The Fiction Behind The Cost of New Drugs

The following column appeared yesterday in The Fiscal Times:

The following column appeared yesterday in The Fiscal Times:

The drug industry usually defends the high price on drugs – the latest cancer therapies are tipping the scales at $100,000 a year – by pointing to the large sums it spends on research and development. It is true that drug firms spend alarger share of their revenue on R&D than most other industries, typically anywhere from 15 to 20 percent of sales. And Eli Lilly ran a Super Bowl ad claiming the cost of developing a new drug has now risen to more than $1.3 billion.

Now Matthew Herper of Forbes Magazine, usually an insightful writer on the drug industry, has raised the stakes. He wrote last week that the cost of developing a new drug had reached breathtaking level of $3.7 billion to $11.8 billion.

Don’t grab two aspirin quite yet. “The Truly Staggering Cost of Developing New Drugs,” read the headline. “The Flat Tax Theory of Drug Industry R&D,” would have been a more appropriate title.

Some background: despite annual R&D expenditures in the billions, large companies like AstraZeneca, Eli Lilly and Merck only develop a handful of new drugs each year. Most new drug candidates fail, usually in the early stages before they even reach clinical trials. But even when they’ve passed safety studies and seem likely to improve outcomes for people with a disease, the potential new drugs will fail their final trials about half the time.

Clearly, drug development is a risky business. Just ask any of the hundreds of start-up biotechnology companies that have burned through their venture capital cash and gone belly-up. When you’ve only got one molecule in development, which is typical of start-ups, you’re banking everything on its success. The simple odds say that half the time you’re going to fail. They are not bad businessmen. It’s science. The drug and biotechnology industries are the most science-intensive and science-driven in the world, and scientific theories sometimes don’t pan out.

For many years, industry supporters have tried to create a metric for measuring the cost of developing a new drug. The Tufts University Center for the Study of Drug Development, which receives industry funding, has developed a methodology that’s widely recognized by the economics profession. Its studies are the basis for the Eli Lilly ad’s claims.

I think they’re wrong, and the $1.3 billion-per-new-drug number is hogwash.

But let me start by saying I’m not a disinterested observer in this debate. I wrote a book claiming the figure was substantially lower (“The $800 Million Pill: The Truth Behind the Cost of New Drugs”) and the title chapter sought to debunk the Tufts methodology.

Here’s the short-hand version of my argument. First, the Tufts economists overestimated total costs by adjusting the multi-year costs for developing any single drug for inflation over the lifetime of its development and treating it like capital investment. That misses the crucial point that R&D isn’t a capital expense. It’s an annual write-off from a company’s top line and thus a deduction from taxable profits. No industry capitalizes R&D costs. The drug industry doesn’t either, including in its accounting statements to the Securities and Exchange Commission. Therefore, it shouldn’t in its public propaganda.

Second, the Tufts researchers failed to account for research that is driven by the marketing needs of the firm, such as coming up with so-called me-too drugs or drugs designed to replace drugs coming off patent. Neither generates new drugs that are any different from what’s already on the market, and should more properly be considered a marketing expense.

Companies also conduct hundreds of clinical trials every year, the most expensive part of drug development, that are not designed to win new approvals or even win new approved indications for old drugs. They are designed by companies to encourage physicians to use their drug. The industry even has a name for this kind of R&D: seeding trials.

My bottom line: after discounting for improper accounting and marketing-driven R&D, the cost of drug development on average was probably less than half of what Tufts claimed, somewhere around $250 million at that time (2004).

So how did Forbes’ Herper get his outlandish numbers? He totaled up individual company R&D expenditures over the past 14 years (the expected patent life of a new medicine) and divided it by the number of new drugs brought to market by each company during that time period. AstraZeneca got the booby prize: it spent $59 billion in inflation-adjusted dollars on R&D and only pushed five new pills through the Food and Drug Administration.

That’s “as much as the top-selling medicine ever generated in annual sales,” he lamented. “At $12 billion per drug, inventing medicines is a pretty unsustainable business.” You’d never know through his accounting that AstraZeneca earned $70 billion in non-inflation-adjusted dollars since 1997, which is far more than it invested in R&D. Indeed, its after-tax profit margin in 2011 – let me repeat, its profits after all taxes – was over $10 billion or 30 percent of sales, nearly twice what it spent on R&D.

Where did Herper go wrong? First he used inflation-adjusted numbers in his R&D calculation. That unjustifiably inflates the total as I explained earlier. R&D is a cost, not an investment.

Then he divided this total by the number of new, approved drugs developed by the firm. This would include, for instance, Nexium, an acid indigestion pill that the company developed to replace Prilosec, now sold over-the-counter. They are exactly the same medicine with a minor chemical tweak. In fact, half of what’s in Prilosec is Nexium, and both halves work exactly the same way.  Yet insurers, including Medicare, paid $6 billion for Nexium last year, when the people in need of acid indigestion relief could have had an over-the-counter alternative for a fraction of the price.

The latest drug from AstraZeneca is Brilinta, a new anticoagulant in an already crowded field. The FDA deemed it a “standard” drug not worthy of rapid review, i.e., not a significant medical advance. In other words, a lot of the company’s R&D is spent on meeting the financial needs of the firm – not the health needs of society.

Herper also erred in dividing the total R&D expenses by only five new drugs. How about all the post-marketing trials paid for out of the company’s R&D budgets that were designed to get more doctors to prescribe all the medicines in its portfolio? The governments’ database on clinical trials contains 40 Phase IV clinical trials for Nexium (these are trials done after a drug has been approved) sponsored by AstraZeneca. They included such need-to-know studies as how well does reducing acid indigestion help people sleep at night, and does it help people with chronic coughing. These trials, when they appear in the journals that are read by physicians who see people with those conditions, are nothing more than a way of getting doctors to say, “Hey, take this, it might help.”

The company also made more than two dozen submissions to the FDA last year for labeling revisions and expanded indications for existing drugs, all of which were supported by science produced by its R&D department. These post-marketing changes, too, often depend on expensive clinical trials.

The folks in the R&D departments at major drug companies and biotech start-ups spend lots of money doing a multitude of tasks – including new drug development, all of which are part of what it takes to run a for-profit company in the modern drug industry. But never forget that many of those tasks have little to do with what it “costs” to develop a new drug, which is a metric more useful to companies trying to justify the high prices of their latest medicines than a society trying to figure out how it is going to afford them.