Statement Submitted to the Senate Committee on Finance by Peter Kolchinsky, PhD
Pertaining to Senate Hearing - June 2nd, 2020: COVID-19 and Beyond: Oversight of the FDA’s Foreign Drug Manufacturing Inspection Process
Reliable, high-quality generic drugs are the great value proposition of continued biomedical innovation. They are the ultimate price control on branded drugs and a unique phenomenon in all of healthcare, where nothing else goes generic – not hospitals, not services, not surgery.
Drugs are a manufactured good. The high prices of branded drugs that are necessary to incentivize investment in risky research projects are like a finite set of mortgage payments. Once a mortgage is paid off, America takes ownership of an inexpensive public good. Through their taxes and insurance premiums, our parents paid for branded drugs and passed them on to us as inexpensive generics, as they might a home. Over 90% of all prescriptions in America are for generic drugs and each new drug we invent to improve our standards of care is built on a foundation bought and paid for by past generations.
Consider how easy it is to save money on medicines thanks to generics. If your cholesterol is high, your doctor might prescribe Lipitor, using the brand name of Pfizer’s long-generic drug out of habit instead of the generic name “atorvastatin.” No worries. Your local pharmacy is permitted by law to fill your prescription using pills made by any FDA-approved manufacturer of atorvastatin. Your pharmacy does the shopping around for you, playing dozens of manufacturers off one another to get the lowest price. You do not have to worry about which company makes your atorvastatin because it is now a commodity; everyone has to make it to standards of bioequivalence defined by the FDA. You probably care more about the manufacturer of your toilet paper than you care about who makes your atorvastatin, presumably because you trust the FDA.
False Economy
But here is the dilemma: It is hard to manufacture drugs reliably, consistently, and to the highest standard. A company has to care about getting the conditions just right and run quality tests at every stage to make sure that the intermediates and final product are exactly as they should be. And while the company that sells a new, branded drug for a high price has a strong profit motive to keep quality high, especially because it has to prove to physicians that this new medicine can be trusted, the same cannot necessarily be said for generic drug companies.
Generic drugs do not always work as well as they should, and globalization has greatly exacerbated this problem. Generic drugs have not earned and do not deserve our blind trust. As transplant surgeons, cardiologists, infectious disease specialists, and psychiatrists have increasingly recognized, a generic version of an essential medicine manufactured by one company can be more or less potent than the original version or a generic manufactured by a different company. A generic might not have the stated amount of an active ingredient. It may contain deadly impurities. In some cases a generic might release a day’s worth of drug into the bloodstream all at once; in other cases it might release a drug too slowly. That could mean an infection otherwise easily controlled might instead turn deadly. Blood pressure or high cholesterol could remain unchecked. Or in the case of an organ transplant, a patient taking a generic version of an immunosuppressant might lose their precious new organ to rejection. Not all generics are substandard; some are manufactured correctly and behave the same as branded equivalents. But a growing number of Americans are hesitant to take that risk, and rightly so. Some physicians will only prescribe the branded version of a drug, which often spurs a protracted fight with insurance plans that, like the FDA, consider generics interchangeable with branded drugs and only want to pay for low-cost generics.
One key reason why generics are unreliable is because they are increasingly manufactured overseas, where labor costs and regulatory bars are low. This shift has put factories that manufacture generic drugs practically beyond FDA oversight, which has long been crucial to holding companies accountable for quality standards. In a competitive market that takes quality for granted and prioritizes lower costs, it only takes one bad apple out of a dozen manufacturers to drive all the honest players out of business. As detailed in Katherine Eban’s eye-opening book, Bottle of Lies, the FDA cannot frequently or adequately inspect the Indian and Chinese manufacturing plants that produce so much of our generic drug supply. Instead of making high-quality medicines, many of these companies circumvent quality regulations through an escalating game of cat-and-mouse with the FDA. Quality testing might be done on samples of Pfizer’s own Lipitor, with the resulting data passed off as evidence that these factories’ atorvastatin generics work just as well. When companies are run by people without integrity and regulators cannot hold them accountable, the dark side of human nature can flourish. In this case, cutting costs comes at the expense of Americans’ health, health worldwide, and the value proposition of biomedical innovation.
