Drug discovery process
will be fundamentally
different in the future
Before joining Washington University in St. Louis, Michael Kinch, PhD,
was managing director of the Center for Molecular Discovery at Yale
University. “A few years ago, to motivate the team I gave them what’s
called a Big Hairy Audacious Goal (a B-HAG),” Kinch says. The B-HAG was
many-headed but one of the heads was to make a collection of all
FDA-approved drugs. The idea was that the collection would serve as a
screening library for drug repurposing.
Kinch thought the first step –
pulling a list of drugs – would be easy; they’d go the FDA and get their
list. But it turned out the FDA doesn’t have a complete list. They had a
running list of all prescribable drugs called the Orange Book. But “all
prescribable drugs” isn’t quite the same thing as all the drugs that
have ever been prescribed, since there are drugs that are no longer
marketed or have been withdrawn because of concerns about safety or
“So what we did was compile a comprehensive list of drugs
approved for use in the U.S.,” Kinch said. “By drugs, I mean the actual
molecules that do the work, called new molecular entities (NMEs), as
opposed to the fillers and the flavors. We went all the way back to
morphine, first sold by Merck in Germany in 1827 and shortly thereafter
in the U.S., and worked our way forward to 2013, closing the database at
the end of that year. “How many do you think there were?” he asks.
were 1,453 – only 1,453 drugs for all of the infectious diseases,
cancers, cardiovascular diseases, skin conditions, neurodegenerative
diseases and other ills the flesh is heir to. “I thought it was rather a
small number, myself,” Kinch said.
“I’m a bit of a workaholic and a dataholic,” Kinch continued, “and this list of 1,400 molecules was
irresistible to me because it raised so many questions. I started to
ask, ‘Who got the approval for the drug, and what was the fate of the
company that got the approval for the drug’? ‘Who did the clinical
trials on the drug, who filed the first patent on the drug, who did the
first publication on the drug, who discovered it in the first place’?”
Over weekends and in the evenings the list turned into a database and
the database grew and grew. “I would sit on the couch at home after
everything calmed down for the night and look at the data. The kids
would say, ‘Oh, look, Daddy’s doing drugs again.’ It was the family
joke.” “Finally I had this ridiculously large database of information,”
he said. “And I began to wonder how to mine it for publication.” He sold
the editor of Drug Discovery Today on a long series of peerreviewed
articles before he had even identified the topics of the articles. “It
was the stupidest thing I’ve ever done,” he said, but also accidentally
brilliant, because it forced him to dig the ore out of his goldmine.
begin a paper, I pick a topic or a question and look at a spreadsheet to
see if there is a story there,” Kinch said. “And you know what, every
darn time there has been a story.” The Drug Discovery Today series has
already led to his participation in both New York Times articles about
the pharmaceutical industry by Economic Scene columnist Eduardo Porter
and in a twopart BBC Radio special about antibiotics. Together Kinch’s
stories add up to one shocking revelation: the R&D infrastructure for
drug development is shrinking, perhaps irreversibly, and our ability to
discover and develop new medicines is being progressively dismantled.
What is to be done? Kinch has come to Washington University in St. Louis
to help find an answer. He regards the cleareyed recognition of the
problem as a first, giant step toward solving it. But he also feels we
have no choice but to do the hard work of defining new models for drug
discovery and that the role universities play must change, and not
necessarily in comfortable ways.
Big Pharma has left the building
the past several decades, Kinch said, the pharmaceutical industry has
managed to dismantle itself. “It’s done a really efficient job of it,”
he said. Starting in the mid-1970s, the industry started outsourcing the
discovery of new drug targets and then the early stages of drug
development. “When you compare the number of drugs that were approved to
their R&D costs,” Kinch said, “you see that the cost per drug was going
through the roof. So they were happy to outsource research.”
pulled out, biotech pulled in,” he said. That continued through the ‘70s
and ‘80s, but then pharmaceutical companies began to buy out
biotechnology companies. The number of biotech companies peaked in 2000,
and today they are often acquired before their first product is
“I used to work for a big biotech company called MedImmune
that everyone thought would remain independent indefinitely. It was
bought by AstraZeneca in 2007,” Kinch said. He sees two worrying trends.
