TRIAL

Home MyAAJ Join AAJ Username/Password Find Lawyer Shopping Cart Links Help

Search Tips
Printable Version

Feature
July 2006 | Volume 42, Issue 7

The truth about the drug companies

Drug companies tell us that most of their enormous earnings are poured back into developing new and better drugs. But there’s far more imitation than innovation in drug research. That may be good for business, but it’s a dangerous boondoggle for consumers.

Marcia Angell

The pharmaceutical industry is very different from the image it projects in its public relations. Americans spend about $200 billion a year on prescription drugs—the fastest-growing component of our health care bill, which itself is growing at an alarming rate. The soaring expenditures are due to three factors: greater overall use of prescription drugs (that is, volume); higher prices for new drugs; and rapid price escalation of drugs already on the market, particularly top-selling ones. In fact, prices for the 200 top-selling drugs are now rising more than three times faster than the inflation rate.

The industry, while acknowledging the high prices, portrays this as something of a success story. We get our money’s worth, the drug companies say, in the form of a steady stream of miracle drugs that improve and extend lives and may even save money in the long run because they avert more expensive forms of medical care, like hospitalization. Skyrocketing prices, they say, are necessary to cover their research and development (R&D) costs—a claim that implies that they spend most of their money on R&D and that after they pay for it, they have only modest profits. Curbing prices, they say, would choke off R&D and stifle innovation.

Sounds good, but is any of it true? To answer that question, I will address three other questions:

How much do drug companies spend on R&D?
How much do they have left as profits?
How innovative are they, anyway?

What you will see is a Wizard of Oz industry—full of bluster, and certainly vastly rich and powerful, but with very little to justify its pretensions.

Contrary to the industry’s public relations statements, big drug companies spend relatively little on R&D—far less than they spend on what they call “marketing and administration” and less even than they have left over in profits. Consider a few figures for 2002. That year, the top 10 American drug companies had sales of $217 billion. By their own figures, they spent 14 percent of that on R&D. But they spent over twice as much, a whopping 31 percent of sales (or about $67 billion), on marketing and administration. And they had 17 percent left over as profits. That is a stunning profit margin, as we shall see.

For some reason, nearly all companies lump marketing and administration together in their annual reports, but one company, Novartis, separates the categories. It reported that 85 percent of the combination went to marketing. Assuming that the proportion is about the same for the other big companies—and there is reason to think it is—then they spent nearly twice as much on marketing alone as on R&D.

In its public pronouncements, the industry denies this by counting just four specific items as marketing: direct-to-consumer advertising; visits to doctors by sales representatives; the value of free samples; and advertising in medical journals. But marketing budgets cover a lot more than those four things—most important, what drug companies term the “education” of doctors.

We are all aware of direct-to-consumer ads by drug companies. They spend over $4 billion annually on those, but that is small potatoes compared with their expenditures on physician “education.” Drug companies sponsor the majority of medical meetings, and the cost probably runs to tens of billions of dollars. What doctors are learning is a very drug-intensive style of medical practice, and they come to believe that expensive new drugs are better than cheaper older ones, despite the fact that there is usually no evidence to support that idea.

And what about profits? For well over two decades, drug companies have had consistently higher profits than any other industry in the United States after paying for R&D and all their other expenses. As I mentioned, the top 10 American drug companies had a profit margin of 17 percent of sales in 2002—well above their R&D expenditures. Compare this with the median profit margin of 3.1 percent for the Fortune 500 industries that year. In 2003, for the first time, the industry fell slightly from first place to third, but its profits were still well above the median. There is no way marketing drugs can be construed as a high-risk business, despite industry rhetoric to the contrary.

The recent claim that drug companies spend on average $802 million to bring each new drug to market is based on secret, proprietary data and is wildly inflated. But whatever the companies spend on R&D, if they spend more on marketing and have more left over as profits, they can hardly claim that high prices are necessary to cover their R&D. Instead, high prices are necessary to cover their stupendous marketing expenditures and to maintain their enormous profits. There is now some pushback on prices, but drug companies are compensating by trying to persuade more people to take more drugs, thereby increasing volume.

Me-too drugs

The important issue is not how much drug companies spend on R&D, but whether we get our money’s worth. I argue that we do not.

Remarkable as it seems, only a small fraction of drugs now coming to market are innovative in any meaningful sense of the word. Between 1998 and 2003, the FDA classified 78 percent of the 487 drugs that entered the market as likely to be no better than existing drugs. And 68 percent of these drugs were not new at all, but old drugs in new forms or combinations.