I am not suggesting that generic drugs made in America are inherently safer because Americans are more ethical. But what I find somewhat reassuring is that drugs made in America are made on the FDA’s home turf where the leading drug regulator in the world can do a more effective job of monitoring quality. In the US, where it can conduct surprise inspections, the FDA issues plenty of warning letters: more than 50 in the last 12 months related to Drug Quality Assurance. But the violations overseas are extreme and all the worse considering that the FDA typically gives several weeks’ notice to companies that its inspectors are coming; time used to clean up their operations, or, in some cases, cook their books and coach employees to lie.
An American consumer can hold a local pizza shop accountable for having a dirty bathroom or rancid cheese by writing a bad review and eating elsewhere. But Americans have no such power to demand American-made drugs manufactured under FDA supervision. Just try it.
Look in your medicine cabinet to see which companies make the generic drugs you find there. Odds are good that they are based in India. Now google that company’s name together with the term “483” and odds are good you will see that all of these companies committed drug quality violations in the last 12 months. You might ask your doctor or pharmacist to fill your next prescription only with generics made in America. But neither will know how to do that. Your pharmacist is driven by competitive forces to purchase drugs at the lowest possible cost, which increasingly means from overseas manufacturers. If she even has any control over where her pharmacy purchases its atorvastatin, she would be hard-pressed to order from a more expensive US-based manufacturer if her competitors are ordering cheaper Indian-made drugs. Other than writing a prescription for the branded version together with the phrase “Dispense as Written” (to ensure that the pharmacist does not instead dispense a generic), there is nothing your doctor can do. But before you think that prescribing only branded drugs is the solution, consider the fact that this will make most drugs unaffordable. Insurance plans balk at paying brand prices for drugs that have gone generic unless a physician personally fights for an exception, which few have the time to do. The system is fundamentally based on the idea that, for a given drug, all generic versions available in the US are equally safe and effective as one another and the brand. The consumer has no choice!
Covid and Contracting: An Opportunity
Policymakers are considering repatriating America’s drug supply chain to avoid future shortages in the context of temporary disruptions caused by COVID-19. But they should think bigger. We should repatriate the American drug supply to restore and preserve the integrity of all generic drugs in America.
That might seem like overkill. Better quality testing of final products, as the pharmacy Valisure has started doing, might suffice in some cases. The trouble is that the kinds of analytical assays done in a lab can only detect some of the defects in a poorly made generic. A drug can be made so that it meets all the lab test specifications, for example, containing the right amount of the active drug and dissolving at the right rate in a beaker, and yet it could still be made in such a way that it acts differently in the human body. That is why the FDA requires that generics companies prove their final products behave very similarly to branded originals in human bioequivalence studies. Once they do, then the process that makes that product must be locked down and the company must document that it remains religiously adherent to that process. The presumption is that, as long as they do, then the final product can be trusted to continue to act as documented in humans. But if generics manufacturers never even made their generic to appropriate specs, lied about their human bioequivalence data, or did not continue to adhere to a process that might have been the right process initially, then it is not a given that subsequent lab tests will detect the failure of their final products to function in patients as intended. Those drugs might still cause dose dumping or not release enough drug at the right point in the gut, resulting in potentially important differences in therapeutic outcome. The bottom line is that generic manufacturing must be done to careful specifications, and Americans need both the quality of the final products and the processes by which they continue to be made to be carefully monitored by a regulator, which in the current framework is the FDA. That can only be done truly reliably in the US (though close, trusting cooperation with certain countries that have similar concerns about low-quality generics is conceivable).
Repatriating the entire competitive generic drug market as it exists today would be counterproductive. We can rebuild the market smarter and more cost-effectively by using long-term procurement contracts: the model we already trust to ensure that America has pandemic flu vaccines and drugs for smallpox. This idea is already gaining traction. Just recently, the federal government awarded a four-year $354 million contract to Phlow, a Virginia-based generic drug manufacturer, to produce certain essential generic medicines. That is the right idea. Now we need to scale that approach many times over with several other companies.
This type of long-term contracting can achieve better quality and might also result in even lower prices than we have today. No doubt competition achieves the lowest profit margins possible under free market principles. But the fixed costs stack up. Even with low profit margins, the total costs of maintaining all those competitors can be high. Instead of trying to get twenty different generics manufacturers to all produce atorvastatin and compete on price, America can negotiate long-term contracts with a smaller number of companies, allowing them to enjoy greater economies of scale and greater profits while America pays less overall. Let’s call this “Contractual Genericization.”
