for Pfizer.

  9. Mike B. says:

    Did glycoforms have any role to play? Change one monosaccharide and you can radically alter PK and possibly even immunogenecity.

Leave a Reply

Your email address will not be published. Required fields are marked *

Time limit is exhausted. Please reload CAPTCHA.