The Golden Age of Medical Innovation
From the March/April 2007 Issue
Critics grouse about a sluggish drug pipeline, but they are looking in the wrong place. John Calfee shows that innovation in drugs and medical devices in the U.S. is rampant. New treatments are fighting breast cancer, macular degeneration, rheumatoid arthritis, and many more diseases. Why the success?
Last December, the editors of Science magazine—the widely respected journal of the American Association for the Advancement of Science—selected ten “breakthroughs” of the year.
Number one was the solution to the “Poincaré Conjecture,” a topological puzzle that had confounded mathematicians for a century. Number four was the discovery of a 375-million-year-old fossil that filled an evolutionary gap between fish and birds. But it was breakthrough number six that got my attention—“a ray of hope for macular degeneration patients.” It related the stunning clinical-trial results for ranibizumab, a drug sold under the brand name Lucentis by Genentech, its developer and manufacturer.
Lucentis is a monoclonal antibody—made, using biotech methods, from genetically engineered bacteria—that attacks a protein responsible for the leading cause of blindness in seniors. Age-related macular degeneration, or AMD, affects a part of the retina and causes blind spots, which expand over time, in the center of the field of vision. It is a disease that had previously defied all attempts even at delaying the inevitable decline toward blindness. But with Lucentis, the eyesight of about 95 percent of AMD patients either improved or stopped getting worse—a stunning achievement.
With the drug Lucentis, the eyesight of about 95 percent of patients with macular degeneration either improved or stopped getting worse—a stunning achievement.
Biotechnology has lately provided many other breakthroughs. In January, an article in The Lancet, a British medical journal, described the latest striking results for another Genentech drug, Herceptin, which treats an aggressive form of breast cancer that affects about 50,000 women a year. The cancer is characterized by an overabundance of a protein called human epithelial growth factor receptor-2 (HER2), which stimulates the growth of tumors. In 2005, research published in The New England Journal of Medicine found that, when administered after surgery, Herceptin reduced the odds of a recurrence by half. An accompanying editorial concluded that the key result “suggests a dramatic and perhaps permanent perturbation of the natural history of the disease, maybe even a cure,” and that “our care of patients with HER2-positive breast cancer must change today.” The new Lancet study went even further, documenting mortality declines after only two years of use, an unheard-of result in treating breast cancer. Other biotech treatments approved recently include Enbrel, for psoriasis, a painful skin condition; Remicade, for Crohn’s disease, an inflammatory digestive-tract disorder; Rituxan, for rheumatoid arthritis; and Avastin, for several different cancers.
Enbrel was developed by Immunex, a Seattle company that was purchased in 2001 by Amgen, the world’s largest biotech firm. Remicade is a product of Centocor, the biomedical firm that pioneered monoclonal antibody technology. And Rituxan and Avastin are from Genentech, one of the early biotech firms that is now seeing a boom in revenues, up 40 percent in 2006 to $9 billion.
The remarkable treatments these firms have developed seem at odds with the widespread notion that medical innovation has all but come to a halt. A recent report by the Government Accountability Office, for example, has warned, in the words of a popular website that covers drug research news, “that the pharmaceutical industry is not producing enough new drugs despite spending more on research and development.” In a similar vein, Democratic Representative Henry Waxman, the new chairman of the House Committee on Government Reform, stated, “Many aspects of the drug development system need to be examined to determine how to encourage research that focuses on breakthrough treatments rather than drug industry profits.”
Yet breakthroughs are precisely what are being achieved. Why aren’t they more widely recognized? One reason is that none of the treatments I have just described would appear on a list of innovative new drug approvals over the past few years. The results all came from clinical trials involving drugs that had already been approved for something else—often a different stage of cancer, sometimes a different type of cancer, and occasionally a different illness altogether.