In other words, the major output of the industry is not important new drugs, but minor variations of drugs already on the market—called “me-too” drugs. For example, the top-selling drug in the world, Pfizer’s Lipitor, is a me-too drug—the fourth of six cholesterol-lowering drugs of the same type. There are whole families of me-too drugs, and there is little reason to think one is better than another at comparable doses.

The few innovative drugs that emerge usually stem from publicly funded research done at government or university labs. Even in me-too drug families, the original is usually based on government-sponsored work. For example, the first of the Lipitor-type drugs, Mevacor, came on the market in 1987 and was based largely on university research. Most of today’s top-selling drugs have progenitors dating back to the 1980s or even earlier.

Despite industry rhetoric, the drug companies are growing less and less innovative. They re-jigger the same old drugs just enough to get new patents, and rely on their marketing muscle to convince doctors and patients that they are producing medical miracles.

Of the 78 drugs that entered the market in 2002, only seven were new chemical compounds classified as likely to be better than old drugs. And not one of the seven was made by a top-10 American drug company. In fact, the big drug companies openly admit that they rely on licensing drugs from small biotech companies around the world. They are outsourcing their research. That may be legitimate, but it hardly supports their claim to be innovative and to be rewarded as though they were.

Most industry R&D spending goes for clinical trials—the end of the R&D process. While this is the least creative part of R&D, it is the most expensive. Before the FDA permits a new drug to enter the market—or an old one to be sold for a new use—the company must demonstrate in clinical trials that the drug is both safe and effective. But it only has to compare the drug with a placebo, not with a drug already in use. That is why we usually have no idea whether a new drug is any better than an old one. It just has to be better than nothing—a very low standard indeed. This inexplicable loophole makes it possible for drug companies to turn out one me-too drug after another, instead of undertaking the harder task of trying to discover innovative ones.

Me-too drugs are used in already-established, highly profitable markets that are easily expanded. In fact, drug companies often promote diseases to fit drugs instead of the reverse. They use direct-to-consumer ads to persuade essentially healthy people that they have medical conditions that need ongoing treatment. Why? Because there are more healthy people than sick ones, so the market is bigger and more easily expanded. Thus, millions of Americans come to believe they have dubious or exaggerated ailments like “generalized anxiety disorder,” “erectile dysfunction,” or “acid reflux disease.” For every ailment or discontent, there seems to be a drug—or many drugs.

This is beautifully spoofed in a cartoon in the April 11, 2005, New Yorker. It shows a little boy asking the pharmacist, “How much are the home-run pills?”

Many me-too drugs target not diseases but predisposing conditions, like high blood pressure or high cholesterol. I do not want to throw out the baby with the bathwater here. The appropriate pharmaceutical treatment of high blood pressure or high cholesterol is important and even life-saving for many people. But in recent years, the definitions of these conditions have been considerably broadened, so that in my view, many people are now taking drugs when it is not at all clear they are of any benefit or more beneficial than losing weight or exercising more.

Conflicts of interest

We need to be concerned not only about the shift from innovation to imitation, but also about the reliability of the research on which the approval and use of drugs depends. I have been reluctant to believe that clinical research on prescription drugs is generally biased. But in the past couple of years I have had to conclude that it is biased far more often than I realized as an editor of the New England Journal of Medicine. I want to spell this out in some detail.

When a company seeks approval from the FDA, it is required to submit all of the clinical trials done for that purpose, but it need not submit studies that were not done to support an FDA application. Furthermore, the FDA will not release all the trial results in its possession without the consent of the company. Nor does the company have to publish or otherwise publicize any of the results.

That means many clinical trials never see the light of day. Companies, of course, are eager to publicize favorable trials, but unfavorable results remain hidden—often within the FDA, which in this regard seems to put protection of industry “proprietary” interests ahead of the public health.

That is what New York Attorney General Eliot Spitzer found in the case of GlaxoSmithKline’s antidepressant Paxil. Only the favorable trial results were published, even though the FDA knew of studies showing that Paxil was associated with an increased risk of suicidal thoughts in children.

Another example: A few years ago, two researchers, Irving Kirsch and Thomas Moore, used the Freedom of Information Act to obtain FDA reviews of every placebo-controlled clinical trial submitted for initial approval of the six most widely used antidepressant drugs approved between 1987 and 1999—Prozac, Paxil, Zoloft, Celexa, Serzone, and Effexor. What they found was startling. All six drugs were only minimally effective—on average the difference between drug and placebo was 2 points on the 62-point Hamilton Depression Score, not enough to be clinically significant. All were equally ineffective. Yet these drugs are widely considered by both doctors and the public to be highly effective, because only favorable results are publicized.

