Home | This Week's Contents  |  C&EN ClassifiedsSearch C&EN Online

Pharmaceutical Business
Pharmaceutical and biotechnology companies seek ways to reduce the time required to discover and develop medicines

Debate Continues Over Drug Development Costs

Biotech firms flaunt a new assertiveness in dealing with acquisitions and alliances

Big deals
Drug developers join forces through acquisitions and collaborations

Related Sites
Pharmaceutical Research & Manufacturers of America (PhRMA)

Tufts Center for the Study of Drug Development

Food & Drug Administration




Albany Molecular Research Inc. (AMRI)

MediChem Life Sciences



Abbott Laboratories


E-mail this article to a friend
Print this article
E-mail the editor
 Table of Contents
 C&EN Classifieds
 News of the Week
 Cover Story
 Editor's Page
 Government & Policy
  Government & Policy
 ACS News
 Digital Briefs
 ACS Comments
 Career & Employment
 Special Reports
 What's That Stuff?
 Pharmaceutical Century

 Hot Articles
 Safety  Letters

 Back Issues

 How to Subscribe
 Subscription Changes
 About C&EN
 Copyright Permission
 E-mail webmaster
January 28, 2002
Volume 80, Number 4
CENEAR 80 4 pp. 27-33
ISSN 0009-2347
[Previous Story] [Next Story]

Pharmaceutical and biotechnology companies seek ways to reduce the time required to discover and develop medicines

PHASE TRANSITION An AtheroGenics scientist prepares analytical samples of a new anti-inflammatory drug candidate formulation, supporting one of the many steps in the drug development process.


It's no secret that developing a new drug is a slow, painstaking process requiring years of research and lengthy clinical trials. The drug industry trade group Pharmaceutical Research & Manufacturers of America (PhRMA) claims that drug companies invest from 12 to 15 years in each new drug. In biopharmaceuticals, the Tufts Center for the Study of Drug Development has determined that, while the number of new products has been increasing steadily, clinical development times have doubled since 1982 to an average of 68 months.

The reasons for longer biopharmaceutical development times, according to Tufts, include an expansion in the use of complicated technologies, the focus on more complex diseases, the demand for higher standards for safety and efficacy, and the need to develop medicines for global markets. All these factors also apply to the greater world of pharmaceutical products.

The slowness in the development process is a result of the numerous steps that a drug must pass through en route to commercialization.

Drug development includes about six-and-a-half years of discovery, preclinical testing, and toxicity studies; one-and-a-half years in Phase I trials to assess safety in healthy volunteers; then two years in Phase II trials with a few hundred patients to evaluate the drug's effectiveness and side effects.

The development process continues with three-and-a-half years in Phase III trials involving thousands of patients and scores of research centers to confirm effectiveness and evaluate long-term effects, then one-and-a-half years of Food & Drug Administration review, where all the clinical trial data are presented. Even after the drug is approved, it may undergo further Phase IV testing so more safety and efficacy data can be collected.

Each year in this protracted process means one more year before a patient can benefit from the drug and one less year of profit for its inventor before its patent expires and generic competitors jump in. As this patent clock winds down, potential sales of well over $1 million are lost for every extra day spent in bringing the drug to market, says Bruce M. Johnson, senior director of developmental chemistry for Atlanta-based biopharmaceutical firm AtheroGenics. And the development process is hugely expensive--about $802 million on average, according to a recent study by the Tufts Center.

Not surprisingly, all players involved in the drug development process are attacking this time expenditure, including big pharmaceutical companies like Novartis and Pfizer, biotechnology companies like AtheroGenics, chemical support companies like Albany Molecular Research Inc. (AMRI) and MediChem Life Sciences, and firms like Covance that provide direct assistance in clinical trials.

First, however, it is helpful to consider the general comments of Ken Getz, president and chief executive officer of CenterWatch, an organization that monitors drug development. In its "Speed Demon" analysis, CenterWatch looked at 1,000 drug approvals since 1987 to identify which companies outperformed in development time and how they did it. Getz says that outperforming companies--Abbott Laboratories, Pfizer, Merck, GlaxoSmithKline, and AstraZeneca--employ strategies that can be generalized to all companies.

SEPARATION A scientist at Albany Molecular's Syracuse Research Center isolates a compound during a scale-up and process research project.
TO SPEED drug development, CenterWatch says, companies should plan projects and prioritize them globally, interact proactively with regulatory agencies, use realistic protocol designs, employ autonomous project teams, and use data management and communications technologies to control the flood of data.