Rituxan, for example, which was found to treat rheumatoid arthritis, was originally approved for cancer, and Enbrel (psoriasis) and Remicade (Crohn’s disease) had originally been approved for rheumatoid arthritis. Lucentis, the AMD fighter that made Science magazine’s list of breakthroughs, was arguably what pharmaceutical industry critics derisively call a “me-too” drug. It was created by tweaking the molecular structure of a drug that became Avastin.* Avastin and Lucentis were then developed in parallel, Avastin was on the Science magazine top-10 breakthrough list in 2003. It was approved in 2004 for colorectal cancer has since been approved for certain kinds of lung cancer, and has been submitted to the Food and Drug Administration to be used against breast cancer and possibly kidney cancer as well.
These developments illustrate why, at the very time we keep hearing that progress has stalled, medicine today is actually in a new golden age of innovation. In this new era, the most important advances in treatment often come from products which have been on the market for a while but whose properties were not completely understood until intensive research after the drug was introduced. In cancer treatment, Avastin is a mini-pipeline all by itself, with some 20 clinical trials underway for different cancers or stages of cancer.
In cancer treatment, Avastin is a mini-pipeline all by itself, with some 20 clinical trials underway for different cancers or stages of cancer.
The dominant role of post-approval research extends to many other drugs used as what are called “targeted therapies,” which, as the National Cancer Institute defines the term, “block the growth and spread of cancer by interfering with specific molecules involved in carcinogenesis (the process by which normal cells become cancer cells) and tumor growth.” Some of these targeted therapies find value in treating a completely different illness. More successes are, no doubt, on the way. For example, unexpected results in treating patients with Gleevec, developed by Novartis and one of the first and most spectacular of the targeted cancer drugs, sparked curiosity about its ability to prevent diabetes, while researchers are looking at tantalizing evidence that the HIV drug Viread, made by Gilead, may act against hepatitis B.
A second characteristic of this new era is that competition from follow-on or me-too drugs has been raised to extraordinary levels of scientific sophistication. After Avastin finally demonstrated the therapeutic benefits of inhibiting the angiogenesis (growth of new blood vessels) that feeds rapidly growing cancer cells, competition emerged with amazing speed. Several Avastin-like drugs, which, like Lucentis, may treat new diseases, are in late-stage clinical trials. The same thing has happened to other breakthrough biotech drugs for cancer. Herceptin, for example, faces competition from Iressa and Tarceva, currently on the market, while lapatinib and pertuzumab are in late-stage clinical trials. Tykerb, which is also effective against tumors that have evolved to resist Herceptin, is close to FDA approval. Competition has also emerged for targeted drugs to treat rheumatoid arthritis, osteoporosis, and other conditions. Lucentis may soon face competition from another biotech cancer drug in clinical trials for macular degeneration.
The story of spectacular new uses for old drugs is not a new one. It happened to blood pressure drugs, ulcer preventatives, and the statin class of cholesterol-reducing drugs like Lipitor, which we now know also prevents strokes. But the development of new uses and the creation of new competitors happen faster with today’s drugs that are more precisely targeted at very specific biological mechanisms in the human body. This may seem paradoxical, but the targeted mechanisms tend to play many roles in the body and are typically part of a cascade of events that can be interrupted in many different ways—and, thus, by many different cleverly designed drugs.
It is research on entirely new drug mechanisms—that is, drugs that operate in the body in ways that are quite different from the way drugs have operated in the past—which usually marks the origins of therapeutic revolutions. Progress is punctuated with both successes and failures. And often the crucial result tells researchers what will not work. Today, several pharmaceutical firms are trying to figure out the cause of Pfizer’s spectacular failure last year with torcetrapib, a biotech drug designed to increase HDL, or high-density lipoprotein, the “good cholesterol” that prevents heart attacks. HDL research continues, with huge benefits for patients if someone eventually achieves success.
Initial successes, such as Mevacor, the first statin cholesterol-reducing drug, typically provide only a glimmer of what a new mechanism can achieve after it is refined to far greater efficiency through competitive research to find those notorious me-too drugs.
