The pharmaceutical industry is an easy target-too easy sometimes. Take a television camera crew to any African hospital and you will find people dying from diseases-Aids, chest infections, malaria, tuberculosis, meningitis-that are entirely treatable. Thousands of deaths stack up in the world's poorest countries because people are denied access to essential medicines.
NGOs, such as M?decins Sans Fronti?res and Oxfam, have made a persuasive case to governments and the public that big pharma's paramount concern is to protect its intellectual property at the expense of the lives of the poor. Pharmaceutical companies reply that they cannot be a national health service to the world. As businesses, they owe legal responsibilities to their shareholders. The job of a drug company is to design, market, and sell new medicines, while protecting its intellectual property as far as it reasonably can. Western food companies, they point out, are not vilified when starvation strikes Ethiopia.
If big pharma does sometimes seem to be attributed with more than its share of responsibility for the world's medical troubles, there are other issues where it escapes too lightly. Medicines, for example, kill as well as cure. In fact, the pharmaceutical industry is one of the biggest killers in the industrialised world. And it is officially licensed to be so. This seems an extraordinary claim. What is the evidence for it?
For any new drug to receive an official stamp of approval for use (in Britain, it comes from the committee on safety of medicines), it has to be shown to be safe and effective in at least two clinical trials. This final stage in the drug registration process is preceded by often lengthy investigations in the laboratory and at the bedside. The trials leading to a drug's approval involve comparisons of the new medicine with a placebo. Patients are randomly allocated to one or other "treatment." They are then followed up for a predefined period, sometimes several years.
Naturally, for the drug to be proven effective it must outperform the placebo across a range of clinically relevant end points, such as death, disability, or symptom control. These rules for testing a new medicine are evolving. More and more clinical trials sponsored by pharmaceutical companies do not set out to prove that their new drug is better than a placebo. In what are called "equivalence trials," these companies will simply compare their new product with the standard treatment. If the two treatments-new and old-have substantially the same outcomes they are said to be equivalent, and that may be enough for a licence to be granted. The justification is that the new drug may have other advantages which are not easily measured in terms of the usual clinical outcomes. For instance, while its effectiveness might be identical, the new medicine might only need to be taken once a day, instead of three times.
These early clinical trials are also supposed to give important information about safety. If the drug is dangerous, one would hope to see this fact emerge in the trial. But to be sure that one has reliable information about safety, scientists must do large studies over long periods of time to avoid the possibility of what statisticians call a "false-negative" result: one that wrongly suggests that the drug is safe. Although clinical trials are good at showing whether a drug does what it is supposed to do-alleviate symptoms of asthma, cure a person of cancer-they frequently fail to pick up the adverse effects of the new medicine.
One celebrated case is a drug called Tambocor, designed to treat abnormal electrical rhythms in the heart, which was launched in 1985. Tests in animals and early studies in patients seemed to show that this drug was successful at suppressing irregular heartbeats linked to premature death. Only after Tambocor was licensed did reports emerge suggesting that the drug carried unsuspected risks. Eventually, a clinical trial large and long enough to settle the uncertainty was designed. The trial proved that the drug killed many of those people taking it. Some estimate that up to 70,000 people, mainly in the US, have died.
A disturbing question remains unanswered. Let us assume that the case of Tambocor was a rare example of a fatal undetected adverse event. How many less extreme examples are there? And what is the sum total of their effect on our health? Three scientists in Toronto asked this same question several years ago. In a neglected but hugely important research paper published in the Journal of the American Medical Association, Jason Lazarou, Bruce Pomeranz and Paul Corey concluded that the frequency of serious and fatal reactions to commonly prescribed drugs was very high. They searched all the published evidence about adverse effects of drugs and from this information they calculated an overall estimate of risk. They excluded mistakes in the way drugs were administered, failures to take the drugs, overdoses, instances of drug abuse, and examples where the drug had clearly failed to do what it was supposed to.
