by David Healy
The important patent changes in international drug markets came in 1960 when France, the country that had been most opposed to patenting medicines, switched to product patents, followed in 1967 by Germany, the country that had developed more pharmaceuticals than all other countries combined. Once companies knew that applying for product patents in all major countries simultaneously blocked the development of any competing products, the way was cleared for the development of blockbusters. The possibility of truly global blockbusters came in the 1980s with the creation of the World Trade Organization's Trade Related Aspects of Intellectual Property Rights (TRIPS), which extended patent protection worldwide.30 If the patent is valid, this gives a company the possibility of a monopoly on a new product worldwide for twenty years from the date of filing. From that point onward, there could only be one Lipitor, one Nexium, one Prozac, and the way was open for a company to maximize the possibilities inherent in branding—and to go as global as Coca Cola.31
Compared with the vigorous debates in France that led in 1844 to a rollback in patenting medicines and the discussion of the moves that blocked the patenting of penicillin in Britain and polio vaccine in the United States, there was virtual silence in the face of these more recent changes. No one argued that patenting and commerce were incompatible with progress in science and the principles of medicine, as they had earlier.
Several historical factors probably contributed to the silence. World War II had seen heavy state investment in medical research. This investment created partnerships between scientists, universities, and pharmaceutical companies that capitalized knowledge and contributed to the development of what is now termed the knowledge economy. This led in the 1940s and 1950s to an astonishing development of truly novel and extremely effective agents, from the antibiotics and cortisone to the diuretics, antihypertensives, hypoglycemics, and psychotropic drugs, as well as the first chemotherapy for cancer. It seemed we were set on a course in which genuine developments would succeed each other for years to come. The era of snake oil was over. Academic understanding and medical research had developed as never before, laying the basis for real progress, and pharmaceutical companies had played a part in this progress. Besides, even with the change in what could be patented, the spirit of the patent laws and the expectations of the medical community at least notionally remained aimed at providing businesses with a period of monopoly but only in return for a genuine novelty that offered a distinct benefit to the public. Such an arrangement seemed to be an engine for harnessing commercial vigor to public purpose.
It hasn't turned out that way. When assessing the patent application for a drug, the examining officer is supposed to look at whether the structure of a molecule is substantially different from compounds already on the market, and whether it provides a clear clinical benefit, a solution to a problem of medical care for which we have not previously had an answer.32 It is in the interest of a drug company, however, to argue that differences that may appear to be trivial are in fact substantial and innovative, as in some cases they are. But, if a country wishes to build up its pharmaceutical sector, as the United States was intent on doing in the postwar decades, one way to do so is to make it easy to take out patents. The notions of benefit to the community and of novelty can be shaved, so that companies might be awarded patents for trivial variations on a compound that does not clearly confer any benefit in terms of health or other public value.
Against this background let us look at the patenting of Depakote. The American patent on Depakote was taken out in 1991 but the drug in fact came from a French anticonvulsant, sodium valproate, first produced in 1962. By the mid-1960s, it was known that the sedative effects of sodium valproate could be useful in the treatment of mania. When Abbott filed for a patent on semi-sodium valproate in 1991, it was on the basis that minimally reducing the amount of sodium in the compound, which was completely irrelevant to the mode of action of the drug, made it novel. Had Abbott proposed to test this compound, which was trivially different from sodium valproate, for a hitherto incurable disorder, such a stretching of the spirit of the patent law might have been warranted on the basis of clinical need. But all Abbot planned to do was to put it into trials for use in mania, with a result that was a foregone conclusion. That Depakote was granted a patent is indicative of how lax the application of American patent law had become. The reason to go to all this trouble was that sodium valproate was now off-patent—any company could make it—and without marketing exclusivity Abbott thought that it could make little or no money.
