No Refills

In 2009, only 25 new drugs were approved—less than half the number in the mid-’90s. Why are new pharmaceuticals so hard to bring to market? Overcautious regulators and profit-hungry conglomerates make easy scapegoats, but they’re only partly to blame. While we’re waiting for both sides to reinvent themselves, even little things like better monitoring of side effects can lead to big new discoveries.

The golden age of drug discovery.” That’s how Martin Mackay, Pfizer’s president of pharmatherapeutics research and development, recently described the state of his industry. That surprised a lot of drug researchers, many of whom feel more like the Indians than the conquistadors. Although Mackay said the Pfizer pipeline was “replete” with new drugs, nothing on the horizon seems capable of replacing the revenue his firm will lose when Lipitor, the best-selling patented drug of all time, goes off patent at the end of next year. Pfizer is not the only company whose pipeline is tapping out. What most people think of when they imagine “drug discovery” is what the Food and Drug Administration calls “new molecular entities” (NMEs)—entirely new compounds. And the number of these being approved as drugs has declined.

In 1996—arguably the peak of pharmaceutical productivity in the past two decades—the FDA approved 53 NMEs. These days, breaking the 20 mark is rare; last year 19 were approved, plus six “biologics,” substances such as vaccines and antibodies that are based on proteins made by living cells. Most analysts seem to think that U.S. companies just aren’t turning out as many valuable new drugs as they used to.

How can this be happening now, just when we’ve decoded the human genome, developed new techniques for synthesizing drug compounds, and built powerful computer models of how compounds interact with their targets? Why aren’t the pipelines overflowing with new blockbusters? That’s a good question. “Blame the FDA,” say conservatives and drug companies. “Blame the companies,” respond liberal activists and good-government types. But as viscerally satisfying as such finger-pointing may be, it offers half-truths at best. For a variety of reasons, we may be entering an age when a miracle cure is just that.

Ask a conservative—or a chemist whose compound is in the advanced stage of clinical trials—about drug development and you’ll get an earful about the FDA. Over time, critics say, high-profile disasters like Fen-Phen and Vioxx, which killed or seriously harmed some of the people who took them, have encouraged ever-more-stringent review. The number of clinical trials required to support a new-drug application has more than doubled since 1980, while the number of patients needed in each trial has almost tripled. As a result of these and many other factors, the clinical-trial stage now costs more than four times as much, even after adjusting for inflation.

Making drug trials more expensive can have a big effect on development. Once they’re developed, most drugs are nearly pure profit (a topic of much complaint among consumer activists). The cost is all in the R&D—somewhere between hundreds of millions and nearly $2 billion per drug, depending on which estimate you use. And the single biggest portion of that cost is the very, very expensive clinical trials, in which pharmaceutical companies try to show the FDA that their compound is safe and effective.

This means that clinical trials have unwanted side effects. Because of their astronomical expense, one drug with a huge market is more commercially desirable than 25 drugs that each treat a less common disease, because only one set of trials is necessary. If you’re targeting a disease that affects relatively few people, one of two things will happen: the drug will be very expensive, or the drug will be shelved because it’s unlikely to earn back its R&D investment.

Tougher safety and efficacy standards may also be keeping good drugs out of the public’s hands. Most people agree that today’s FDA would not have approved aspirin; even penicillin, the miracle drug that helped dramatically extend the human lifespan when introduced in the early 1940s, is questionable. Allergic reactions to penicillin kill a higher percentage of its takers than Vioxx ever did, while the gastrointestinal bleeding produced by aspirin means it probably would have flunked while still in animal testing.

But there’s another side to the story. For starters, the stodgy FDA may be changing. Roger Longman, a former managing partner at Windhover, a health-care-analysis firm, says that while “the rank and file is sort of resistant” to streamlining its clinical-testing system, “there’s an interesting split at the FDA. At the top of the FDA, they’re not getting more conservative—in fact, I think that they’re going the other way.”

Besides, as Pedro Cuatrecasas, a biochemist and professor of pharmacology at the University of California at San Diego, points out, attrition rates—the percentage of compounds that fail in clinical trials—haven’t actually increased. That makes it hard to blame overfussy regulators for the decline in new drugs—though Derek Lowe, a pharmaceutical researcher who also writes about the industry, notes that firms do take the agency’s newer and stiffer requirements into account when deciding whether to take a drug candidate into trials. However, Lowe adds, some of these safety requirements are less a reflection of increasing FDA risk aversion than of our ability to test for more things. Even without the FDA breathing down their necks, lawsuit-shy companies have plenty of disincentives to pour millions into a drug that causes gastrointestinal bleeding.

Hence the left’s counterclaim to conservative complaints about the FDA: that companies, not regulators, are the ones who are too risk-averse. Pharmaceutical companies are blamed for focusing on marketing at the expense of innovation: they allegedly kill promising compounds because of fears of small markets, then concentrate on “me too” drugs that aren’t really any better than their competition’s. Or they tweak existing drugs in ways that don’t necessarily make the drugs any more effective, but do give companies a new patented drug for which they can charge the Earth. (Remember how we got Clarinex around the time Claritin was going off patent?) Then, when pharma companies have exhausted these strategies, they shore up their finances by merging with other companies and downsizing their research staff.

The complaints about mergers hold some truth. Even without the customary lab closings and shelved projects, mergers can trigger uncertainty and personnel turnover, causing good projects to languish. And yes, pharmaceutical firms do bring out a fair number of tweaks to existing drugs, or old drugs approved for new disease indications.

But those who want to chalk up empty pipelines entirely to mismanagement may be confusing cause with effect. Firms don’t merge because it’s fun; they merge because a big hole in their pipeline can lead to financial troubles. Mergers spread the risk. Companies naturally start obsessing about costs and marketing when their products become less profitable.