In the search for entirely new mechanisms, however, uncertainty abounds. For decades, research on the mechanism that would power Avastin was the lonely obsession of biologist and pediatric surgeon Judah Folkman at Children’s Hospital in Boston. He never quite achieved complete failure, and when success finally came, it opened the door to a new approach to cancer research: controlling tumors by targeting the blood vessels that help them grow.
Far from the public eye, progress is finally being made on autologous vaccines for cancer and Alzheimer’s disease. Such vaccines are created from a sick patient’s own cells.
The past two decades have also seen concerted efforts to devise entirely new ways to treat HIV infection and depression—conditions where therapy has long been dominated by a multitude of drugs employing closely related mechanisms. After years of failure, Merck is finally close to FDA approval of the first integrase inhibitor, employing a completely new method of fighting HIV. Recent years have also seen pharmaceutical-based revolutions in rheumatoid arthritis, osteoporosis, and of course, heart disease, where the death rate has dropped by 29 percent in a decade.
Also, far from the public eye, progress is finally being made on autologous vaccines for cancer and Alzheimer’s disease. (Such vaccines are created from a sick patient’s own cells.) These new treatments are part of a larger search for therapeutic vaccines—that is, drugs that do not merely boost the immune system but harness it to attack specific causes of illness. In the meantime, traditional vaccine R&D is in a renaissance, with brand-new preventatives for cervical cancer and rotavirus (a form of diarrhea that can and often does kill young children).
The usual methods for measuring progress in drug research simply do not work any longer. Industry critics—such as Marcia Angell, former editor of The New England Journal of Medicine and author of The Truth About the Drug Companies, a 2004 book that was praised in The New York Times—focus relentlessly on tallies of new drug approvals while excoriating me-too drugs. In its November 2006 report, entitled “New Drug Development: Science, Business, Regulatory, and Intellectual Property Issues Cited as Hampering Drug Development Efforts,” the GAO also drew its conclusions based on new drug applications, or NDAs, to the FDA, while giving scant attention to follow-on research. The GAO report pointed out that NDAs had risen annually from 74 to 129 between 1993 and 1999, then started sliding, ending at 102 in 2004, the final year of the study.
There are better ways to mark progress. Post-approval randomized clinical trials, to which the marketplace (physicians, third-party payers, and, above all, patients) responds with gusto, provide extremely valuable information on where drug therapy is taking us. Virtually all the results I have described came from such trials, some of them designed for FDA supplemental approval and some just for the medical community at large.
We don’t hear much about these trials in the rancorous debate over what the pharmaceutical industry has done for us lately. Instead, we get a scorecard of NDAs. This is unsurprising, because NDA counts are easy numbers to grab, but the true nature of healthcare innovation is the slow, but immensely productive, adaptation by doctors, hospitals, clinics, patients, and society at large to new therapeutic possibilities.
The treatment of heart attacks and strokes, for example, requires broad changes in behavior by emergency personnel and hospitals as well as better knowledge in the general population; otherwise, the best antidotes will be used too late. This arduous adaption process, which only begins with new drugs (or new breakthrough uses for old ones, like aspirin), has been documented by Harvard’s David Cutler, Mary Beth Landrum, and Kate Stewart in a recent National Bureau of Economic Research paper with an abstract that begins, “There is little empirical evidence to explain why disability declined among the elderly over the past 20 years.”
We don’t see many headlines about the inadequacy of medical-device R&D—because it’s harder to focus on a few misleading numbers as sole indicators of innovation.
A few scholars are measuring the health impact of the broad sweep of drug development. Economist Cutler and his two colleagues found that new medical technology, mainly drugs, accounted for most of the spectacular 55 percent drop in heart disease deaths between 1975 and 1995. A group of public-health scholars determined that HIV drugs have extended the lives of patients by an amazing 18 years. In a series of econometric studies using a wide variety of data sets, economist Frank Lichtenberg has found a strong and consistent connection between the adoption of newer drugs and various health and economic benefits, including longevity. For example, new drugs extended the lives of cancer patients at a cost of less than $3,000 per life-year—even before the arrival of the latest targeted therapies like Herceptin and Avastin.
