The Great American Drug Deal

by Peter Kolchinsky

  Would it be conscionable to make a patient use regular human insulins from over 20 years ago given how much better today’s insulins are? If so, which patients exactly, to save the system money, should be exposed to greater risks of insulin shock or death? Physicians try to adhere to the Hippocratic Oath without letting their judgments be unduly influenced by cost, which is a luxury that our society can afford. No doubt, if patients didn’t have insurance and had to pay out of pocket, physicians should prescribe whatever their patients could afford—better any insulin than no insulin. But the essence of the Biotech Social Contract is that a society as wealthy as America’s can sustainably afford to make standard-of-care therapies available to all patients, rich and poor (and with few exceptions, to the poor in developing nations),289 as long as those payments are investments (i.e., mortgage payments) into genericizable drugs.

  Assessing the Value of Incremental Improvements

  As long as all improved versions of drugs eventually go generic, every upgrade is desirable. But it’s reasonable to ask whether some of these upgrades could be had for less, if we could have a system that matches each quantum of innovation with a commensurate incentive. How do we evaluate upgrades and incremental improvements? That’s the question that drives politicians, patient advocates, journalists, health economics, physicians, and biotech executives to quarrel over which drugs are “worth it” and how to price drugs more intelligently. It’s a critical question, and the answer reveals the inadequacies of our current systems for rewarding innovation.

  In getting to that answer, we must not base the pricing debate on the affordability of any one drug for any individual patient. At the level of the individual, the cost of a drug is purely a function of how society structures its healthcare insurance system, and we covered that in Chapter 4. Here, we are focused on whether treatment upgrades are worth the cost when that cost is borne by society as a whole.

  Keep in mind that every upgrade to a drug is patented, whether the innovation is the invention of an entirely new drug or a formulation tweak that makes an older drug more convenient. So the period of market exclusivity is typically over a decade. But the patent protection of the upgrade doesn’t delay genericization of the older version. So while some upgrades result in the displacement of prior, potentially less-effective versions, many others co-exist and even compete with their older versions.

  The Old Competes with the New

  When payers set higher copays for new versions that offer a small convenience (once-daily pill) and lower copays for the older generic versions (twice-daily pill) that are still effective when taken as prescribed, patients often prefer the generics and only some might pay the higher copay for extra convenience. Eliminating all copays would likely drive all patients to prefer the more convenient once-daily pill, which would result in a huge reward for the company for a modest innovation. So while patents govern how long a period of market exclusivity an upgraded version of a drug will have, it’s good that insurance companies can regulate patients’ access to the slightly better drug. A company, knowing that its branded once-daily pill will have to compete with generics of its off-patent twice-daily pill, will offer the payer rebates to get them to keep the copay relatively low for the once-daily brand. But switching away from a patent-based reward system for incentivizing small upgrades to a regulatory market exclusivity system discussed in this chapter would allow all patients to enjoy the benefits of the upgraded version of a drug without society paying too much. See the sidebar “Lyrica Competes with Itself” for a case study.

  Since eventually the better version will also go generic, it makes sense to preserve incentives (i.e., some financial return) for companies to keep making incremental upgrade. But the question is how high that return needs to be. This is not the same as asking what the right amount of profit a company should get for inventing any new drug. Here I’m only talking about incremental upgrades that are relatively predictable, inexpensive, and fast.

  Pursuing a cure for Alzheimer’s would be too uncertain a venture to set the kind of reward in advance that we could confidently say would justify the risk and incentivize the industry to pursue it in earnest (except maybe an extremely high reward, akin to letting the winner set their own price for 10-15 years, which is what we have now). But skilled drug developers already know what it would take to turn a twice-daily pill into a once-daily one, and therefore it should be possible to size the reward for that upgrade to the cost, time, and risk of the work. Essentially, society should commission straightforward upgrades of older drugs with appropriately sized incentives instead of relying on patents for this incremental innovation, which sometimes result in too much reward and sometimes not enough.