Besides, if companies have really stopped caring about R&D, why is it that according to Joseph DiMasi, an economist at Tufts who specializes in the pharmaceutical industry, spending on R&D is 11 times what it was 30 years ago? (To be sure, R&D spending has dropped in the past year, but it is not clear whether that change represents a new trend or a reaction to the financial crisis.) And why are firms having so many dramatic failures in Phase III, when the compound is tested on a large number of patients and failure is most costly? If they were really so conservative, wouldn’t they be killing those projects a lot earlier?

In fact, while some companies are diversifying into generics and animal medicines, Longman thinks that overall the industry is spending not too little but too much on massive internal research teams hunting for the next blockbuster drugs. He believes the major pharmaceutical companies should make investments in lots of promising compounds at biotech firms—sort of like taking an option on a drug.

No firm has yet taken Longman’s advice and dramatically shrunk its internal research operations. But some companies, notably GlaxoSmithKline and AstraZeneca, are using external partnerships with smaller firms to replenish their pipelines. So far, however, this approach hasn’t been enough to stop the slide. Which raises the worrying question: What if creating new drugs is just getting harder and harder?

In many ways, the task facing researchers is simply more difficult than it was 20 or 30 years ago. Back then, chemists had big fat targets like angiotensin, a protein that causes blood vessels to constrict. Scientists knew it caused high blood pressure, and better yet, they were pretty sure they could develop a small molecule (that is, one that can easily enter the bloodstream) that would hit what they were aiming at. Best of all, hypertension provided an enormous market. Driven by similar finds in other areas, an age of blockbusters dawned: the Lipitors and Prilosecs and Allegras.

These days the targets seem smaller, fewer, and farther away. The best-understood diseases already have a lot of good drugs treating them. New treatments need to prove that they have better efficacy, fewer side effects, or something like a longer-lasting dose that makes them superior to the pills already on the market. Longman likens this process to chasing an Olympic sprinter—who has a head start.

In other words, complaints about me-too drugs are overblown, since they’re actually harder to get approved than something with a novel target. Because clinical trials are the most expensive part of development, such projects are still pretty risky. When you’re up against nothing, it’s relatively easy to show that you’re more effective than the alternative. But when you’re up against already-state-of-the-art treatments, you’re looking for small improvements. That means you need huge numbers of patients to generate a statistically significant result. And since good treatments already exist, the safety hurdles are also higher—the FDA is less likely to approve a statin that causes internal bleeding than a pancreatic-cancer drug that does the same.

Meanwhile, in the areas where we don’t have good drugs, we don’t have so many easy targets, either. The great hopes for finding drug prospects by decoding the human genome have largely faltered; so far, reading our DNA seems mostly to have taught us how little we still understand about our own biology. So researchers are left with complex problems like cancer, which is really not one disease at all, but several thousand different ways that a human cell can go wrong. Or Alzheimer’s: after decades, researchers are still trying to decide whether the disease’s signature beta-amyloid plaques in the brain are a cause of dementia, or a side effect.

Of course it is true that advances in genetic science have yielded some relatively simple targets to shoot for: diseases like Gaucher’s, which is caused by a hereditary deficiency in a single enzyme, and which can be treated with a synthetic version of the missing substance. But it is almost axiomatically true that simple genetic defects with debilitating or fatal consequences tend to be relatively rare. The more-prevalent conditions that we’d really like to find blockbusters to cure are, increasingly, complex metabolic processes like obesity, which may be caused by many things going wrong at once.

Drugs are still being developed for those simpler, rarer conditions—Genzyme, among other companies, has made a good living developing compounds for diseases like Gaucher’s. But the task is harder, and more expensive. A year’s worth of Gaucher’s treatment can cost $200,000.

As Longman suggests, it’s hard for big pharmaceutical companies to target such little markets. Genzyme’s $4.5 billion in annual revenue is less than Pfizer takes in just on Lipitor. Big Pharma isn’t set up for micro-research.

Then there are the customers. Drug-plan managers like Medco and Caremark have gotten more aggressive. Drugs that 10 years ago would have been blockbusters—like the anti-platelet drug Effient, Eli Lilly’s answer to Plavix—now face tough scrutiny. In fact, Medco is funding research that it hopes will make the medical case for the continued use of Plavix (which is cheaper) by most patients, while identifying the subset of patients who would benefit from the new drug. That would be great news for health-care costs, but terrible news for Eli Lilly.

The fact is, no single pipeline theory works well on its own; they all interact. A stodgier FDA can mean stuffier decisions inside companies. Smaller pipelines make for bigger mergers. Bigger companies, in turn, may mean smaller pipelines.

But the way that all these things are intertwined might actually make it easier, rather than harder, to boost our research output: any change has ripple effects. If Big Pharma can look outside its own walls more, and if the FDA can reinvent itself, the whole landscape may well alter. And though we might not quite understand our genome now, give us another decade, and it may start yielding some interesting drug targets for the diseases that we’d still very much like to cure.

Even the smaller targets we’ve already identified may yield more drugs if we can hold down the cost of clinical trials. Strategies like outsourcing our drug trials to countries with lower costs have already helped. And Vijay Vaitheeswaran, the health-care correspondent for The Economist, believes that we will soon be employing electronic reminder systems to ensure that patients take their pills, and monitoring systems that can better track side effects. These two changes alone could make smaller clinical trials more viable and effective; other innovations may follow.

None of this may work—Panacea, after all, was just a Greek goddess. But even if nothing works, look on the bright side: at least we won’t have to pay for so many pricey new drugs.