Like pharmaceuticals, most medical devices must receive FDA approval before they are sold to the public. The variety of devices is astonishing, ranging from bedpans, bandages, and syringes to cochlear implants (which help the profoundly deaf to hear), implantable heart defibrillators (which provide a jolt if the heartbeat becomes irregular), and drug-eluting stents (which keep formerly clogged arteries open). While the FDA approves a few hundred new drugs and variants (like special forms of dosing) annually, the agency has recently been approving new medical devices at the rate of 3,000 per year, and permitting the introduction of another 5,000 without formal approval. The technically sophisticated part of the device market is so new that it was not even subject to direct regulation by the FDA until 1976.
One reason we don’t hear much about FDA statistics on new device approvals is that Congress constructed a bizarre regulatory regime that classifies some of the most innovative new products as variants of old ones—the so-called 510(k) class. But even for devices that, like drugs, go through the pre-marketing approval process, what really counts, again, is what we learn after approval. Most cardiac stents, for example, are used in ways not strictly in accordance with the FDA label, while new applications of MRI (magnetic resonance imaging) machines are a staple of the medical literature.
We don’t see many headlines about the inadequacy of medical-device research and development—not because R&D works better in the device market than in the pharmaceutical market, but because it’s harder to focus on a few misleading numbers as sole indicators of innovation. What we do see, if we look closely enough, is steady and accelerating progress.
In contrast to pharmaceuticals, where progress often arrives in the form of new information about old drugs, progress in the device industry typically consists of improvements in the devices themselves and, often, in the training and skills necessary for physicians to get greater benefits from them. With cardiac stents and minimally invasive surgery, for example, the skill of the practitioner can determine whether a new model is better than the old one. Progress also comes from sharper displays, smaller manufacturing tolerances, more precisely controlled robotics, new packaging for better sterilization, and on and on.
This environment of incremental improvements, speedy and pervasive competition, and better physician training results in changes that cannot be summarized by simple benchmarks. Even clinical trial data are scarce because it is hard to conduct random testing for devices whose utility depends on the skill of users. Also, incremental advances make it hard to finish a trial and analyze its results before the device being tested (a miniature TV transmitter for gastrointestinal endoscopy, for example) becomes obsolete. So even profoundly useful innovations in medical devices often fail to generate the sorts of data that make it to the newspaper pages.
Important progress has been made on diagnostics for HIV, other sexually transmitted diseases such as chlamydia and gonorrhea, congestive heart failure, hepatitis B and C, and hundreds of other targets.
In the meantime, innovation is moving ahead on another critical healthcare front: diagnostics—that is, techniques for judging whether someone has a disease and how virulent it is, how the disease might react to a drug, and the likelihood of severe side effects. Important progress has been made on diagnostics for HIV, other sexually transmitted diseases such as chlamydia and gonorrhea, congestive heart failure, hepatitis B and C, and hundreds of other targets. The DNA decoding revolution has strongly affected diagnostics, generating tests for genetic mutations (including those involved in breast and lung cancer) and the remarkable creation of genomic arrays, which are sheets typically made of glass with snippets of DNA that can quickly reveal links between, say, a drug and cancer cells.
There’s a symbiotic relationship between diagnostics and treatment. For example, a series of trials has demonstrated the striking benefits of adjusting blood glucose levels in diabetics daily or even several times a day. Manufacturers have introduced more accurate, more convenient, and less painful glucose measurement devices that are now critical for better treatment. For some drugs, a precise diagnostic tool can reduce costs and improve safety. In fact, some of the targeted cancer drugs, including Herceptin and Iressa, would be nearly useless without gene-based diagnostics that indicate whether the medicines will work at all.
It’s no surprise that advances in applied science—biology, materials (ceramics, glass, metals, plastics), chemistry, computing, and the like—would profoundly affect R&D in medical technology. But what’s not appreciated is the extent to which different parts of the medical technology industry have themselves converged to form a nearly seamless entity.