Their calculations produced shockingly high numbers. They worked out that in 1994, well over 2m hospitalised Americans had an adverse reaction to their medications. An average of over 100,000 people died from these reactions. A league table of annual causes of death in the US therefore runs like this: heart disease (743,460), cancer (529,904), stroke (150,108), adverse reactions to drugs (106,000), lung disease (101,077), accidents (90,523), pneumonia (75,719), and diabetes (53,894). The best-case scenario is that only 76,000 people die from licensed pharmaceuticals annually. The worst possible outcome is 137,000 deaths. Thus pharmaceutical medicines are between the fourth and sixth most common cause of death in the US-and have been so over the past 30 years.
Why are these facts not more widely known? The pharmaceutical industry is a skilled manager (some would say manipulator) of information. The pharma business depends on carefully shaping the opinions of doctors, and increasingly patients, to prescribe or take large quantities of their products in a ruthlessly competitive market. What gives one drug a market edge is the way in which often marginal evidence of advantage is woven into a compelling story of effectiveness. The history of a drug called Lotronex shows what imagination can do to propel a new product into the limelight-and how dangerous the system is that allows this to take place with too little restraint.
Lotronex was launched by GlaxoWellcome for the treatment of women with irritable bowel syndrome, a non-life-threatening condition that affects as many as one in five adults. An irritable bowel causes pain, bloating and diarrhoea. Clinical trials seemed to show that the drug worked quite well at reducing the frequency of these unpleasant but manageable symptoms. But there were worrying drawbacks. About one in ten patients in the trial stopped taking Lotronex because of constipation. One patient developed a more serious condition called ischaemic colitis, in which the intestine is damaged by a shutdown in its blood supply.
GlaxoWellcome first sought approval of Lotronex in the US, the world's largest drug market, through the food and drug administration (FDA) and it was registered for use in February 2000. By November of that year, at least five people taking the drug had died. GlaxoWellcome withdrew Lotronex from the market.
During the FDA's review procedures, its office of drug evaluation concluded that Lotronex was safe and effective. But it said that the company had to conduct a further investigation into the risk of ischaemic colitis. The FDA was anxious about this risk, but not so much that it wanted to hold up approval of the drug. The FDA crossed its regulatory fingers and gave Glaxo Wellcome the benefit of its doubt.
Once Lotronex was given a licence, GlaxoWellcome's marketing division went into overdrive. Six-page advertisements appeared in specialist medical journals read by doctors who would treat patients with irritable bowel syndrome. These advertisements claimed that Lotronex had "a favourable safety profile." They did mention the problem of ischaemic colitis, although the company downplayed the risk by arguing that cause and effect was unproven.
But after only a few months on the market, concerns about the safety of Lotronex began to resurface. Seven patients had developed serious complications with constipation-three of them requiring surgery. Eight further cases of ischaemic colitis were reported. The drug carried a higher risk than even the pre-approval doubts had intimated. This course of events is common for drugs that turn out badly-a few inconclusive but nevertheless troubling problems before approval, followed by a widening cause for concern after the licence has been granted. The lesson for regulators is that it is best to act sooner rather than later to avert a public health disaster.
Instead, the FDA decided simply to warn patients about the risks of the drug. And so, most bizarrely, it was left to the company, not the FDA, to withdraw its own drug after five deaths proved that warning patients did no good at all. By the time Lotronex was taken off the market, there had been 49 cases of ischaemic colitis and 21 episodes of severe constipation, including horrific instances of ruptured bowels.
What was so shocking was the FDA's coolness about the Lotronex debacle. Janet Woodcock, director of the agency's centre for drug evaluation, declared only that the "FDA is committed to working with pharmaceutical sponsors to facilitate the development and availability of treatment options for patients." There were no words of regret.
In a final twist, after publicity surrounding the FDA's failure to protect patients from Lotronex had died down, GlaxoSmithKline (as it is now called) successfully lobbied the agency to bring its drug back on the market-but with heavy restrictions on its use.
Concerns about the deep involvement of the pharmaceutical industry in the management of drug information were revealed in a 1998 survey of the FDA's medical officers. Many of these physicians reported that since the 1992 Prescription Drug User Fee Act was passed, standards for drug approval had fallen sharply. The 1992 act enabled companies to give money to the FDA to hire medical officers to review drug applications from those same companies-supposedly creating a stronger FDA without requiring more taxpayers' money, and benefiting the industry by shortening approval times. But many of the FDA's own staff complain that the culture of the agency is now too pro-industry, too secretive and more prone to make mistakes.