Faced with an application for its use in mania, the FDA then licensed Depakote. Surely clinicians would not use the much more expensive on-patent semi-sodium valproate over the far less costly off-patent but essentially identical sodium valproate? Such a prediction ignores the power of the kind of branding that product patents made possible. Clinicians were faced with a brand new compound, a brand new class of drug—a mood stabilizer—and a brand new illness—bipolar disorder—and they fell hard for the package. Depakote became a billion dollar global blockbuster and manic-depressive illness was consigned to the dustbin of history, greatly increasing the costs of healthcare in the process. The success of Depakote lay entirely in Abbott's ability to distinguish between two drops of water—but it was the ability to take out a product patent with global reach that made it worth their while to do so.
In the case of Zyprexa, an antipsychotic and mood stabilizer, the story is as extraordinary. The first generation of antipsychotics ran into problems in the 1970s with million-dollar legal settlements against their manufacturers for a disfiguring neurological side effect of treatment—tardive dyskinesia. This led to a period of almost twenty years when no new antipsychotic came on the market. The only antipsychotic that did not cause this problem was clozapine, but clozapine had been withdrawn in 1975 because it was associated with a higher rate of mortality than other antipsychotics.
The way forward seemed to lie in producing a safe clozapine. There were two ways to attempt this. One was to develop a drug that bound to the key brain receptors that clozapine bound to; this method underpinned the patenting of Risperdal (risperidone) and Geodon (ziprasidone). Another way was to make minor adjustments to the clozapine molecule. Tweaking a molecule risks producing a compound with all the hazards and none of the benefits of the parent. This is what Lilly did: in 1974 the company produced a series of compounds that were all abandoned because of toxicity.
As the patent life of that series ebbed away, Lilly had to decide whether to abandon the hunt. This was a company in serious financial trouble, facing potential takeover. On April 29, 1982, they opted to move forward with a compound from the original series that by definition was not novel—olanzapine, later branded as Zyprexa. To make Zyprexa commercially viable, they needed a new patent, which meant demonstrating some benefit not found with other antipsychotics. In 1991, the only novelty presented in the company's new patent application, which was approved, was a study in dogs in which Zyprexa produced less elevation of blood cholesterol levels than another never- marketed drug.
Zyprexa has since turned out to be one of the drugs most likely in all of medicine to increase cholesterol levels in man. Lilly has settled over $2 billion worth of claims that Zyprexa has raised cholesterol and caused diabetes and other metabolic problems. There was arguably a better case to be made for patenting it to raise cholesterol than to treat psychosis.33 Lilly's patent was declared invalid in Canada, though not in the United States or Europe. Despite this, Zyprexa has been one of the biggest selling drugs of all time, grossing $4–5 billion per annum from the late 1990s through 2010. There was no basis to think this drug was any more effective than dozens of others and a lot of reasons to think it was more problematic for patients, but the marketing power that came with its patented status enabled Lilly to hype its benefits and conceal its hazards and steer doctors to write enough Zyprexa prescriptions to save the company.
In our brave new world, companies can make blockbuster profits
out of a Depakote or Zyprexa. If these two compounds were exceptions, the price might be worth paying for a set of drugs that were otherwise innovative and were leading to treatments for serious conditions that previously went untreated. Many might sigh but most would reconcile themselves to the situation—this is the way the world works. But that world does not seem to be working anymore. Where there were a handful of new tranquilizers, antipsychotics, antidepressants, and stimulants introduced annually year after year from the 1950s onward, the flow of novel psychotropic drugs dried up in the mid-1980s.