Consider breast cancer treatment, which today involves targeted drugs (along with older ones, still very useful), several molecular diagnostics that determine what drugs to use, improved imaging tools (with digital spot view mammography rapidly replacing analog), and a variety of devices such as less-invasive stereotactic core needle biopsies, minimally invasive lymph node surgery, better brachytherapy (radiation therapy using tiny “seeds”), and intensity-modulated radiation therapy. A recent academic review of two decades of drug treatment concluded, “Though much remains to be discovered, and some hard battles no doubt remain to be fought, the end of breast cancer as a serious cause of human mortality is now in sight.” This convergence process is also happening with the rapidly advancing treatments for colorectal and lung cancer and for diabetes, which threatens to become the single most expensive illness in the United States.
Rapid technological change usually threatens entrenched interests, including competitors facing eclipse and trade unions wanting to protect their membership. In medical technology markets, threats to continuing innovation seem to arise from different sources. Here are the kinds of threats that can end a golden age:
Intellectual Property: The surest way to hobble medical technology is to damage intellectual property (IP) protections, mainly patents. Drug companies simply won’t spend the nearly $1 billion needed to develop the average new medicine if they have uncertain ownership when they’re through. While the lack of a rigorous IP regime in developing nations has deterred the development of new drugs and vaccines for tuberculosis and malaria, the good news is that the World Trade Organization’s TRIPS agreement is slowly buttressing R&D incentives to attack diseases in these poor countries.
Price Controls: Innovation is senseless if successful products can’t be priced well above the marginal costs of manufacturing and distribution. But consumers, encouraged by politicians, resent paying those prices—especially when the same products are sold for less in other wealthy nations. As a result, the power of price controls in many European countries and in Canada raises the specter of similar controls in the United States. Other than the dismantling of intellectual property, no policy would be more destructive to innovation than price controls.
Other than the dismantling of intellectual property, no policy would be more destructive to medical innovation than price controls.
Unfortunately, public discussion of drug pricing is informed by fundamental misunderstandings of the economics of medical-technology R&D on the part of journalists and even prominent academic writers. Especially dangerous is the view that, since recent years have not brought much innovation, little of value would be sacrificed even if price controls impose disincentives for research in the private sector. The illusion that publicly funded R&D can replace private markets contributes to this thinking. In fact, financial risk-taking in pursuit of exceptional profits has been the only reliable source of pharmaceutical innovation that proceeds all the way from laboratory to bedside.
The most immediate threat to pricing freedom is entirely novel. The National Institutes of Health, encouraged by a New England Journal of Medicine editorial, has embarked on a clinical trial to compare Lucentis to off-label Avastin in treating macular degeneration. The explicit purpose is to demonstrate that AMD can be treated at a fraction of the cost of Lucentis by using tiny amounts of Avastin, whose primary use in cancer therapy involves far larger doses. The goal is not to control prices directly but to undermine a pricing structure that is essential to R&D. If firms cannot charge a higher price per milligram for a new use at a much smaller dose, there is no incentive to research it. The NIH seeks to prevent Genentech from reaping a payoff from developing the very therapy that was on Science magazine’s list of the ten breakthroughs of the year. It is hard to think of a worse signal to send to the developers of targeted drugs.
FDA Regulation: The Food and Drug Administration has come under vigorous attack from Congress, the news media, and, most important, medical academics. Although the FDA acquitted itself well in the most important episode, the withdrawal of the pain reliever Vioxx, overzealous attacks on its handling of drug safety are bound to cause the agency to tighten its standards for new drug approvals. Look, for example, at the agency’s refusal to describe its new standards for the testing of antibiotics, including such basic matters as when trials must include a placebo control.
A recent Institute of Medicine report strongly criticized FDA handling of drug safety and proposed many changes, including marketing restrictions and stronger agency powers, even to the point of making controls on physician-prescribing a precondition for some new drug approvals. The vehicle for legislating changes will no doubt be the Prescription Drug User Fee Act (PDUFA), which is up for renewal this year. User fees collected under PDUFA cover the costs of faster new-drug reviews.
