Skip to Content

Business and Markets

Infinity’s Research, From Someone Who Was There


Here’s a look from the inside at Infinity Pharmaceuticals, whose R&D operations hit the wall just recently. Keith Robison knows the company and its targets well, having worked there until about five years ago, and he provides a perspective that some others in the readership will have experienced as well:

So it was a very great shock to many, including myself, when the big trial (pancreatic adenocarcinoma) for the hedgehog program was stopped after interim analysis showed that not only was it not going to succeed, but mortality was worse in the treatment arm.  I ran into the head of biology at a networking event just after the trial halt, and he was still in complete shock.  Unsurprising, as the pre-clinical data was excellent, including engineered mouse models.  The trial designed to provide early market approval for that compound, in chondrosarcoma, followed soon after, stopped after an interim analysis showed success was unlikely.  A third trial was soon dropped.  The HSP90 inhibitor failed to impress, and was also dropped.
Going into clinical science, one should have no illusions.  Certainly any I had were dispelled before I was even officially employed by Infinity.  I had just sent in an email to the hiring manager saying I’d take the position, when he called me back.  Yes, I still had a job.  WHAAATTTT?  I hadn’t seen the news that Infinity had halted their Phase III trial for the HSP90 inhibitor in gastrointestinal stromal tumor (GIST), which had been a path to marketing a drug soon.  Infinity survived that setback, as did my employment (a friend in the clinical space would have to wait longer).  Infinity then spent a lot of effort showing that the deaths in the GIST trial were probably due to liver toxicity — this is a major issue with HSP90 inhibition and many of these patients had extensively resected livers due to metastases.  So Infinity pushed on with other trials for HSP90 inhibition.
We get ambushed all the time like this in this business – it happens to the small companies and to the big ones, to the programs that aren’t very well thought out (especially to those), but also to the ones that look to have everything lined up. The fundamental fact about doing biopharma R&D is that we don’t know enough about human biology and disease to ever feel really secure that we’re doing the right thing for the right reasons. It’s still startling to me, the number of people outside the field (journalists, politicians, activists, and just plain members of the general public) who have no idea that this is the case.
We have to take our best shots, with the best information we have, and then we just have to hold on and hope for the best. That doesn’t always work. But it’s what we have.

9 comments on “Infinity’s Research, From Someone Who Was There”

  1. bhip says:

    Animal models in oncology (or insert therapy are here) are largely exercises in PK/PD- at this dose, are drug levels sufficient to inhibit growth/increase survival (insert PD endpoint here).
    If the effects on tumor growth/survival are spectacular (you know it when you see it), the odds of success in the clinic simply improve somewhat.
    Don’t waste the funds & patent life (should it ultimately be relevant) mucking about with multiple, likely meaningless models. If you think it could work (good pk, interesting/relevant MOA, human mutations, etc) & the therapeutic window seems acceptable, spin the wheel & hope for the best…

    1. K says:

      That was certainly true of preclinical oncology models in the past. But it is increasingly not true thanks to the advent of patient-derived xenograft (PDX) models in which actual patient tumor biopsies are implanted into mice, rather than laboratory-adapted cancer cell lines.

      The body of literature suggesting that these new models accurately predict the clinical response to drugs in an indication, or in individual patients, is substantial and continuing to grow. Obviously, we can’t prove that will be true for every indication and every drug class, but there is hope that oncology will go from being an therapeutic area with some of the worst preclinical models, to having some of the best.

      Which is not to say this technology is a panacea to all our problems in oncology (and, incidentally, it is much more difficult to apply in the immunoncology space that is so hot right now due to the immune-deficient nature of most of these models), but this is a remarkable advancement over what was the state of the art only 10 years ago.

  2. Former Infinite says:

    I spent several years as a scientist at Infinity, including the early days before Julian Adams came on board, and worked on several early stage programs. I’ve got to say that I have great memories of working there – most of my colleagues were very intelligent, hard working and pleasant to work with. There was much less BS there than at most other organizations, and a real communal sense of purpose and desire to excel. The level of scientific rigour was high – I saw very little obfuscation and hand waving when it came to interpreting data and designing the right experiments to move things forward. I’m very saddened by the recent news and wish the best for the current staff. Drug discovery is a harsh mistress.

  3. tally ho says:

    I’ve always been leery of chaperone protein inhibitors (e.g. HSP90), since by functional design chaperone proteins have many interactions with a wide variety of protein partners – seems like a dodgy target target class. Can any experts comment on the on-target tox/therapeutic window for HSP90 inhibitors, assuming a lower bar for oncology indications?

    1. hn says:

      I feel the same way. Is there evidence that known hsp90 inhibitors work by other mechanisms?

  4. Curious Wavefunction says:

    I would second the quip about Julian Adams who in my opinion evidences the right combination of critical skepticism and hopeful determination. He would be the first one to poke holes in questionable ideas, but also the first to suggest a way forward. A really valuable guy to have in any organization; I hope he finds another challenging position soon. Best wishes to all affected.

  5. The point Derek makes regarding the misconceptions about drug development is an important one, and resonates with something exGlaxoid said in yesterday’s migraine comments: People want and expect pills (or injections) for everything.

    We live in an age of wonders and it seems perhaps people have been conditioned to think that everything is doable just given a little time and research. There’s a new iPhone every year! Computers keep getting smaller and more powerful! For many people, I’d conjecture, there is a belief that the engineering prowess that goes into making a iPhone and creating new products regularly should be applicable to other fields like drug development. From the outside, these technology/science intensive enterprises look the same to them, and especially given the way some in the media hype every new academic mouse study (New target for Alzheimer’s! Never mentioning all the studies were done in cell cultures and mice). Certainly fuel for the conspiracy minded who believe biopharma is keeping cures away from the public and jacking up prices for no reason other than sheer, malevolent profiteering.

    So what can be done to make people understand just how difficult the enterprise is? It seems creating a somewhat more savvy general population is one (but only one) element of improving the overall reputation of the drug development industry.

    1. Pennpenn says:

      I’d say it’s largely a problem of it being very difficult to make people understand something they don’t want to understand or accept.

      1. anonao says:

        The fact that large pharma fire people while spending fortune in advertising and stocking billions of dollars in tax heavens is not helping either!
        When news are companies looking to reduce their tax and not about drug discovery the public will tend to see them as “evil” and not companies looking to find cures and make a profit out of it to reinvest in research.

Comments are closed.