  Enantiomers: Small Tweaks, Outsized Rewards

  In the late 1980s and early 1990s, pharmaceutical companies figured out how to synthesize or purify especially pure versions of certain molecules known as chiral compounds. Chiral compounds are those for which multiple structural variants are possible using the same chemical formula. For example, your left and right hands are the same (four fingers and a thumb) and yet are structurally distinct such that your right hand won’t fit into a left-handed glove. In chemistry, such variants are known as enantiomers of one another, which is to say mirror-image versions of one another, and they can act quite differently.

  In some cases, drugs containing a mixture of enantiomers were not necessarily harmful; they were just less effective than drugs containing only the most effective enantiomer. In other cases, enantiomers might cause side effects, so a drug containing multiple enantiomers would not only be less effective, it might have serious consequences (as is the case with levodopa, a Parkinson’s disease treatment, one enantiomer of which causes a blood disorder called granulocytopenia).

  As with most innovation, what was once novel and special is eventually taken for granted as a reliable, inexpensive commodity. As with books, cars, computers, and phones, so with formulation technology and chemical synthesis techniques. This means that incremental drug innovations that previously were only rolled out as one-off upgrades are increasingly likely to be incorporated into the first versions of novel drugs. Chiral drugs that once might have been launched as mixtures of different enantiomers are more likely to be launched at the outset as a single, pure chemical.290 Indeed, once companies were able to use this technology to make purer versions of drugs—at scale—the FDA strongly encouraged all companies to develop only the most therapeutically relevant enantiomer of their drugs.291

  But sometimes small improvement can lead to disproportionate financial rewards. After its launch in 1989, Prilosec (omeprazole), a leading proton pump inhibitor for the treatment of heartburn292 and a mixture of two enantiomers, generated billions of dollars in sales for AstraZeneca. As Prilosec reached the end of its patent life in 2001, AstraZeneca launched a purer version called Nexium (esomeprazole). It consisted of only the effective enantiomer, which meant that a 40mg dose contained twice the effective amount present in 40mg of Prilosec.293

  Think of Prilosec being like a mixture of red and green M&Ms of which only the green ones are therapeutically active, and the red ones are useless, possibly even causing side effects. In this example, Nexium is like a drug consisting of just the green M&Ms. Not surprisingly, AstraZeneca was able to demonstrate that a 40mg dose of Nexium was more effective than 40mg of Prilosec, the highest marketed dose of Prilosec (which makes sense since 40mg of Prilosec includes only 20mg of Nexium green M&Ms diluted with 20mg of useless red M&Ms). AstraZeneca also formulated Nexium into a once-daily pill, which was a nice convenience. Once Prilosec went generic, physicians who wanted to prescribe a more effective treatment to patients for whom the 40mg dose of Prilosec wasn’t enough had two options available to them: (a) They could risk prescribing an off-label (i.e., non-FDA-approved) 80mg dose of Prilosec (40mg of green M&Ms mixed with 40mg of the useless red M&Ms, the safety of which had not been formally demonstrated),294 or (b) They co
uld prescribe the FDA-approved 40mg dose of Nexium (40mg of pure green M&Ms).295 As you would expect, most doctors and patients preferred Nexium, which remained a successful branded drug long after Prilosec went generic, generating over $50 billion of additional revenue for AstraZeneca—and drew not a small amount of public and payer ire over an old dog being paid so handsomely for its new trick.296

  Competition within a drug class to launch a best-in-class product means a company can’t afford to come to market with anything less than the best that it can do (try buying a new car without modern safety features like a rearview camera or ABS brakes). AstraZeneca was competing with other companies to treat heartburn. If it had been able to, AstraZeneca would have launched Nexium, a once-daily, pure drug, from the start, instead of starting with a twice-daily, chirally mixed Prilosec. But the ability to make chirally pure drugs inexpensively at a large, commercial scale emerged after Prilosec’s launch and therefore had to be worked into a future upgrade. Today, making pure drugs is much more straightforward. So while society paid for what seemed like small tweaks as many older drugs were re-launched in their purer forms as high-priced branded drugs, that novelty now comes standard and is paying off in new ways, sometimes enabling new drugs to come to market that couldn’t even have been developed before.