Say 20 companies end up making generic atorvastatin in the US, competing on price such that no one company makes very much in profit. But each of those companies has to cover its fixed costs without enjoying economies of scale, and the FDA has that many facilities to inspect to ensure high quality. Instead of 20 companies each having to cover $5M of fixed cost to make a drug ($100M total), we might have two companies serve the US market that each only need to cover $20M of fixed costs ($40M total) due to economies of scale (though each company would still manufacture the drug at two or more locations to mitigate against shortages) – two management teams instead of twenty; fewer, larger factories, instead of many smaller ones. Because of the $60M difference, America could pay those two companies more generously under contract, allowing them to make more profit on an absolute basis, and America would still save money compared to the free-market competitive system. Their greater profitability would give them a strong incentive not to screw up, since failure could mean transfer of the contract (essentially management of high-quality manufacturing facilities) to another company.
Still, there is no doubt that the underlying costs of making drugs in America are higher than doing so elsewhere. Even under long-term procurement contracts, American-made generic drugs might not be less expensive than the ones we get now. But when the drugs we get now are not actually reliable, then whatever low price we pay for them is too high. Better to pay what we must for generic drugs that actually work as intended than pay less for inferior and dangerous products.
The real benefit to Americans of onshoring generic drug manufacturing through contracting would be that we could rely on much tighter quality controls. Not only would generic drugs be American-made, creating American jobs by companies paying taxes in America, but they would be high-quality American-made. And with proper funding and incentives, innovation and automation of manufacturing – so-called advanced manufacturing techniques – can help reduce the costs of producing America’s drug supply on American soil.
Repatriating generic drug manufacturing does not require that every atom of every drug be made in America. American manufacturers could still purchase commodity chemicals from overseas, provided that they were able to pivot to other sources in case of disruptions and could guarantee quality control of those chemicals before they go into final products. The final steps of packaging exactly the right chemicals in the right combinations at the right pressures in the right formulations with the right coatings has to happen on US soil, where FDA oversight is rigorous.
We can’t do this all at once. We need to start by repatriating our drug supply from countries like India and China that have poor track records of adhering to Current Good Manufacturing Practices (cGMP) per FDA requirements. We can leave countries like Ireland (where many drug companies manufacture products for lucrative tax breaks) for last and decide at that time whether it is prudent to cut ourselves off entirely from foreign production.
It is also important that high-quality generics are available globally. As it is, many countries do not trust generic drugs, and rightfully so if one considers that overseas manufacturers have a history of sending the best of their tainted goods to the US and selling the worst to Africa, South America, and their own countries. If the US led the way in making high-quality generic drugs, we could end up exporting them to other countries. At the very least, out of pure self-interest if not compassion, America should ensure that everyone in the world gets the proper doses of all antibiotics, since underdosing leads to antimicrobial resistance that could land on our shores.
Generic Drug Contract with America
The framework I am proposing, the Generic Drug Contract with America, would accomplish at least three important goals:
Restore the quality of the American generic drug supply by repatriating most or all generic drug manufacturing to the US where the FDA can keep a close eye on the process.
Protect the American drug supply from disruptions like we have seen due to COVID-19.
Create tens or even hundreds of thousands of high-quality American jobs where they are most needed by tying government contracts for American-made generics to requirements (with federal subsidies) that these companies build their factories in regions and communities that have been hollowed out by globalization.
This model can also help solve another growing problem that Congress has not even begun to contemplate. Some drugs cannot go generic under our existing legal, ethical, and regulatory frameworks. Increasingly, the technologies we are using to treat diseases are complex and some are near-impossible to copy. If we cannot trust generic manufactures to make atorvastatin reliably, they may never be able to make some of the antibody-drug conjugates and gene therapies that have recently come to market and are on the rise.
The cost of branded drugs that will go generic are worthy, finite mortgage payments that America makes towards a medicine that it will eventually own as a public good (i.e. generic). But the costs of ungenericizable drugs are rent from day one. Companies that sell ungenericizable drugs need never worry about a patent cliff. They need never hustle to invent a new drug to replace lost revenues when older ones face competition. They can just keep making the same thing indefinitely. And while even branded drugs often compete with other branded drugs (for example, there are several statins, several SGLT2 diabetes drugs, and several insulin analogs), having two or three competitors in a class makes for an oligopoly, which still can tacitly collude to extract rents. But there are many cases where only one drug in a class is particularly well suited to some patients and therefore represents a true monopoly. Reliable generics are the true disruptors of oligopolies and monopolies, but that path is not available to certain types of drugs.