"Fast companies minimize development risk by communicating early on with agencies for input on how to best design the study," Getz says. "Many have engaged research centers as consultants early on to assist in the design of protocols."

Successful firms use new technologies to speed the routing of documents--"a critical strategic step," he says. Also, they attempt to "more strategically and intelligently select places where clinical trials will be conducted, by looking at past performance and using a more systematic selection process."

Patient enrollment is a particularly wide roadblock, accounting for one out of every four delays in CenterWatch's study, Getz says. Companies must reach "a tremendous number of patients before they get a few to come in for initial screening," he notes. This problem results from a combination of factors, including study design, exclusion criteria, and media reports of deaths in clinical trials.

Another major area of drug development delay, he continues, is the data collection and management activity of clinical trials. "It takes a tremendous amount of time and is very labor intensive to copy relevant data into the case report and source documentation, to go to research centers and review information, and to evaluate data for problems with integrity and quality."

For a few drugs, the development process can be accelerated significantly. One is Gleevec (known as Glivec outside the U.S.), which was discovered and developed by Novartis for chronic myeloid leukemia and was approved by FDA in May 2001.

Gleevec was first synthesized and tested in 1992; the first reports on it were published in 1996, and a Phase I clinical trial was initiated in 1998. Years were cut off the development process because the promise of the drug was so great that patients demanded it be made available. Normally, it takes months or years to recruit patients, but Gleevec patients asked to be signed up or they immediately joined the study when approached.

"You have to be steeped in chemistry, and appreciate its finer points, in order to avoid pitfalls like chirality and toxicity."

"The company was able to recruit patients faster, was able to collect and analyze and review data a lot faster," says Gloria Stone, head of public affairs for oncology at Novartis. "Everything we did, we did more quickly without cutting corners on any parameters in the clinical trials. That's what helped us develop the drug faster than other drugs." Gleevec was also approved less than three months after its submission to FDA, a record for drug review and approval.

"Novartis invested a lot of resources, including manpower, to accelerate development that companies normally don't devote to a drug because you don't know if it's going to work," Stone says. For instance, Novartis management committed to accelerated construction of production facilities because of the potential of the drug--there was no therapeutic alternative and patients were doing very well on it--and because their instincts told them the project would succeed, she says.

But not every company has surefire drug candidates, and most must support prospective drugs through other means. In the case of Pfizer, one strategy was the establishment three years ago of a Discovery Technology Center in Cambridge, Mass., that's focused on developing drug discovery tools and disseminating them throughout the company's six research centers.

Rod MacKenzie, vice president of the center, says such an effort is not possible for a firm that does not approach the size of Pfizer, which became the world's largest drug company in 1999 when it acquired Warner-Lambert. "Pfizer decided to leverage its size and resources, not just to speed up drug development, but to make better drugs and more drugs--to solve the conundrum of R&D productivity," he says.

Drugs that make it into clinical trials fail with alarming frequency because of unexpected problems. This attrition--not the resources required for clinical trials--is the real driver behind drug costs, MacKenzie contends, and accounts for the high risk of pharmaceutical projects.

To reduce attrition at later stages and develop high-quality drugs faster, Pfizer concentrates on converting genomic data into chemical leads that it can use to validate new disease targets. Because of its size and because it "covers every conceivable disease of therapeutic interest in the industry," Pfizer is able to take a different tack on drug development, MacKenzie says.

Instead of identifying a disease target and looking for molecules to fit the target, the company works from the genome to identify "drugable" gene families and studies them in parallel to develop small-molecule drugs. By working on the targets simultaneously, the company achieves "terrific savings in screening time and cost in the discovery phase," MacKenzie claims.

Another scale-related aspect of Pfizer's research, he adds, is its huge library of chemical compounds. Often a drug target is identified and time is spent searching unsuccessfully for a matching molecule. Large libraries provide time savings because they offer a better chance of finding an appropriate molecule without toxicity, absorption, or metabolic problems.

IN THIS WAY, the company does not have to give up on an approach in which it has already invested precious time and resources. And by using only high-quality molecules, the company does not waste effort generating data on nondruglike or unsafe molecules. "Only now have we built up a body of knowledge that we can really make sing," MacKenzie says.

Beyond the discovery phase, Pfizer is working to speed a drug's journey through the development process. Before a drug can go into a human, it must pass hurdles such as toxicity and formulation, says Peter B. Corr, senior vice president and head of development at Pfizer. "We are working on ways to predict this early in the process--with attrition before human studies."