  For example, Johnson & Johnson’s intranasally delivered depression drug esketamine (an enantiomer of the impure chiral drug ketamine) was approved in 2019, despite ketamine being DEA-monitored, only approved for hospital use (for anesthesia), and otherwise considered an illegal, commonly abused street drug.297 And, although ketamine’s anti-depressive properties have been understood for several years, J&J had to conduct a full clinical development program for esketamine—with all the usual costs that come along with that and, in this case, not without significant risk. So J&J took a decades-old dangerous anesthetic and repurposed it into a proven treatment for depression by employing intranasal formulation technology and chirally pure chemical synthesis to make it safe enough to approve for a relatively large population of patients (although its drug label still contains plenty of warnings). Yet, according to those who see esketamine as a low-risk reformulation of a variant of an old drug that some people snorted to get high and that was already suspected to help with depression, J&J merits the public ire that was once directed at Nexium and the EpiPen.298

  Incremental Incentives for Incremental Innovation

  The question is this: Is $50 billion (the amount Nexium generated for AstraZeneca) a fair price for society to pay for a clinically proven, FDA-approved, once-daily, more effective derivation of a generic drug? Are there any ways by which society could have had it for less?

  One way would be for the FDA to have offered to extend any company selling an impure chiral branded drug extra market exclusivity, let’s say an extra six months of sales, in exchange for using newer manufacturing techniques to switch the drug to a pure enantiomer. At the time, Prilosec was the world’s best-selling drug, with revenue peaking at more than $6 billion/year—so every extra six months of those sales would be extremely attractive to any company. And further upgrading the drug to a once-daily and developing a higher, more-effective dose could have been incentivized with additional, similarly short exclusivity extensions. In other words, AstraZeneca could have been incentivized to upgrade Prilosec to the pure, higher-dose, once-daily version that we know as Nexium 40mg for only an extra 18 months of Prilosec revenue.299

  The idea of offering additional market exclusivity by regulatory fiat is not novel. Exclusivity periods of varying duration have been legislated to incent companies to test drugs in children, to develop drugs for small, special populations of patients, or to bring off-patent molecules to market in the US.

  The Hatch-Waxman Act of 1984 and subsequent acts, such as the Food and Drug Administration Modernization Act (FDAMA) in 1997, recognized that sometimes a new patent is impossible or can’t reliably encourage a company to develop a drug candidate into a product or to characterize it in ways that would benefit society. For example, FDAMA gave the FDA the ability to grant companies a six-month extension of an existing patent in exchange for running studies of their drugs to determine whether they could be safely and effectively used for children.

  Consider that children have different metabolisms and are smaller, yet sometimes they need anti-depressants, analgesics, antibiotics, chemotherapy, and, increasingly, drugs for diabetes and hypertension. In many cases, it was not worth the expense and risk for companies to test their drugs in the much smaller pediatric population when it’s so much easier to run trials in adults, the larger market.300 So, for example, without trials incentivized by the FDA granting six months of extra exclusivity for pediatric studies, physicians would have to guess what dose of Lyrica should be given to a seven-year old for neuropathic pain, as opposed to being able to refer to the drug’s label. It’s far better that drugs go generic after we know how to use them properly, not only in adults but also in children.301

  There are even some drugs that lack any patent coverage at the time they come to market. For example, there may be a drug that has been used in Europe for decades for a particular disorder that could also benefit American patients. However, no company would bother to go through the trouble of developing this drug for the US market, investing the money to run new trials to FDA standards, without some period of market exclusivity. Generics would simply be able to swarm in, preventing the company that did the work of bringing the drug to the US market from generating a return on its investment.