The order established by the Hatch-Waxman Act in 1984, which introduced the idea of modern generics that are interchangeable amongst themselves and with the original branded version, did not contemplate biologics or how such drugs could ever go generic. Recombinant insulins launched in the early 1980s, and other biologics only really started coming to market in the 1990s. Yet not until the ACA passed in 2009 with a component called the Biologics Price Competition and Innovation Act (BPCIA) did we get a pathway for biosimilars. Biosimilars are as close to generics for biologics as we can get and yet far from close enough.
The trouble is that the BPCIA was always on weaker ground than Hatch-Waxman in terms of driving cost savings because it is harder to establish interchangeability for a biologic than for a small molecule drug (though that is arguably not as easy as we once thought either). Without interchangeability at the pharmacy level, an insurer cannot just have a pharmacist switch all patients over from a branded biologic to its biosimilar and save a plan money. That is because insurers (and even their employer clients) are addicted to the rebates they have negotiated on branded biologics, which they risk losing by covering cheaper biosimilars of a given drug without the assurance of successfully saving money by having patients actually switch to those biosimilars.[1]
If trying to create a US-based competitive market for generic atorvastatin seems hard and expensive, then doing it for insulins and antibodies will be impossible. For example, Abbvie’s Humira, an autoimmune disease antibody, is the most lucrative drug in history. It’s been on the market for 18 years, much longer than the 10-15 years typical of more easily genericized small molecule drugs. It does not make sense to wait until five or ten other US companies figure out how to make Humira before America considers its $15B/year mortgage on that drug paid off.
The most reliable manufacturer of any biologic is the company that has been making it for years as a branded drug. The same contracting mechanism that we use to contract with a few companies to make the US generic drug supply can also be used to contract with biopharmaceutical companies to continue to make their biologics at a contracted price once their patents expire – one that is low, but remains profitable. Yes, that is a price control. But it is a fix for the market’s failure to achieve the same end through competition.
The Generic Drug Contract with America would achieve a key fourth goal:
Ensure that even conventionally ungenericizable drugs become inexpensive after their patents expire, bringing modern drugs in line with the intent of the groundbreaking Hatch-Waxman Act of 1984 that established generic drugs as we know them.
Just as Hatch-Waxman allowed for exclusivity extensions to incentivize certain kinds of development of branded drugs, such as demonstrating how they should be used in children (i.e. six-month pediatric extension), similarly this Contractual Genericization would allow for exclusivity extensions to incentivize companies to continue to upgrade their branded drugs.
The Great American Drug Deal
The biotechnology industry I know thrives on solving healthcare problems. Its efforts to stop Covid are on full display, ingenuity flowing like water through the cracks of a rock, searching out its weaknesses. This same ingenuity is also targeted at thousands of more familiar healthcare problems. Hepatitis C infection can now be readily cured with a short course of pills. The lives of many cystic fibrosis patients have been transformed over the past five years with oral medications. At the pace we are developing new treatments, there is a good chance that a mother giving birth to a girl today will never have to worry about her daughter dying prematurely of breast cancer.
This scientific hustle exists on a foundation of public funding for basic science and some public support for drug development. But it is largely fueled by private capital: the cumulative savings of the world, from billionaires to schoolteachers’ pension funds, searching for an attractive return. And it is America’s willingness to pay high prices for new drugs during their patented period of market exclusivity that is the draw for all this private capital. The fact that other countries pay less than we do for branded drugs is a function of their willingness to deny breakthrough medicines to their citizens and America’s resolve to back innovation even if it has to do so on its own. But America relies on drugs going generic to ensure that our country gets value for its investment. And the fact that branded drug revenues are finite keeps scientists and investors constantly working to develop the next breakthrough. Thanks to the mortgage model brought about by the Hatch-Waxman Act, the biotechnology industry evolved into a community of builders that charge finite mortgages. If generic drug quality remains unreliable and newer drugs remain difficult or impossible to genericize, the biotechnology industry risks regressing into rent-extracting landlords.
For a free market to work, it cannot allow monopolies to extract high rents indefinitely. America created the FTC to regulate companies that grew into natural monopolies, in some cases breaking them up. America also passed laws that regulated natural monopolies, such as PURPA in 1978, to make sure Americans are not price gouged by electric companies.