So we need to worry a lot about the suppression of unfavorable research. But we also need to ask whether we can rely on the favorable research that is published. I spent much of my professional life evaluating clinical trials for publication in the New England Journal of Medicine, and I can tell you that there are many ways to design a trial to make a drug look better than it really is—in addition to the obvious one of comparing it with a placebo.

For example, companies sometimes enroll only patients at low risk of side effects, even though the drug is intended for use in more vulnerable people. That way, a drug will seem to have fewer side effects than it will when it comes into widespread use. Or a new drug may be compared with an old one administered at too low a dose. That makes the new one look more effective, and it can be promoted as being stronger.

Sometimes, unfavorable data is simply omitted from a publication. Pfizer, for example, launched a one-year study to determine whether the painkiller Celebrex was easier on the stomach than older painkillers, but it published only the first six months of the results, since the favorable effects disappeared in the second six months.

The fundamental problem is that drug companies have far too much control over research on their own products—how the research is designed, conducted, and published. They usually hire private research firms to oversee the trials, which in turn hire private doctors to enter their patients in the trials. But the firms and their doctors are all on the company payroll, and if they want the business, they follow company instructions. Even when trials are conducted in academic centers, the researchers are often on company payrolls as consultants or advisers. These pervasive conflicts of interest raise concerns that much clinical research on prescription drugs is unreliable and that, in general, both doctors and patients have come to believe that drugs are more effective and safer than they are.

In the wake of the Paxil case settlement, there were calls to register all clinical trials, whether favorable or not. That is a good idea, but it needs to be done right. Trials should be registered at inception in a central, publicly administered database. Initial registration would detail the design of the study—the kinds of patients to be enrolled, the drugs and doses, the outcomes to be measured, and how long the trial would last. That would prevent companies from changing the rules to suit the results, as Pfizer did when it published only the first six months of a one-year study and almost got away with it. At the end of the trial, the salient results would be added.

Registration should include all trials and be a requirement for enrolling human subjects. After all, using people for experimentation should entail public accountability. Clearly, half-hearted industry promises of voluntary registries are not enough.

But as valuable as a proper registry would be, a more important reform would be to deal with the underlying conflicts of interest. In my book, The Truth About the Drug Companies: How They Deceive Us and What to Do About It, I suggest ways to do this, including creating a separate institute within the National Institutes of Health to oversee clinical trials of prescription drugs before FDA approval. It makes no sense to rely on investor-owned businesses to evaluate their own products. And in the case of prescription drugs, it is way too dangerous.

Perhaps the single most important reform we could institute would be to require that drug companies compare their new drugs with old ones already on the market that treat the same condition. Drugs generally should not be approved by the FDA unless they are shown to be better in some way—more effective, with fewer side effects, or easier to take. Sometimes it may be deemed desirable to have more than one drug in a class on the market, but there is no excuse for having four, six, or eight very similar drugs.

The argument is often made that me-too drugs introduce price competition, but there is no evidence of that. Prices never drop in response to the introduction of a similar drug, and new drugs almost never are promoted on the basis of price. They are promoted as though they were better, even though there is usually no evidence on that score.

This reform would pull the rug out from under the me-too market and force the companies to do what they claim they are already doing—working on truly innovative drugs. It would also eliminate the need for those gigantic marketing expenditures. You can see from direct-to-consumer ads that the drugs most heavily promoted are me-too drugs. The companies have to persuade both doctors and the public that there is something special about their particular me-too drug—despite the fact that there is seldom any scientific evidence to that effect. If they had a truly important and unique drug, it would not need to be advertised. Any company that comes up with a cure for cancer, for example, will not have to promote it.

The industry portrays itself as a model of American free enterprise, but it is anything but that. Of the top 10 drug companies in 2002, half are European. While the industry is free to decide what drugs to develop and what to charge for them, it depends on government-funded research and government-granted monopolies in the form of patents and exclusive marketing rights to protect sales income.

It is time to demand something in return. Every other advanced country regulates prescription drug prices in some way. If the drug companies continue their price-gouging, we should consider importing from Canada—not drugs, but a similar system for regulating prices. We need not worry about stifling innovative R&D. Drug companies do much less of it than they claim, and what they do they can easily afford.

Marcia Angell, a physician and former editor of the New England Journal of Medicine, is senior lecturer in the Harvard Medical School Department of Social Medicine. The text of this lecture, which she delivered last year at the Arizona State University College of Law, was published in the Summer 2005 issue of Jurimetrics. Reprinted with permission of the American Bar Association.