  So for these cases and a number of others, the Orphan Drug Act of 1983, Hatch-Waxman Act, and, more recently, the Affordable Care Act of 2010, offered anywhere from 3-12 years of market exclusivity for a variety of drug categories, including biologics (12 years), drugs intended for an orphan market of under <200,000 patients (seven years), a small-molecule drug not available in any form in the US for a non-orphan market (five years), or a different dose or formulation of a drug that is already available in the US (three years).

  Unlike pediatric exclusivity, which tacks six months onto the end of a period of patent protection, these other exclusivities run in parallel to a patent (and therefore are redundant if a company has a patent, which typically offers a longer period of market exclusivity). These FDA-granted exclusivities therefore only offer a fallback incentive in case there is nothing about a drug that the company can patent reliably to get the monopoly it needs to justify developing the drug.302

  Given the rules of the game, AstraZeneca’s Nexium strategy was logical. Unfortunately for society, relying on patents to incentivize Prilosec’s incremental upgrades into Nexium instead of granting incremental exclusivity extension earned AstraZeneca far more profit than would have been enough to elicit those upgrades. To any proponent of a market-based economy (as opposed to a centrally planned one run by a government that claims to know the value of all things), this may sound heretical. But I picked the extreme risk/reward imbalance of the Prilosec/Nexium example in the hope that all capitalists would agree that this innovation could have been incentivized under a different scheme. Simply granting AstraZeneca six months of extra exclusivity for the whole Prilosec franchise to upgrade it to a pure enantiomer, six months more to upgrade it to a once-daily pill, and another six months to explore the safety and effectiveness of a higher dose would have generated an extra ~$9 billion for AstraZeneca, more than most drugs generate during their entire patent life. This new best form of Prilosec would have gone generic around 2003, twelve years earlier than Nexium actually did, which would have saved society over $40 billion.

  Lyrica Competes with Itself

  Pfizer’s original version of the pain drug Lyrica (generic name pregabalin) must be taken twice a day and went generic in 2019. Patents for the once-daily version of pregabalin, known as Lyrica CR, won’t expire for another eight years. Insurance companies can leverage the availability of twice-daily pregabalin, deciding they won’t cover Lyri
ca CR unless Pfizer drops the price to something close to the generic. In response, Pfizer will no doubt offer generous copay assistance to patients who are prescribed Lyrica CR and offer PBMs generous rebates to keep the drug on their formularies (i.e., society gets a discount thanks to Lyrica CR competing with pregabalin generics). Most patients will get regular pregabalin and some will get Lyrica CR. The goal, per the Biotech Social Contract, would be for Pfizer to generate enough revenue from Lyrica CR while it’s still branded to have justified its development, after which time Lyrica CR too will go generic. What society most likely will wind up paying for this incremental upgrade will be modest compared to the investment necessary to get pregabalin in the first place (i.e., Pfizer’s sales of the original Lyrica), and yet eventually everyone who needs pregabalin will be taking an inexpensive once-daily generic version. But what if Pfizer had been offered just six more months of exclusivity for Lyrica in exchange for letting Lyrica CR go generic at the same time? All patients who need pregabalin would already be enjoying the convenience and cost-effective efficacy of generic pregabalin CR.

  Seeking New Uses for Old Drugs

  The Orphan Drug Act of 1983 was specifically passed to incentivize companies to develop drugs for smaller, orphan populations (up to 200,000 patients in the US).303 When companies can invent a new drug to treat an orphan disease, they know they will have a long period of patent-protected market exclusivity but still might feel that the risks and costs of pursuing the program are too high. In these cases, the Act offers them tax breaks, research subsidies, a more open dialogue with the FDA on how to develop the drug, and accelerated review of their application to launch the drug earlier than would normally be possible.304 If a drug has no patent protection, the law grants the company seven years of exclusivity. That’s not bad, though Europe is more generous, offering ten years of exclusivity to any new drug.