For example, he says, a drug company can speed the process by including in Phase I--the safety studies--additional tests to see if the drug is working. "This is a key leverage point," Corr emphasizes. "The question is whether the drug works against the molecular target. It might in preclinical trials, but will it in the human?"

Pfizer uses special methods in the early clinical stages to help predict a molecule's activity before full-blown human trials. One indirect technique is biomarkers--"something that can be measured in the blood or urine or skin saying we have influenced that particular target," Corr says.

If the problem is a tumor, for example, the researcher might do a skin biopsy and relate the results to those found from the tumor biopsy. If the drug affects the skin as it does the tumor, then the skin test is a clue that the drug will be effective.

On a smaller scale, AtheroGenics has identified roadblocks to drug development and ways to speed things up. Its mission is the discovery, development, and commercialization of drugs for the treatment of chronic inflammatory diseases such as atherosclerosis, asthma, and arthritis. The company recently announced Phase II results for its lead product, used for restenosis, the narrowing or closing of a coronary artery that has been opened by angioplasty.

Johnson notes that the speed of drug development depends upon the type of disease addressed. While a bacterial infection is cured in two weeks if an antibiotic works, 10 years might be required to show any improvement in the heart attack rate by taking an atherosclerosis medicine.

If a disease is fatal in six months, the amount of data needed to prove that a drug works is much less than that required for drugs that are taken over long periods. Because AtheroGenics deals with chronic conditions, it, like Pfizer, is looking for a "quick-response indicator" to shorten drug development time, Johnson says.

Rather than following the relatively linear traditional timeline, in which a company does only the amount of work necessary for the next step, Johnson suggests investing heavily up front in time-consuming studies. "The only thing that matters is time--plus being first on the market," he emphasizes. However, the risk is much higher this way.

Johnson also suggests a strategy for avoiding spending valuable time on losers. "We focus on the weaknesses of a new drug instead of its strengths," he says. "The sooner you learn there is a problem with a new drug, the sooner you stop spending money." The objective should be "to kill a drug as soon as possible so you don't spend money on something that will turn out to be unacceptable."

Offering drugmakers outsourcing and collaboration--and a promise of time savings--are contracting companies such as AMRI, a chemistry-based company that conducts R&D projects and partnerships aimed at small-molecule prescription drug discovery and development.

LIKELY CANDIDATE Lee K. Hoong, a medicinal chemist at AtheroGenics, carries out extraction experiments as part of the development of a new arthritis drug.
DRUG COMPANIES are under pressure from Wall Street for 10 to 15% growth, says Lawrence D. Jones, senior vice president of business development for AMRI, "which translates into one to three blockbuster drugs per year that these companies must produce." To get more candidates through the pipeline faster, the companies can either build more internal resources--a costly and time-consuming affair--or go to outside resources like AMRI.

AMRI's goal is to help its clients avoid bringing into clinical trials drugs that are toxic or otherwise unacceptable. It also tries to speed up the discovery process by understanding how structure and functionality affect what a molecule will do.

"We have our own technologies that augment the capability for drug discovery," Jones says. "When a client comes to us with a protein target, we can provide compounds to test it against our vast arrays of natural product libraries, botanical libraries, chemical libraries, and virtual libraries. From these libraries we can choose small-molecule 'hits,' or active compounds, to synthesize and optimize."

AMRI also saves discovery and development time by using combinatorial biocatalysis. Using enzymes and microorganisms to produce unusual compounds provides "a different avenue to create a derivative in a short period of time, in a unique way that organic chemistry couldn't achieve in one chemical step," he says.

AMRI accompanies its clients not only through the discovery phase, but also all the way to commercialization with chemistry and scale-up synthesis. "Making this stuff can be expensive," Jones says. "You need process research to identify cost-effective routes to make the molecule."

The company also smooths the way with FDA by making sure it has made the right compound reproducibly in the purity and strength required.

AMRI attempts to supply resources at the earlier stages, where letting a clunker drug get through the screen of testing and evaluation is a costly mistake. "You have to be steeped in chemistry, and appreciate its finer points, in order to avoid pitfalls like chirality and toxicity," he says.

MediChem is another company offering chemistry support for clinical trials by providing Good Manufacturing Practices quality material to its partners. It says its partners include a majority of the major pharmaceutical and biotech firms. David E. Zembower, senior vice president of chemistry R&D, says the company provides both process R&D and medicinal chemistry services--for example, structural proteomics and protein crystallography. "We work with partners to develop scalable routes to develop candidates," he says.