One is that the number of biotech companies has followed a Bell curve.
“What’s scary to me,” he said, “is how symmetrical that curve is. There
really aren’t that many independent biotechs left, and there aren’t that
many entering the field. Where is the drug discovery and early
development going to come from?”
The second worry is the rise of drug
companies with limited R&D capabilities. One example is Valeant
Pharmaceuticals, a Canadian company that now controls as many drugs as
the more familiar Eli Lilly but has a research budget that is one
percent of Eli Lilly’s.
The only research Valeant does is post-approval
trials for the FDA or market research, Kinch said. They don’t do new
“A growing number of drugs,” Kinch said, “are now
controlled by marketing organizations that have little or no internal
drug discovery or development activities. “We all see it coming,” Kinch
said. “And no one has an answer for it.
The pharmaceutical industry has
made rational business decisions. It is unrealistic to expect them to
repopulate their labs; they’re out of it and they’re not going back.”
Because of the shrinkage in the pharmaceutical industry, Kinch estimates
that we have lost more than 75 percent of the “expertise” that supported
By expertise he means scientists with the experience to
modify chemical compounds to improve their efficacy and decrease their
toxicity, the hard part of drug development. While Kinch was at Yale he
hired three computational chemists who had been laid off from Pfizer’s
Groton CT labs.
One was teaching high school chemistry, one was a
50-something postdoctoral associate, and one was working part-time.
“These are not just good people; they are world leaders in their field
with decades of experience,” Kinch said. This is the hidden sting in the
scorpion’s tail because, once gone, there is no quick way to replace
these highly skilled people.
“They learn their craft through a mentoring
relationship and over many years in the lab,” Kinch said. “Some of what
they know is written down, but most of it is passed on in the lab.
“These guys are getting ready to retire or they’ve retired, or given up.
“Who’s going to train the next generation?” he asked. “I started my
professional life as a professor and I can tell you academia doesn’t do
it. Is academia going to take over this role or are we going to find
another way to do it? “I don’t know how this story’s going to end,” he
said, “but right now it’s not looking like it ends well.”
“We need to recognize that this is going to be a completely new
game and that means we basically throw out all past assumptions and
start from scratch,” Kinch said. Although he is still in the listening
phase of his new assignment, he points out the drug landscape is very
uneven these days. Pharmaceutical companies still make money in cancer
drugs and drugs for “orphan” (rare) diseases such as cystic fibrosis.
“Antibiotics and drugs targeting psychiatric, neurological and pain or
itch are at the bottom,” he said. “Few in the industry want to touch
either of those two areas.” Washington University has deep expertise in
the microbiome relevant to infectious diseases and in neurological
diseases such as Alzheimer’s, he said, so there’s an opportunity there.
But the model for research has to be different than it has been in the
past. University scientists are used to publication being the end-point
of a project. We have to take drug candidates further than the
peer-reviewed journals and find ways to ferry them across the “Valley of
Death” between the university lab and the company lab, Kinch said.
in mind,” he added, “that the majority of the US$1.2 to $1.5 billion it
takes to develop a new drug is spent on late-stage clinical trials. We
shouldn’t be making that kind of bet.
But on the front end, you’re
talking millions, not billions, maybe $10 million to discover a class of
drugs and identify a lead candidate.” The market cannot be the only
mechanism by which we meet the need for new drugs, Kinch said, because
the common good and the stockholders’ good are too often disastrously
misaligned. Answers will have to come out of the interaction and
coordination of the government, universities, venture capital firms and
foundations, as well as the private sector.
This all sounds a bit scary, Kinch said. But remember that the research university itself dates back
only to the beginning of the 19th century, and the double-blinded,
placebo controlled trial only until the 1950s and 1960s.
are not static but rather have continually changed as new challenges
have arisen. The collapse of the pharmaceutical industry’s research
infrastructure is our challenge, Kinch said, and we will define the
future of medicine by the way we address it.
of upload: 14th Nov 2014