By Randall Osborne


How much does drug development cost? Uh, a lot.

A better question might be: What does that cost to the extent it can be determined ¿ have to do with how much companies can get away with charging for their products?

An even better one might be: Who is making the cost estimates, and what interests do they have in shooting high or low?

On the heels of a study released late last month by the Tufts Center for Drug Development at Tufts University in Boston, several camps are brawling over drug development numbers and what they mean.

The study says the "full capitalized resource cost" of development per new drug is about $802 million. Since 1991, when Tufts conducted a similar study, the out-of-pocket cost (adjusting for inflation) for each new approved drug jumped by 7.6 percent per year.

Clinical costs, considered alone, went up 11.8 percent yearly, which is more than five times greater than costs for preclinical research and development, the study says.

Put another way, the average cost of new drug development would have risen $231 million (in 1987 dollars) to $318 million (in 2000 dollars), if the rate had only kept up with inflation. But the study came up with the $802 million number.

Not everybody agrees. Not everybody agrees with those who disagree.

It's a great big fight.

The Tufts study is based on surveys of 10 undisclosed drug companies, and data were collected on clinical costs from a randomly chosen sample of investigational compounds. No in-licensed drugs were considered ¿ only those on which all work was done by the company itself. The attrition rate became an important part of the figuring, tricky as it is.

The study even purports to take into account for expenditures capitalized at what Tufts called "an appropriate discount rate," which "should be the expected return that investors forego during development when they invest in pharmaceutical [research and development] instead of an equally risky portfolio of investment securities."

That rate was figured by checking stock market returns and debt-equity ratios over a specified period.

"Capitalizing both the pre-human and clinical testing out-of-pocket cost per approved drug estimates at such a discount rate yield our fully allocated R&D cost estimate," said Tufts, in a prepared release about the study.

Public Citizen, a Washington-based consumer activist group, disputed the numbers, calling them "skewed" toward the very high end. None of the 68 drugs involved in the study got any government support in their development (as many drugs do), Public Citizen said. Also, the group said, the $802 million includes expenses that are tax deductible, as well as "theoretical costs" not incurred by the companies.

Only about half of the study's figure is realistic, Public Citizen claimed, since true out-of-pocket costs total (by the study's own reckoning) about $403 million per drug. The rest is "opportunity cost of capital," which adds what R&D expenditures could be worth if the cash had been invested somewhere else.

The study will be used by drug firms ¿ which provide 65 percent of the Tuft Center's funding ¿ to justify the exorbitant prices of their products and big profits, Public Citizen said.

Next to enter the fray was the Pharmaceutical Research and Manufacturers of America, which hired Ernst & Young to refute Public Citizen's claims. E&Y, in its report, said the assertions by the consumer group about the Tufts allegedly cooked figures "do not stand up to close scrutiny."

The scenario gets even more complicated. In fact, it got more complicated a decade ago.

Tufts conducted a study in 1991 that used methodology similar to the just-publicized one, and found the cost for developing a new drug then to be $231 million. PhRMA did its own calculations, and came up with a figure of $500 million. Public Citizen disputed all that, too, saying the real after-tax cash outlay for each new drug (even factoring the failures), was more like $110 million. And the consumer group said it derived its numbers from yet another study of the Tufts study, this one by the congressional Office of Technology Assessment.

At the eye of the storm is Joseph DiMasi, director of economic analysis at the Tufts Center. He acted as lead investigator in both Tufts studies and said he has been keeping his head low as the parties battle over the findings, but spoke to BioWorld Financial Watch about the main objections.

Calling Public Citizen "apoplectic," DiMasi said he tries "not to get into a full frontal debate with them. You're never going to convince them."

First, he said, the sample was not deliberately taken to exclude drugs boosted by governments.

"Certainly, the [National Institutes of Health] does a tremendous amount of basic research that results in leads," he said, but whether consumers pay for that research as taxpayers or as drug buyers may not matter much in the big picture.

"One could argue that the total cost to society is even greater when you consider these other [NIH] costs," he said.

Anyway, the study is "a random sample of self-originated drugs, developed under the auspices of a single firm," which was more important to the outcome, he said.

Measuring the price of developing in-licensed drugs, "you have to go beyond the firm that's willing to participate into other firms, and track down those costs," DiMasi said. "Certainly, you can generalize from the self-originated drugs to an overall sort of 'resource cost' estimate. There's a reasonably efficient market for licensed-in drugs, and that value will be accounted for through the licensing arrangements. There's no free lunch."

As for including the "opportunity cost of capital," DiMasi noted it's "fairly standard economic and financial practice to [do so]. That's the way investors would look at R&D," which also would be examined in terms of duration.

"It's costlier if expenses are spread over a period of time, instead of being incurred at once," he said. Critics, DiMasi said, "were throwing everything up against the wall" to see if it might stick.

"They kept pushing this $110 million figure in the summer," he said. "They took down opportunity cost completely, and deducted a portion of what's left," since it could be tax deductible. The latter was a mistake, he added.

"They think you can take the corporate income tax rate as a percentage and take that out of whatever estimate we have, and say that what you've got left is the effective cost," DiMasi said. "But we're looking at trends in R&D costs, and tax structures change over time."

He noted that "my numbers are broken down so that one can see all these components" the critics want to subtract from the total.

Still, he said, "if you understand the nature of the corporate income tax, it's intended to be a tax on profits, not on sales with a special dispensation for R&D."

Critics, he added, "view this as corporate welfare, but what's really going on is that the appropriate base for the tax is profits ¿ revenues minus costs. I'm not sure it's right to view [the study] this way, but they view it as an impediment to their agendas."

The take-away point, in DiMasi's view, is that drug prices are not tied directly to development costs anyway, but to the value of the drugs in treating and preventing disease ¿ or at least that pharmaceutical companies don't use such costs to justify what they charge.

"They don't," he said, adding that there was "no funding tied to this particular study," which is expected to appear in a peer-reviewed journal during the next few months, and then will be freely available.

"We don't have any interest in looking at [development costs] any particular way," DiMasi said. "We're seeking the truth here. If we'd thought our best estimate was $80 million instead of $800 million, we would have reported that."