A system of regulation that depends financially on the very entities it is supposed to regulate is looking for trouble. And, indeed, ten drugs approved by the FDA have been withdrawn since that 1992 law authorising large industry payments to the agency.
Matters at the FDA do not seem to be improving. Inspectors from the US department of health and human services reported this year that over half of the agency's reviewers believed that the time allowed for a priority review of the safety and efficacy of new drugs was inadequate. A quarter felt the time allowed for a standard review was insufficient. The department's report concluded that, "overall, these findings present a significant warning signal." They recommended that the review time could be extended by one or two months. But the FDA's new head, Mark McClellan, an economist by training, disagrees. He has declared that the FDA will speed up its review procedures.
The public is largely unaware of these flaws in the world's most important drug regulator. They are flaws that also afflict Britain's supposedly independent committee on safety of medicines (CSM), many of whose members have close ties-such as shareholdings and consultancies-with pharmaceutical companies. One recent example was the safety review commissioned on Seroxat, a drug for depression. A committee created by the CSM had to be disbanded when it emerged that several key members held shares in companies making this class of drug.
This hidden story about big pharma threatens an already vulnerable industry. The growth in pharmaceutical sales is, at least in the short term, in steep decline. The year 2001 saw an overall 12 per cent growth rate, with a 17 per cent growth in the North American market alone. But by the end of 2002, growth had slowed overall to 8 per cent, and 12 per cent in North America. After the booming 1990s there is a problem of inflated shareholder expectation. The giant corporate mergers were designed to keep up the earlier momentum but have failed to do so.
But there are reasons to believe that this is not just a short-term dip. Manufacturers of generics-drugs that no longer receive patent protection-are now competing for more top-selling products. Generics-based companies do not have to carry the research costs and development risks incurred by the innovation-driven sector. The NHS likes generics because they are cheap. Fifteen years ago, just a third of all prescriptions were generics. Today that proportion is over half. While this trend continues, the decline in sales growth for medicines is likely to be sustained.
And high-earning "blockbusters" are in short supply. The FDA approved only 17 new drugs in 2002, compared with 27 in 2000 and 53 in 1996. The 2002 figure was the lowest in 20 years. The same is true in Europe. In 1997 there were 60 new drug applications to European drug regulatory authorities. That figure fell to 19 in 2002. Thomas L?nngren, executive director of the European agency for the evaluation of medicinal products (EMEA), argued last year that, "Never has so much money been spent on research and development with so little result."
The sales figures, which have continued to decline in 2003, look even more dismal in the light of a range of patent expiries hitting companies. Nineteen drugs will lose patent protection in Britain this year. These medicines cost the NHS over ?210m each year. One drug, Merck Sharp and Dohme's anti-cholesterol medicine, Zocor, costs the NHS ?144m annually. It lost its patent in Britain on 5th May.
Moreover, sales of newly launched drugs, which should be the industry's most lucrative medicines, have collapsed. US sales of new drugs reached $1.2bn in 2001. They fell to just $156m last year. And this problem is also going to get worse thanks to the greater understanding about how drugs interact with the human genome. Roche recently launched a gene test to help doctors predict how a person will respond to a wide range of common medicines. If an individual's genetic makeup indicates that the drug will be ineffective, the doctor can avoid a wasted prescription-and possibly a nasty side-effect. Gene tests such as Roche's AmpliChip are projected to earn $100m by 2008. While this is a tidy income, it falls far short of the $1bn a year blockbusters can earn.
Given these escalating pressures, it is not surprising that companies are cutting corners. When Pharmacia-now owned by Pfizer-saw that a clinical trial of one of its products for high blood pressure was failing to show the success it had hoped for, the company pulled out of funding the remainder of the study. Doctors accused the company of breaking its promise to the thousands of sick individuals who had agreed to take part in the trial. But from Pharmacia's point of view a short-term squall of protest is better than a drawn-out visible failure.
