|
||
| |||||||
Productivity
Improvements Needed to Increase Pharmaceutical Profits
Pharmaceutical
companies are seeking better ways to obtain innovative drugs for all their
spending on research and development Despite
the fact that his company can lay claim to being industry leader, when he
gave the keynote address at the FT World Pharmaceutical Conference in
London on November 5, 2001, Pfizer chairman Hank McKinnell complained
that, even as the costs of research and development are spiraling upwards,
the incentives for making that investment appear to be diminishing. Development
times are lengthening, so that it currently takes between 12 and 20 years
for a product to reach the market, he claimed. “Most researchers never
see a product that they have worked on succeed in the market,” McKinnell
pointed out. To
highlight the huge costs involved, he cited the latest Boston Consulting
Group (BCG) estimate that a pharmaceutical company needs to spend €900
million ($810 million) on R&D to bring a product successfully to
market. “That’s why Pfizer is this year spending €5 billion ($4.5
billion) on R&D,” explained McKinnell. Current
Drive to Improve Productivity
With
costs and risks on this scale, there is every justification for the
pharmaceutical companies’ current drive to improve productivity
throughout R&D. Completion of the rough sequencing of the human genome
and the application of genetics to biopharmaceutical R&D have been
seized upon as holding enormous potential. In July 2001, a BCG report, A Revolution in R&D Part II: The Impact of Genetics, claimed
savings, in the best case, of more than $500 million and up to two years
in developing each new drug. “The
potential rewards from applying genetics tools to R&D are enormous,”
said BCG vice president Peter Tollman. “It’s possible that companies
could, under certain plausible assumptions, cut the costs of drug
development by more than half.” So
much for the promise, what is the reality? Backlash
from US society?
GlaxoSmithKline
(GSK) senior vice president Peter Goodfellow warned delegates at the FT
conference that the pharma industry was going to feel a backlash from US
society, if people there appreciated that while National Institutes of
Health spending on drug research had increased from $4.8 billion in 1980
to $20 billion in 2000, the number of new medical entities (NMEs) approved
had gone down to 20 this year. He also pointed out that the industry
itself could question the justification for increased spending on R&D
because CMR International data showed that while R&D costs had risen
slightly more rapidly than worldwide sales between 1988 and 2000, the
output of NMEs had declined from its peak in 1991 Goodfellow
proposed a number of solutions, starting with the financial. Pointing to
the series of mergers that had brought about GSK, he stated, with all the
conviction of someone who knows what he is talking about: “It is almost
inevitable that consolidation will continue.” He
outlined three current technology solutions: genomics, genetics, and
chemistry. As an example of the wealth of data generated by genomics, he
drew attention to some 400 GPCRs (G protein–coupled receptors) in three
families in the human genome, with some 160 unliganded orphans. However,
the question remains as to the practicability of applying all this sort of
data. Pharmacogenetics
is already being used to study the response of individuals to drugs.
Goodfellow cited as an example the paper that Wayne Anderson and
colleagues presented at the European Respiratory Society meeting in 2000
showing the effects of polymorphisms in the promoter region of
5-lipoxygenase and LTC4 synthase on the clinical response of asthmatics to
zafirlukast and fluticasone. These sorts of studies are allowing
pharmaceutical companies to look for genetic variations that affect
patients’ responses to medicines. This type of research will lead to
more effective targeting of medicines by helping physicians to predict
which patients are likely to benefit from a given medicine before it is
prescribed. The
automation of chemistry had produced a tenfold improvement in the
efficiency of generating leads between 1996 and 2000, Goodfellow
maintained. He
reported that with the formation of GSK last year, the company had taken
the opportunity to devise a new structure in an attempt to improve R&D
productivity. “The organization needs to be large where scale is
important—genomics, genetics and high-throughput screening (target
generation to lead identification)—and Phase III clinical trials,”
said Goodfellow. “To promote agility and entrepreneurial spirit, the
organization needs to be small where focus is important—lead
optimization through to ‘proof of concept’ in the clinic.” Despite
all the gains in productivity in the centralized generation of validated
targets and leads, lead optimization is still a cottage industry,
Goodfellow maintained. GSK has therefore organized it in Centres of
Excellence of Drug Discovery (CEDDs), each of which comprises a group of
250–350 scientists under a single management. Each CEDD specializes in a
single therapeutic area or collection of diseases, with skills including
‘lead-optimization’ chemistry and ‘disease-specific’ biology, and
preclinical and clinical expertise. Goodfellow
stressed that there is a serious human resources issue facing the
pharmaceutical industry, as its scientists will need to be better
organized and trained. Automation is increasing and there is a need to
further industrialize the process of drug development, he said. “The
basic problem is that scientists are trained to use their hands and not
their brains,” said Goodfellow. There is therefore a need for the
universities to prepare young scientists for a future that would require
them to use their brains more than their hands, he argued. Reducing