The decline of the antidepressants illustrates this all too well. The antidepressant drugs produced from 1958 to 1982 were used primarily for severe mood disorders and as such had a much smaller volume of sales than the benzodiazepine group of drugs, of which the best known, Valium and Librium, quite literally became household names—these were mother's little helpers. Valium and the other benzodiazepines were marketed as tranquilizers for anxiety from 1960 onward. In the 1980s, claims that they caused dependence led to a backlash against the benzodiazepines, leaving the market open for a new group of drugs which, however, could not be called tranquilizers as this term was now too closely linked to dependence and withdrawal. The strategy seemed clear to the major drug companies—to persuade doctors that behind every case of anxiety lay a case of depression. And to persuade them that a new group of drugs, the SSRIs (selective serotonin reuptake inhibitors), were both antidepressant and a therapeutic advance, when in fact the companies had almost consigned the SSRIs to the dustbin in the early 1980s as they were not as effective as either the tranquilizers or older antidepressants. They were also not especially novel, most of them being simple derivatives of preexisting antihistamines, many of which work as well as the SSRIs for nervous problems. Nevertheless this molecular group appeared to offer a modest but patentable amount of novelty and therapeutic benefit. The profits that came with the patent status, amounting to $15 billion per year for the group as a whole, provided the means to transform psychiatry's views of common nervous disorders—until the patent on these drugs expired soon after 2000 and clinicians had to be reeducated that the very same patients were now suffering from bipolar disorder and in reality needed a mood stabilizer.
If the SSRIs had been a bridge to a more effective group of compounds this again might have seemed acceptable, but since 2000 almost the only novelty has come from three isomers of earlier SSRIs. Many drugs come in mirror image, or left and right hand (isomer) forms. Typically only one of the hands is active. Until the 1990s it was inconvenient to attempt to separate these isomers. But then in the mid-1990s Sepracor isolated esfluoxetine and dexfluoxetine (Zalutria) from Prozac (fluoxetine). Lundbeck isolated escitalopram (Lexapro) from Celexa (citalopram), and Wyeth isolated desvenlafaxine (Pristiq) from Efexor (venlafaxine). The astonishing thing is that companies have been permitted to take patents out on these compounds, which are as alike to the parent compounds as two drops of water.34
The paucity of genuinely new drugs coming on the market in recent years is not some odd quirk of psychotropic drug development. The best-selling drug for minimizing acid secretion in the gut in the 1980s and 1990s was the proton pump inhibitor Prilosec (omeprazole). Before the patent on omeprazole expired in 2002, its parent company, Astra Zeneca, simply introduced Nexium (es-omeprazole), an isomer of Prilosec, and clinicians shifted from a cheap drop of water to an identical but vastly more expensive drop of water. If a drug does not come in isomeric forms, companies have instead in recent years patented the metabolites of a parent compound and released these as a novel drug with the approval of the US patent office and almost no resistance from medicine.
When the breakthrough drugs of the 1950s emerged, there were great hopes that not only would they offer remedies for illnesses that we did not have treatments for, but they would also shed light on the nature of the illnesses being treated. In 1896, the advent of diphtheria antitoxin had demonstrated that not all throat infections were diphtheria, thus opening up the idea that bacteriology would be able to carve a mass of respiratory, gut, and other problems into a series of discrete illnesses each of which could be tackled individually. In the 1950s, there were great hopes that the new treatments for disorders like arthritis, depression, hypertension, or schizophrenia might similarly clarify whether these were single diseases or the end result of common pathways into which several different diseases fed.
The new drugs, to use a celebrated phrase, would help us carve nature at its joints, just as distinctions among bacteria had enabled us to carve infectious diseases at their joints. But we have now arrived at almost the precisely opposite point. Rather than drugs being used to carve nature at its joints, nature instead is being used to differentiate drugs whose differences are essentially trivial.
With the new possibilities for profit opened up by a lax to nonexistent application of patent laws, the medical arena has ceased to be a domain in which scientists using new molecular tools push back boundaries. Indeed since the passage of the Bayh-Dole Act in 1980, which encouraged scientists, including medical academics, to consider patenting the products of their research, clinicians and scientists have seemed keener on making patent applications themselves and setting up start-up companies than in advancing medical knowledge or healthcare. Pharmaceutical companies meanwhile have no interest in what molecules might reveal about how humans work. Molecules are only interesting insofar as they can be used to capture market niches. Medicine may look the same as it has always done to onlookers; the marketers know it's not.