To get around bottlenecks in the discovery and preclinical phases, MediChem uses in vitro techniques such as growing monolayers of Caco-2 cells to determine which molecules can pass from one side of the cell to the other--a measure that correlates with how well a molecule is absorbed across the intestinal wall. Using techniques like these, "we can engineer out problems before clinical trials," Zembower says.

This concern about a drug's absorption, metabolism, and excretion properties early in the development process marks a shift in philosophy for the industry, Zembower says. Historically, he adds, the emphasis was on finding a precise molecule, putting it in an animal, and then uncovering technical issues such as oral delivery problems. By using tools to predict these problems, precious development time is saved.

Patient enrollment is a particularly wide roadblock, accounting for one out of every four delays in one study.

FURTHERMORE, like AtheroGenics, MediChem is directed toward "killing" bad molecules as soon as possible or reinventing them to circumvent problems--all in the interest of accelerating the development of good drugs.

In early January, the Icelandic biotech firm deCODE Genetics purchased MediChem in order to augment its own drug development capability. Zembower says MediChem will continue offering its services to third parties.

Some chemical-support firms are more oriented toward late drug development. One of these is Covance, whose business is 40% early drug development services and 60% "late stage" services, including Phase II and Phase III clinical trials.

Matt Palazzolo, Covance's vice president of strategic product development, says the company aims to use various screening methods to expedite delivering into clinical trials and to eliminate drugs with a low potential for success. Like MediChem and other companies, Covance approaches each project by trying "to screen out losers really early in the process, before Phase II efficacy trials," he says.

Covance is positioned to do this since it is removed from the internal machinery of a drug company: It has "an unbiased, objective view of that drug's probability of being successful," Palazzolo says. He emphasizes that eliminating a drug early in the development process is as important as developing it quickly through clinical trials, owing to the "astronomical" time and money spent in such trials.

Like Pfizer, Covance uses biomarkers, such as enzymes or other clinical markers that establish surrogate biological activity. This may relate to changes in gene expression or other clinical indicators of biological activity. "Whether it has biological activity in the human is a critical go/no-go decision," Palazzolo says.

One of Covance's thriving businesses is its Central Laboratory Services, which fills a clinical support function. This business handles only clinical-trial samples. At any given time, they will have thousands of sample kits being delivered, especially those in Phase II and III, from investigator sites all over the world.

In fact, Covance has worked with about 25% of all scientists in the world investigating new drugs in clinical trials, says Marietta Henry, vice president of medical affairs and global laboratory medical director at Covance. "We offer consistency of lab methods, which leads to consistency of results, which leads to more efficient statistical analysis for NDA [new-drug application] submission," she says.

The main laboratory, in Indianapolis, handles 3,000 to 4,000 sample kits per day, each containing four to 10 tubes, which may require up to 20 tests for each tube. The kits are delivered early each morning, and results are communicated to the investigator sites by late afternoon--using a workforce of only about 200 people. Everything is bar-coded, and machinery is automated as much as possible to minimize human handling and human error. Tests to be run are specified in the project's protocol so that only those samples agreed on in consultation with FDA are taken and tested.

All this "really shortens the length of the trial," Henry says. "Since 80% of an NDA is laboratory work, the central lab is very important when clients go to submit an NDA to FDA."

Considering the roadblocks in the way of new drugs and the linear way in which the drug discovery and development procedure has been done traditionally, the advances in drug development speed are impressive.

Still, the drug development process continues to lengthen, particularly as regulatory agencies require more patients to prove long-term safety. By reducing the number of drugs that get to clinical trials, and by knocking off time in late-stage clinical trials, pharmaceutical and biotech companies are trying to reverse the trend toward increasingly long development times.

"Companies should not accept the rising cost of drug development as a given," says Raymond V. Gilmartin, chairman, president, and CEO of Merck. "Research-based organizations like ours must become more efficient and focused in order to master the challenges of rising costs."

[Previous Story] [Next Story]



Debate Continues Over Drug Development Costs

There's no question that drug development costs escalate when development time lengthens. But the question of just how much it really costs to bring a drug to market has been argued over the past several years, and it remains a subject of fierce debate.