Marketing Restrictions: On the surface, marketing may appear to be wasteful. You often hear, even from intelligent critics, the argument that advertising and other marketing costs detract from R&D expenditures. (The truth is that marketing, by boosting sales, increases returns to R&D and provides the cash needed for research.) Certainly, marketing can encourage people to use drugs inefficiently—but at the root of that problem is the fact that third parties, mainly insurance companies, are footing most of the bill, so patients have little incentive to monitor costs.
‘Though much remains to be discovered, and some hard battles no doubt remain to be fought, the end of breast cancer as a serious cause of human mortality is now in sight.’
In fact, for a wide variety of drugs, devices, and diagnostics, the real problem is underuse; people are suffering and dying because they do not know about the technologies that can help them. Certainly, promotion can also deceive patients and physicians, but a rich literature demonstrates Americans’ innate skepticism toward advertising. People are smarter than many politicians think. Still, we can probably expect continued attacks on pharmaceutical marketing at both the state and federal level.
Litigation: Abusive lawsuits can destroy medical innovation. They nearly did so with childhood vaccines before those preventatives were moved out of the tort liability system in 1986. Most litigation against drug companies runs in cycles. The 1990s were relatively quiet, other than the exceptional case of the Fen-Phen (fenfluramine-phentermine) diet drug combination, which never involved active marketing by the manufacturer before it was removed for safety reasons. Now, we are in a powerful upturn in the cycle, fueled by demonization of the industry by the media, resentment at prices seen as too high, and the drug-safety furor triggered mainly by Merck’s Vioxx, a popular painkiller that was removed from the market because of cardiovascular risk. Much of Vioxx’s appeal to doctors and patients was that it could substitute for medicines that have serious side effects—including death—in treating such diseases as osteoarthritis.
The current wave of litigation against Merck has yet to reveal its full contours, with final costs likely to range from a billion dollars to tens of billions. But at the core of the matter lies the unsettling fact that juries are being required to answer a scientific question—whether Vioxx actually caused the heart attack or stroke being litigated—that can’t accurately be decided. So far, no diagnostic tool exists to make the judgment.
Other drugs in litigation, such as hormone replacement therapy and certain forms of birth control, present similar issues. Will a litigation explosion pose serious threats to further innovation in medical technology? There’s no clear answer, but the danger is real.
For now, however, far from being in a slowdown, innovation in pharmaceuticals, medical devices, and diagnostics is accelerating. The forces that led to the creation of the breakthroughs—advances in computer software and hardware, imaging, genomics, exotic materials, and much more—are far from mature stages of development.
We are in a golden age of medical innovation—but not in the sense that this is a shining era upon which historians will look with fondness (like the great age of classical music) distant decades from now. In free markets, every glittering age of technology is succeeded by another that is even more glittering. Only unwise public policy can make the fire go out.
John Calfee is a resident scholar at the American Enterprise Institute. * Correction: The published version of this article incorrectly described the drug Lucentis as being derived from an older drug, Avastin. The corrected version notes that Lucentis and Avastin were developed in parallel, with Lucentis being a modified form of the monoclonal antibody that became Avastin.
John Calfee is a resident scholar at the American Enterprise Institute.
* Correction: The published version of this article incorrectly described the drug Lucentis as being derived from an older drug, Avastin. The corrected version notes that Lucentis and Avastin were developed in parallel, with Lucentis being a modified form of the monoclonal antibody that became Avastin.
 Smith I, Procter M, Gelber RD for the HERA study team. 2-year follow-up of trastuzumab after adjuvant chemotherapy in HER2-positive breast cancer: a randomised controlled trial. Lancet 2007; 369: 29-36. Abstract available here.
 M. Flanagan, “Avastin’s Progression,” BioCentury 14, no. 11 (2006): A1–A5.
 Peters MG, Andersen J, and ACTG Protocol A5127 Team. Randomized controlled study of tenofovir and adefovir in chronic hepatitis B virus and HIV infection: ACTG A5127. Hepatology. 2006 Nov;44(5):1110-6. Abstract available here.