Attrition Rates in Drug Development
Even
with more than 10,000 employees in R&D, AstraZeneca executive director
Claes Wilhelmsson confessed that it was very difficult to do things
quicker, better, and more predictively. But he said that his company
believed that it could reduce the attrition rate for taking hits through
to validated compounds from 90% to 80% in the years ahead. “Research
is moving from conceptual models of cells to mathematical simulation of
cellular systems,” he said. He described how such in-silico
drug discovery complements the traditional in
vitro and in vivo
approaches, although it is still difficult to use predictive models in silico. Wilhelmsson
favored more aggressive target validation, earlier DMPK (drug metabolism
and pharmacokinetics) and toxicity studies, the development of biomarkers,
and predictive modeling of DMPK and toxicity. He
also set out what he called key strategic approaches to speeding up drug
development: ·
manage information flow within an organization and to external and
internal partners; ·
build an optimum platform for launch and commercialization; ·
build processes and organization to promote faster data capture; ·
build alliances to manage communication infrastructure; and ·
build an organization and tools that can quickly globalize
solutions and manage rapid change. “The
challenge is to identify the golden nuggets and discard the rest,” said
Wilhelmsson. To
speed up clinical trials, he saw Web-based data capture (WBDC) as the way
forward. By spending a little more time designing and planning clinical
trials, it should be possible electronically to save much more time in the
phases of execution, analysis, and reporting, he maintained. The designing
and planning would involve setting up the study and the sites, involving
online recruitment and collaborative tools for document exchange. Wilhelmsson
described AstraZeneca’s experience as the first company to use WBDC in
full-scale clinical studies. So far, five WBDC-based studies have been
finalized, with 5,000 patients in more than 15 countries on four
continents. He reported that there had been significant gains: 30% faster
recruitment with quality maintained or improved and study data being
available up to three months ahead of schedule. Leveraging
Scientific Partnerships
Novartis
global research head Paul Herrling points to leveraging scientific
partnerships as a way of enhancing R&D productivity on the basis that
not even the largest pharmaceutical companies can satisfy all their
research-innovation needs in-house. “External collaborations are used to
strengthen dedicated in-house scientific expertise and technology
platforms and maintain them at the leading edge of innovation,” he said.
Currently,
his company spends 30% of its discovery research budget (CHF300 million;
$186 million) on external collaborations. “We hope to get beyond that by
optimizing collaboration,” said Herrling. He gave good examples of
collaborations in which Novartis was engaged: ·
to access new enabling tools and technologies to support in-house
research, it looks to EVOTEC, which provides the complementary EVOscreen
ultrahigh-throughput and bead-based screening systems; ·
to access new disease hypotheses and new targets, it looks to the
Scripps Research Institute and the Dana-Farber Cancer Institute, the
latter being the most important for its oncology research therapeutic
area; and ·
to access therapies and leads, it looks to Vertex to leverage
similarities within the kinase family—in a deal that could earn Vertex
$800 million, if it leads to the successful discovery and development of
eight compounds to Phase II within eight years. Bristol-Myers
Squibb Pharmaceuticals Research Institute executive director Nicholas
Dracopoli was sanguine about the benefits that might accrue from the
application of genomics, pointing to optimistic and pessimistic views
taken by analysts, with Lehman Brothers and McKinsey saying: “. . .
genomics threatens to increase not only the overall associated R&D
costs but also the average cost per new chemical entity (NCE),” and the
Boston Consulting Group saying: “. . . application of disease genetics
and pharmacogenetics together could, in the very best case, save as much
as two-thirds of the current cost to develop a drug and nearly two
years”. Target
Validation Is Major Bottleneck Dracopoli
also identified validation of the many targets generated by genomic and
proteomic methods as the major bottleneck in drug discovery. One way of
short-circuiting this bottleneck is to find appropriate biomarkers to
indicate pharmacologic response to a therapeutic intervention. Dracopoli
highlighted the example of using selected markers to predict the outcome
in ovarian cancer treated with Taxol® (paclitaxel)/Paraplatin®
(carboplatin). The success of this kind of approach opens the way for a
molecular classification of cancer, he said. Dracopoli
claimed already to have evidence that biomarker-driven decisions can
reduce R&D costs by taking a year off target-validation times and by
prioritizing the portfolio of drug candidates, although there is the
danger that such a rational approach would have missed the serendipitous
opportunity that arose to launch Viagra® as an erectile
dysfunction drug, after it failed as a heart disease treatment. So,
whatever the fears that senior managers might have about the justification
for current levels of spending on R&D, they seem to have little choice
but to continue to ramp it up, as new technologies promise improvements in
R&D productivity, and to seek to cut back now would mean rapid loss of
market opportunity in the future. This
article was written by Alex Crawford, (May
1 2002) a regular contributor to
D&MD Newsletter. To view and purchase D&MD reports click here! |
|
|