The same was true for Abbott, which faced an investigation into the way it had sold products for feeding the seriously ill. An inquiry by US prosecutors in Illinois began in 2001. Abbott agreed in June to a $622m settlement fee, thus terminating the investigation before it reached a verdict. AstraZeneca has just agreed to end federal fraud charges by paying $355m in criminal penalties and civil fines.
We should not take pleasure in these negative trends for big pharma, for they are also unwelcome news for patients. Take prospects for cancer. Demographic changes mean that cancer will become little short of an epidemic in coming years. The World Health Organisation estimates that annual cancer deaths will more than double to 15m deaths by 2020. We need a successful pharmaceutical industry to help deliver good health in the face of these threats. Notwithstanding a few promising products big pharma is failing to measure up to the demands placed on it.
These failures in both safety and innovation are prompting regulatory bodies to adopt a higher public profile to balance the drug industry's dominance over pharmaceutical information. EMEA's goal is to talk directly with doctors and patients, inserting itself between industry and the public. The FDA's advisory committees are also more eager to press industry to conduct safety studies after a licence is granted.
What measures need to be taken to protect the public from the adverse effects of drugs, while at the same time creating the conditions to encourage industry to develop new medicines to meet the challenges of the future? First, the bar for licensing a new drug must be raised. Before a medicine is approved, the public should expect that enough evidence has been amassed to ensure its safety. That means completing proper safety studies before a drug is licensed. Second, pharmaceutical companies will have to upgrade their commitment to research if they are to drive innovation. Until recently the shift has been the other way, from research into marketing. Since 1997, US pharmaceutical companies have doubled their promotional spending to $18.5bn a year. But the industry seems to recognise that it must do more research. Merck, for example, increased its research budget in the US by a third to $720m in the first quarter of this year. Pfizer's global research budget last year was $5bn. The British pharma sector spent ?3.2bn on research in 2002.
Innovation does not, however, depend on a company's total research and development budget alone. The industry needs to exist, so it argues, in a strongly competitive market to encourage risk-taking in research. It is easy for a company to design a "me-too" drug to match a competitor's product and as only one in five drugs that enter early clinical trials actually make it to market, you can understand why companies might prefer less risky research strategies in order to maximise their success rate.
Genuine innovation comes from developing entirely new classes of drug. The knowledge that enables companies to do this in turn depends on government commitment to funding a strong science base in universities and research institutions. Without that undertaking, the fundamental knowledge on which the drug industry depends is absent. But the mistake that industry has made in recent years is in isolating its research efforts from those in more academic settings. Companies used to fund researchers to pursue their studies in universities as well as in industry laboratories. This collaboration was not only designed to foster goodwill with doctors who would eventually be recommending their products. The links put the best traditions of academic excellence next to the competitive instincts of industry. Historical studies of how modern treatments came to be developed show that a strong predictor of success was partnership between these two sectors, which presently see themselves as working to different values and perhaps even with differing goals in mind. The biotech sector-the small private companies often set up by academics and backed by big company money-has in some places taken over as the intermediary between big pharma and academia. But so far the results have been disappointing.
Innovations in the laboratory which flowed freely in the 1970s and 1980s seemed to dry up by the end of the century, at least for common diseases that kill millions of people every year. There is the odd exception-such as Viagra,which is hardly a lifesaver. The biotech sector has a few promising developments, especially in cancer and diabetes, to its name. The huge investment in studying the human genome will also eventually bring forth a fresh era of innovation, but this will follow the discovery of which proteins, and not which genes, are involved in the complex genesis of disease. We are at least a decade away from translating this into better treatments.
In the meantime, how can public policy help companies to overcome this innovation drought more quickly? For diseases in the developed world, the best incentive is longer patent protection, perhaps lengthened by as much as a decade, to 30 years. Patent protection would initially raise healthcare costs by shutting out cheaper generics from the market. But it would, in the longer run, provide more funds for research and take the pressure off companies to rush inadequately tested products to market.
These shifts in public policy and commercial strategy would make big pharma less of a target for critics and more of a partner with the public. But who will lead a debate to revive an industry that is presently ailing? Here, there is only a vacuum.
