The more drugs on the shelf, the better.
For many years, high drug prices have raised the ire of politicians of all stripes, from Donald Trump to Bernie Sanders, with many others in between. The proposed solutions almost always focus on the final step in the drug development and marketing process: the sale of the drugs to various government payers. Would-be reformers argue that Medicare, Medicaid, hospitals, the military, and other institutions, such as the Veterans Administration, should be able to purchase drugs at mandated lower prices. However, drug development and marketing usually involve a decades-long process. By focusing on the entire process instead of just the final step, we can find better solutions that neither industry critics nor advocates have explored. If we address the real problem causing high drug prices, then, in the vernacular of Donald Trump, we can make drug pricing great again.
The solutions proposed by politicians, which have included mandated discounts, price negotiations, and so-called most-favored-nation pricing requirements, expose Americans to a risk that most people do not fully understand or appreciate. Yes, prices can be reduced legislatively. However, those reduced prices also reduce the incentives for drug companies to endure the expensive, lengthy, and risky drug development process. This leads to a fundamental problem. Because drugs are one of the primary factors that have made us healthier and have increased well-being and lifespans, these measures would likely make us worse off on net. Our solution is to allow more competition, and a clear way to do that is to end the requirement that drug companies show efficacy for a particular use before the drug may be sold. Doing so would not only reduce prices via competition rather than regulation, but also get drugs to market as much as a decade earlier. As a bonus, the lower cost of getting drugs to market would make it more likely that drugs could be developed for rare diseases.
Before 1962, developing a new drug took just two years; it now takes 12 to 14 years. Since 1975, capitalized drug development and approval costs have increased at 7.5% per year in real terms, doubling every 10 years. Given this growth rate, we estimate that drug development costs are now, on average, nine billion dollars per successful new drug.
Why does the process take so much longer and cost so much more? Because, since 1962, drug companies have needed to prove, to the Food and Drug Administration’s satisfaction, that new drugs are safe {and} effective for a particular disease or condition. Between 1938 and 1962, drug companies were required to prove safety, but not efficacy. With our proposal, the safety requirement would still stand. Safety, after all, is the most important aspect of any drug, given that all past drug tragedies have been about safety, not efficacy. The proof of efficacy requirement—where most drug development costs, risks, and time are incurred—provides little value for patients and doctors.
Why were the rules changed in 1962 to require proof of efficacy? Some members of Congress believed that they needed to do something after a tragedy in the early 1960s in which more than 10,000 babies, mainly in Europe and Australia, were born deformed after their pregnant mothers took a sedative called thalidomide. Ironically and illogically, because of thalidomide’s safety problem the rules were changed to require proof of {efficacy}—a puzzling non sequitur. The FDA had been charged with ensuring that new drugs were safe for more than two decades {before} the thalidomide situation. Thalidomide was effective; what it wasn’t was safe.
To understand why the FDA’s efficacy requirement provides little value, consider this quote from Ethan Lazarus, an obesity medicine and family physician, discussing the common practice of stepped therapy:
For example, we might start a generic medication, like phentermine. If it is not tolerated, or we do not get the amount of weight loss needed to improve obesity complications, we would transition to a branded oral medication, [like] Contrave or Qsymia, or switch to or add a GLP-1, like Wegovy or Zepbound.
Let’s parse his statement. He might start with drug A and, if A doesn’t work, try drug B and, if B doesn’t work, move on to drug C. He might instead start with A and, if he isn’t satisfied, add B or C.
An understanding of the practice of medicine allows us to see that doctors need more than one therapeutic option because not all drugs work for all patients. How many therapeutic alternatives? The more the better. Doctors can’t know a priori which medicine will work for which patient, and so they rely on the ancient practice of trial and error.
A clinical trial might show that a new drug provides therapeutic benefits to 60% of patients. And yet we will never know, before a patient has actually tried the new drug, whether he will be in the 60% group or the 40% group. The FDA doesn’t know. The treating physician doesn’t know. And the patient doesn’t know. Whether the FDA requires proof of efficacy before approval has no effect on whether the patient and doctor must rely on trial and error to find the right pharmaceutical at the right dose. This necessarily means that the FDA is wasting tremendous resources by requiring drug companies to prove efficacy before receiving marketing approval. Why? Because all that time, money, and energy provides little value to doctors and patients who still won’t know, even after “proof of efficacy” is established and the FDA authorizes marketing approval, whether the new drug will work for them or not.