BRANDS AND PATENTS
The emergence of product patents transformed the importance of branding. By the 1980s, when H-2 blockers for ulcers, statins for cholesterol, SSRIs for depression, and other drugs were in the pipeline, branding had become so important to companies that the job was outsourced to specialist Manhattan-based companies like InterBrand and MediBrand. No longer would drugs be called Diuril or Tryptizol. We were about to get Prozac, Viagra, Zestril, and Nexium, names that bore no relation to the underlying chemical or disease but were aimed at differentiating between the new Nikes and Reeboks of the medical world and hinted at the restoration to youthful vigor that nineteenth-century brands had shamelessly promoted.
Branding now extends far beyond generating and market-testing fancy names for drugs. Brands nest within brands. The new marketers brand drug classes and diseases with far-reaching implications for medicine and society.
For instance when the makers of the new antidepressants of the 1990s needed to distinguish their drugs from older drug treatments, the term SSRI (selective serotonin reuptake inhibitor) emerged. This is not a medical or scientific term. Serotonin is a neurotransmitter in the brain, but both new and old antidepressants acted on it and the new drugs were in fact no more selective than some older drugs. The term SSRI came from the marketing department of SmithKline Beecham as part of their effort to distinguish their Paxil from Lilly's Prozac and Pfizer's Zoloft, but all three companies used the term to create the appearance of a new class of drugs and provide a common platform from which to launch marketing efforts designed to marginalize older—and demonstrably more effective—treatments.35
To this day, the brand names of drugs do not feature in medical textbooks, but these same books all include sections on statins, SSRIs, and ACE (angiotensin-converting enzyme) inhibitors as though these are medical terms, when in fact they are brand-like names that replace medical terminology. Statins such as Lipitor are just one subset of lipid-lowering drugs that include equally effective older drugs such as nicotinic acid. Zestril and its sister compounds hit the market in the 1980s as “ACE inhibitors,” rather than simply as antihypertensives, and became bestsellers as the SSRIs did—replacing cheaper and more effective antihypertensives.
One of the most striking instances of the branding of a new drug class has been creation of the idea of a “mood stabilizer.” This once rarely used term was summoned up by Abbott Laboratories in the 1990s and pressed into use in the marketing
of their newly patented Depakote. Depakote as we have seen was approved by the FDA in 1995. But it was only approved for the treatment of the manic pole of what was once called manic-depressive illness. Such approval was not surprising—giving any sedative to manic patients will produce a change that can be portrayed as a benefit. More surprising was the company's application for approval. There are comparatively few manic patients, and a lot of sedatives were already in use to manage their illness. If there was to be any serious money in Abbott's move, it had to lie in the much larger market of people whose moods could be portrayed as fluctuating unhelpfully, who were in need of “mood stabilization.” But Abbott's license did not include warrants to claim Depakote was prophylactic—they couldn't claim it would stop moods swinging—or indeed even that it was a treatment for manic-depressive illness.
However, from the start ads for Depakote carried a claim that it was a mood stabilizer. Had Abbott said prophylactic, indicating that this drug had been shown to prevent mood swings, they would have broken the law. The beauty of the term mood stabilizer is that it had no precise meaning. But what else would a mood stabilizer be if not prophylactic? And this verbal construction would lead prescription writers to use it for that purpose, even though no controlled trials have ever demonstrated Depakote to be prophylactic. Far from being a well-grounded scientific idea, the term mood stabilizer was an almost perfect advertising term— as successful a brand as the term tranquilizer had been in the 1950s and SSRI in the 1990s.
All of a sudden everyone seemed to know what a mood stabilizer was. There was an exponential increase in the number of articles in medical journals with this term in the title—from none in 1990 to over a hundred per year by 2000. Within a few years, all psychopharmacology books had sections on mood stabilizers. It was as if in the middle of a TV drama series like Buffy the Vampire Slayer the main character is given a sister she never knew she had. When it comes to entertainment we can accommodate developments like this without blinking, but it is not the kind of thing we expect to be happening in science or medicine without solid evidence.