[Ip-health] The Ugly Cost Of Developing New Drugs -- Can We Make It Prettier?

Claire Cassedy claire.cassedy at keionline.org
Thu Nov 20 12:07:12 PST 2014


The Ugly Cost Of Developing New Drugs -- Can We Make It Prettier?

11/20/2014 @ 11:15AM

Bernard Munos

You have to salute the bravery of Dr. DiMasi and his colleagues. They had
to know, when they set upon updating their 2003 study on the cost of
developing new drugs, that regardless of the estimate they would produce,
few would like it. That did not fail.

In a preview of a forthcoming publication, the economists from the Tufts
Center for the Study of Drug Development disclosed on Tuesday that it took
$2.6 billion in 2013 to bring a new drug to market. This includes the costs
of failures – 92% of compounds were abandoned or did not get approved – as
well as the cost of capital on the money spent during the lengthy research
process, before it can return a profit. If we add the cost of the
post-approval research required by FDA or needed to add claims to the
label, the figure rises to $2.9 billion.

We will have to wait for the publication of the manuscript for a detailed
analysis, but one can already flag several potential problems:

The data analyzed came from a “randomly selected sample” of compounds
provided by 10 pharmaceutical companies. The selection may have been
random, but what was the population from which the compounds were
extracted? Was it the entire set of drug candidates that entered the clinic
during the study, or was it a smaller set edited to exclude outliers? The
exclusion of outliers is not illegitimate in statistical analysis, but the
problem with pharma is that the economics of the industry are driven by
outliers. The multibillion dollar blockbusters that make up the bulk of the
revenues are outliers, and so are the oversized or protracted trials whose
costs can easily dwarf the average. Since the authors were not involved in
the “random” selection process, they may not even be aware of it, but their
estimate could be biased as a result, through no fault of theirs.

Since 2000, the historic big pharma have produced an average 12 new drugs
(NCEs) per year, with a flat trend. Their annual spending on drug R&D is
currently about $71 billion, an amount that has changed little since 2010.
If we take the authors’ estimate at face value, it explains about 50% of
this spending, which begs the question, “where does the rest of the money
go”? Does it support legitimate activities that foster innovation? Is it
waste? Misclassified costs? Marketing support relabeled as research? The
authors should raise the question because the sheer magnitude of this
unexplained research spending raises the possibility that something big has
been missed. For the industry as a whole, the picture is the same: the
average output since 2000 is 27 new drug per year, also with a flat trend.
The annual R&D spending is $140 billion. This leaves about $60 billion
(43%) of annual unexplained spending.

The average is not very informative, and misses perhaps some key points.
Since 2000, 416 new drugs (NCEs) have been approved by FDA (not including
cosmetics and imaging agents). Of these, 123 have been licensed to
companies that just produced one drug, and most of them did not come close
to spending $1.4 billion out-of-pocket to bring their drug to market. For
example, Chelsea Therapeutics received approval for Northera, a drug for
neurogenic orthostatic hypotension, in 2014. The company, which was founded
in 2002, is public and discloses its R&D spending. Since its creation, it
adds up to $172 million. Likewise, Vanda received approval to market
Hetlioz for sleep-wake-disorder in 2014. But its cumulative R&D spending
since its creation in 2002 is about $276 million. I would submit that the
distribution of the costs of developing new drugs is heavily lopsided, with
the highest figures likely to be 10- or 20-times bigger than the lowest
ones. In such situations, the “average” is not very meaningful, as it is
hardly representative of the population – unlike in a bell-curve situation.
Rather than just showing the average, the authors would do the industry a
great service by disclosing the entire distribution of their figures. We
should not regard high-costs as a fixture of the industry. The fact that
some companies can do the job much cheaper is a call to the high-spenders
to transform themselves in order to deliver affordable drugs.

Lastly, the paper perhaps misses an essential point, that is: is the cost
of producing new drugs a relevant metric? What really matters to all of us,
is how much are we really getting for what we are spending. On that count,
we have a real problem, since the least efficient pharma companies spend
well over $5 billion in R&D for each drug they bring to market, while the
big pharma average stands at $5.8 billion. This is not tenable as it puts a
crushing burden on society. There are encouraging signs, however, that it
is a solvable problem, as illustrated by two companies, Novartis and J&J,
that have worked very hard at transforming themselves. Since 2000, they
have brought to market 20 and 18 new drugs respectively, for a cumulative
pharma R&D spending of $67 billion (Novartis) and $59 billion (J&J). In
both cases, that translates into an R&D spending of about $3.3 billion per
new drug, which is tantalizingly close to the $2.9 billion “average” figure
calculated by the authors. There is reason to hope that their performance
will keep improving as their powerful innovation engines continue to churn
out new drugs at an unprecedented pace. There are indications that other
companies such as GlaxoSmithKline and Bristol-Myers Squibb are also on the

The cost of drug R&D is what it is, and it is ugly. But it does not need to
be so. Small companies do it economically, but their ability to deliver new
drugs on a repeated basis has not been demonstrated. The big companies are
more reliable – at least some of them – but their ponderous legacy
structures are not sustainable. The transformation of the industry around
new, more nimble and effective innovation models is both possible and
inevitable. But it takes leadership and boards that are up to the job. Many
pharma companies – though not all – are now blessed with both. They must
speed their transformation, and make drugs affordable again. Taking on
society is not a viable strategy.

Dr. DiMasi and his colleagues have made a positive contribution to our
understanding of pharma economics. In a way, it was an impossible task, but
they had the courage to take it. Despite all the challenges and questions
that have already been raised, their work is a solid piece of scholarship
that helps us all. I hope they remain engaged in the debate about it.

More information about the Ip-health mailing list