Consequently, requiring proof of efficacy before approval is unnecessary and wasteful. It provides some additional information, but the information is expensive, and it doesn’t eliminate the necessity for trial and error.
Doctors need more than one therapeutic alternative, and the more drugs on the shelf, the better. When the FDA rejects a new drug application because the demonstrated efficacy is less than what the FDA considers “enough” for approval, it takes away a legitimate therapy option for patients. Even if the new drug is effective in only 30% of patients, less than the 60% for the existing therapy, the new drug might be effective in a substantial fraction of the 40% of patients who weren’t helped with the current drug.
To give a real example, about one in six patients prescribed a direct oral anticoagulant will not respond to the one they were first prescribed but will respond to a subsequent one. For instance, patient Jones may not respond to Pradaxa but will respond to Xarelto, while patient Smith will not respond to Xarelto but will respond to Pradaxa.
How would our proposal lower drug prices?
First, it would dramatically decrease the costs, timelines, and risk of developing novel drugs. That means that drug companies would face a lower threshold for drug development. This would increase the number of new drugs developed and reduce the revenues required for each drug to break even financially. Lower revenues required for breakeven mean that lower prices could be offered.
Second, the increased competition that necessarily comes from additional competitors has historically been the best way to keep drug prices low. Every new drug must have a value proposition—a reason to justify its purchase—or else customers will see no reason to purchase it. Consider the following scenario:
The value proposition of the first drug is that it’s the first drug of its type. If you want a drug with that mechanism of action, it’s the only one available. The second drug to launch has advantages over the first—perhaps safety, efficacy, dosing, tolerability, or convenience—and it is therefore priced higher. Its value proposition is that it’s a better drug. Next, the third drug that launches, which has no clear advantages over the other two, is priced lower, with the price low enough to create an appealing value proposition. Every product and service must have a reason for customers to purchase it. If the product or service doesn’t have a clear advantage over the competing products and services, the company will often fall back on pricing. We see this in practice. New biosimilar competitors to biologics—the biosimilar copies have nothing to compete on except price—have, on average, reduced prices by 30%. When four or more generic drugs enter the market, prices drop by an average of nearly 80%.
Will large randomized, controlled clinical trials still be conducted to demonstrate efficacy? Yes, whenever the value of that information to the drug company exceeds the cost. Currently, that assessment is never made. Why? In most cases, companies must conduct two independent, large, randomized, controlled clinical trials simply because the FDA requires it. Our proposal separates FDA approval from a determination of efficacy and, because the drug company will run a trial only if the benefits exceed the costs, interjects a cost/benefit analysis step before any expensive trials are commenced. What benefits would a drug company consider? The value of hard clinical evidence to convince potential customers of the merits of the new drug.
We predict that drug companies will choose to conduct many studies demonstrating efficacy. If this is so, has our proposal improved anything? Yes. Currently, efficacy studies are conducted to address questions posed by the FDA. With our proposal, efficacy studies will be addressed to answer questions posed by thought leaders in the medical community. Therein lies a world of difference. While there is certainly some overlap, the requirements of a bureaucracy checking off boxes are different from the informational needs of a medical community trying to understand which types of patients, if any, will benefit from a new therapy.
One particularly hard problem today is that of rare diseases. It is difficult to run the proper complement of randomized, controlled trials for such diseases because there are so few patients, and those patients might be hard to track down. The fact that the number of patients is limited means that a trial of a very good drug could fail to show statistical significance. For many drugs being developed for rare diseases, the FDA has already relaxed some of its requirements. We propose taking this one step further by ending the efficacy requirement. Our proposal would almost certainly be a boon for patients with rare diseases.
There is little benefit but huge cost in requiring drug companies to prove efficacy before the FDA approves a drug for marketing. The FDA is maladapted to helping doctors select the best drugs to promote the health and well-being of American patients. We should roll back the FDA’s requirements to what they were between 1938 and 1962—before the thalidomide non sequitur—and stop requiring drug companies to conduct the trials necessary for proof of efficacy before receiving marketing approval. Instead, those clinical trials should be run by drug companies when the value to them of the information gained exceeds the cost of the trials.
Both industry supporters and industry critics should support streamlining the process of drug development and approval to speed up the development of more affordable cures. Our proposal would lower drug prices and produce a plethora of new drugs to improve the lives and well-being of Americans now and for many years into the future. The answer to high drug prices is more drugs.
