Translationing new ideas into novel drugs

Re: The Importance of New Companies for Drug Discovery:”

A recent report published in Nature finds that a majority of novel drugs are the result of research at universities and new biotechs.  This is good news, showing that the system is working.  Academic researchers with more freedom to follow the road less traveled, and the biotechs that often spring from their discoveries, are responsible for many or most of the novel therapeutic agencies.  The only problem with this finding is that this isn’t the problem.  It is critical that universities continue to receive sufficient funding to allow for a broad range of new ideas to be followed up.  But the critical issue in translating new ideas into new therapies isn’t the lack of new ideas. The volume of new research findings is astounding.  In fact, the volume is overwhelming!  And what makes this a particularly acute problem is the minimal depth of information available about these new ideas and novel targets.  There is still too much basic translational work to be done to separate the wheat from the chaff, and besides being a very expensive process, it also means that promising ideas can go overlooked for years.

The byword in the drug discovery and development business is robustness, specifically reproducibility.  Unfortunately reproducing research findings is almost a curse for academic and many biotech investigators.  A graduate student or postdoc who has gotten some new exciting data doesn’t immediately set up to run the experiment again.  What they do is run to publish it, and also to use some additional technique to approach the same question.  Of course, it is important to make data public, and as soon as reasonable.  It is important for graduate students and postdocs to gain working knowledge and experience with multiple techniques.  But often, very often, the original data cannot be easily reproduced.  Often, there are unrecognized experimental conditions or details that render reproduction tedious at best.  And yet, without reproducibility there can be no drug discovery project.  The FDA, at least, frowns on results that were only observed once.

And yet, it isn’t the perceived role of the basic science investigator to demonstrate exactly how to reproduce their results.  This is left to other investigators in other labs who may want to follow up on the original exciting results.  The primary job of the discoverer, the explorer, is to instigate new studies.  Responsibility for getting things right is not a part of basic science.  One has only to look at many of the most published psychopharmacologists to see that their batting record isn’t necessarily so high; their value was in inspiring more people to join the new game.  No, I don’t mean they can make results up. But, their perceived role is to publicize new data, new ideas, and new hypotheses;  not to show that they can get the same result over and over again.

Unfortunately this leaves pharmaceutical/translational scientists in a lurch.  How can one prioritize which, among the huge number of exciting new ideas and data, to assign limited resources to follow up when robustness has not been demonstrated.  I can vividly recall reading through 20-30 new idea statements each month from brilliant scientists at a world-renowned university, and generally have to tell them that there wasn’t enough information to follow up on.  This took a lot of my time, took it away from ongoing discovery projects, and certainly made no friends among the academic stars.

An interim function needs to be installed; one whose goals involve standardizing some assays enough so that they are carried out essentially the same in every lab and in setting out to reproduce potentially exciting results in these standardized tests.  Some effort in this direction has been ongoing, most critically in the realm of standard behavioral pharmacology assays, but this needs to become one of the highest priorities in pharma and in academic centers.  No longer can an expert electrophysiologist decide one day that they need to run the Morris water maze and expect anyone to accept their single study results.  Assays must be consistent in the results they give and they need to be sensitive to specific challenges – gold standard compounds, for example.  A new and exciting result needs to be reproduced by the original investigator, but now blinded to conditions, and then by another lab member, that is, using the same set-up, and also blinded to conditions.  Or there needs to be a translational institute that focuses on reproducing results.  Nothing less should be acceptable.

So, how do pharmaceutical companies decide what new ideas to follow up on.  Well, they either work on second- and third-generation projects related to drugs already known to work, reproducibly, or they find new uses for known compounds.  Or they go with their gut… and then suffer their departments being right-sized or given “new direction” after two or three years without clear progress.

A new type of research organization needs to come into existence.  At one time, pharmaceutical companies with “money to burn” could allow research departments to take a lot of chances on new ideas.  Those days are gone.  Perhaps what is needed is a multi-company reproducibility/research organization that can plow through a large number of new ideas, while also working to standardize certain assays.  Being funded by multiple entities would reduce the financial risk for all.  One might snipe at this saying that companies only want to work on novel ideas, ones that aren’t being worked on by other companies.   Anyone who has been in the pharma business more than five years knows this argument is a crock…   Companies want to work on ideas they believe others also see as valuable – robust.   Value comes from the unique properties of specific compounds each company finds.  And, companies have different strengths, in biology and chemistry, and so each project will really only appeal to a fairly small number of companies.