As the medical historian Jeremy Greene points out in Generic, his excellent overview of how America’s generic drug model was negotiated into existence, while other countries tried to control the price of branded drugs, America kept its drug costs in check by requiring branded drugs to go generic. That has been our model and it has worked to control costs and to incentivize innovation. Without genericization, America would be spending hundreds of billions of dollars more per year on branded drugs. Because of generics, America spends about $271B on branded drugs and only $73B on generics. As long as we ensure that all drugs go generic, through competition or contract, then if we are still spending $271B on branded drugs in 2035, it is because all currently expensive drugs have become inexpensive generics and the biotechnology industry has invented an entirely new set of branded drugs. All of these brands will necessarily have to be better than the generic drug foundation on which they stand. We are builders and what we are building will ensure that our children and grandchildren live healthier lives than we do.
We need a Generic Drug Contract with America to ensure the quality of our generic drugs, to ensure that all drugs go generic and offer America value for its investment, to protect the incentives that drive further biomedical innovation from being throttled by the alternate and far worse measure of price controls on branded drugs, to ensure America’s drug supply against global disruption, and to revitalize America’s heartland with good jobs.
Of course, it is impossible to talk about drugs without also addressing affordability. As we call for quality generics, we must also call on our society to make appropriately prescribed branded drugs affordable to patients via proper insurance, which means not allowing out-of-pocket costs to exceed what patients can afford. Consider why an insurance plan demands that a physician get prior authorization before prescribing a drug and then, upon deciding the drug is appropriate for the patient and granting the authorization, still requires a high out-of-pocket cost. Is that insurer nudging that patient to disregard her physician’s recommendation? That is not insurance and therein lies America’s problem with affordability: it lacks a proper healthcare insurance system. The entire drug industry operates on profit margins in the low teens, which means that a price cut across the board of just 20% would wipe out profits and company valuations (and with them the value of executive stock-based compensation), and yet does absolutely nothing to make a $10,000 drug affordable to a patient that struggles with a $6,000 deductible (which means that even if a drug costs $8,000, the patient would still have to cover the entire deductible).
So the Generic Drug Contract with America is really part of the greater Biotech Social Contract, which requires insurance reform to lower out-of-pocket costs for all Americans and extend insurance to everyone. Amid the Covid crisis, the federal government has already seen the wisdom in covering Covid-related costs for the uninsured, a logical public health measure. Hopefully those temporary patches on the gap-ridden system of American healthcare coverage will be made permanent and expanded to other diseases, because cancer, diabetes, and many other disorders are no less a personal crisis for each patient who suffers from them and their families. Indeed, there are reform bills in Congress that propose caps on out-of-pocket costs and pair those limits with reforms targeting drug manufacturers. This Generic Drug Contract with America embraces the genericization of drugs, ensuring that they are efficiently manufactured at high quality in the US, as the reform necessary for the biopharmaceutical industry to hold up its end of the bargain.
There will be those who oppose the Generic Drug Contract with America. They may have good reasons. However, we should dispatch the ones that are clearly untrue. This is not an assault on free trade; this is a necessary response to unreliable trade. This would not be an unprecedented incursion of government price regulations into pharmaceuticals; the government already contracts with individual companies to make the entire US supply of particular drugs where there could not be a reliable and cost-effective “free market,” such as pandemic vaccines and other biodefense countermeasures. California is already exploring the possibility of making all of its own generics. The federal government has contracted with Phlow to make many generic drugs that normally would be sourced from overseas.
I am proposing that we think bigger and, over the course of this next decade, make bold strides to repatriate most if not all of our generic drug supply under contracts. We must ensure that we always can look forward to paying off the mortgage of a drug to take possession of an inexpensive, reliable, public good while innovators move on to the next set of upgrades of our medical armamentarium. And someday those too will go generic.
Footnotes
[1] If a payer spends $200M/year on a branded drug but gets a 40% rebate ($80M) in exchange for exclusive formulary status, that payer spends a net $120M on the branded drug. If a biosimilar comes along at a 65% discount, then converting all patients to it would result the drug costing the payer $70M a year, $50M less than they are spending now. That is attractive. But if only 10% of patients switch to the biosimilar, then the payer spends $7M on the biosimilar and $180M on the branded drug (since the rebate would be gone), resulting in greater cost. That is how a rebate without interchangeability protects a brand’s monopoly from biosimilars.