Perhaps the numbers most widely accepted by economists and the government are those produced by economist Joseph A. DiMasi at the Tufts Center for the Study of Drug Development at Tufts University. On Nov. 30, 2001, Tufts announced an update of DiMasi's 1991 study, which pegged the average cost to develop a new prescription drug at $231 million in 1987 dollars. The new study estimates the average "fully capitalized resource cost" at $802 million in 2000 dollars--a 250% increase, adjusted for inflation, over 11 years.

Not everyone agrees with this figure. "The $802 million is extremely unreasonable," says Richard DiCicco, president and founder of consulting firm Technology Catalysts International. "It doesn't cost that much to develop drugs. But it costs that much to pay for the mistakes of people who have tried to develop drugs."

Two aspects of the Tufts number are particularly controversial: its inclusion of the cost of failures and its inclusion of opportunity cost--the amount of money that could be earned by a comparable alternative investment, which amounted to $399 million in the Tufts study.

"The opportunity cost is a theoretical cost, not a real cost," says Larry Sasich, spokesman for Public Citizen, Ralph Nader's watchdog group. Public Citizen claims that the Tufts figure exaggerates R&D costs not only by including the cost of capital and failed drugs but also by overstating the actual after-tax outlay for development costs and by excluding drugs that receive government support. "The pharmaceutical industry gets an awful lot of tax breaks, like credits for research," Sasich says.

Public Citizen believes that a more accurate figure would include only out-of-pocket expenses and would account for the fact that many drugs receive financial backing from the government sometime during their path to commercialization. The group says the drug industry's main trade group, Pharmaceutical Research & Manufacturers of America (PhRMA), is misleading policymakers and the public to scare them into accepting prices that result in excessive profits.

Public Citizen conducted its own study in which it came up with a cost of less than $240 million to develop a drug. Pharmaceutical companies "foster and nurture the belief that the industry spends a lot on R&D," Sasich says. Companies price drugs according to what the market will bear, not what they spent on R&D, he claims.

Another issue is the extent to which the Tufts report includes marketing costs. "We don't know if marketing goes into development costs," Sasich says. "I imagine there are some marketing costs in the Tufts numbers, but we may never know because it is confidential information."

In fact, Sasich continues, it is difficult to divide some late-stage development costs from marketing costs. For example, the writing and presentation of a scientific paper and the publication of reprints is an important part of the research process but can also be used to help sell the drug.

"Our study gave some idea of the out-of-pocket cost to develop a real drug--what you would see on a tax return," Sasich adds. This number, he says, is what it costs to develop in-house a new drug in the U.S. and get it to the pharmacy shelf.

However, standard accounting practice includes opportunity cost, says accounting firm Ernst & Young. It was hired by PhRMA to rebut claims made by Public Citizen before the latest Tufts report was announced. Opportunity cost takes into account the fact that R&D investment may be very risky. Furthermore, Ernst & Young claims that rather than being lightly taxed, the pharmaceutical industry pays a greater percentage of its revenues in taxes than any other industry.

Equally notable as the size of the Tufts cost estimate is its increase since the 1991 study. DiMasi finds that much of the increase is due to rising clinical-trial costs. At a time when drug development programs are expanding, it is difficult to recruit patients, he says. Furthermore, drug companies are increasingly looking at chronic and degenerative diseases, which require longer trials.

Drug companies and PhRMA consider the Tufts figure to accurately represent the state of pharmaceutical research today. "The $800 million cited by DiMasi is important confirmatory evidence of the increasingly complex nature of developing an innovative new medicine today," said Raymond V. Gilmartin, chairman, president, and chief executive officer at Merck, in a speech after the release of the latest Tufts study.

He added that this estimate means that a "large and vibrant" drug industry is needed to translate new research into new medicines, and that continued innovation depends upon the ability of the industry to make a profit and protect its intellectual property.

The validity of the Tufts figure aside, the cost of drug development is high compared with R&D costs in other industries. In addition, Tufts has found that it takes between 10 and 15 years to develop and approve a new pharmaceutical in the U.S. "The single largest challenge facing drug developers--both pharmaceutical and biotechnology companies--is to contain R&D costs and reduce development times without compromising clinical test design," Tufts Center Director Kenneth I. Kaitin says.

[Previous Story] [Next Story]


Chemical & Engineering News
Copyright © 2002 American Chemical Society

Home | Table of Contents | News of the Week | Cover Story
Business | Government & Policy | Science/Technology
Chemical & Engineering News
Copyright © 2002 American Chemical Society - All Right Reserved
1155 16th Street NW • Washington DC 20036 • (202) 872-4600 • (800) 227-5558

CASChemPortChemCenterPubs Page