Last week, the Wall Street Journal editors wrote a stirring defense of the Food and Drug Administration’s decision to approve Biogen’s new drug Aduhelm. Aduhelm was approved to treat Alzheimer’s. Critics of the federal approval argue that there is no direct evidence that Aduhelm will cure Alzheimer’s. Supporters agree.
So what’s the issue? The Journal‘s editors put it well:
Nobody has said Aduhelm is a cure, but it is the first treatment following hundreds of failures that has shown evidence in clinical trials of removing amyloid plaque—a hallmark of the disease—and slowing cognitive decline.
Critics are right that it’s not clear that amyloid causes Alzheimer’s. But the leading research hypothesis is that a buildup of harmful amyloid plaque in the brain triggers a cascade of chemical changes that interfere with neuron communication and cause brain loss. Critics also note that some two dozen drugs aimed at removing amyloid have failed to meaningfully affect the course of the disease.
The Journal continues:
Yet many trials hadn’t properly screened patients to ensure they had Alzheimer’s. Some drugs were also tested on late-stage patients who had irreversible brain loss. And some failed to clear amyloid because they didn’t target the right molecules in the brain.
Biogen learned from these failures, and the FDA noted that Aduhelm was the first drug to show “proof of concept” in an early stage trial that clearing amyloid could slow decline. This was supported by a Phase 3 trial in which patients receiving the highest dosage showed 25% to 28% less decline in memory and problem-solving compared to the placebo group after 78 weeks.
In short, there’s a good case for allowing it. As more and more people use Aduhelm, we’ll get more information. In a few years, we’re likely to know much more.
But there’s a wrinkle. If Medicare pays for this drug, whose positive effects in treating Alzheimer’s are somewhat speculative, we the taxpayers will be forced to pay. And the price is not low. The Journal quotes Senator Ron Wyden (D-Oregon):
It’s unconscionable to ask seniors and taxpayers to pay $56,000 a year for a drug that has yet to be proven effective,” Oregon Sen. Ron Wyden tweeted after the Food and Drug Administration approved Aduhelm this month.
Ask? I wish. Taxpayers don’t get “asked.”
How does the Journal deal with Wyden’s point? By claiming that having Medicare negotiate a lower price is an imposition of a price control.
It’s not. When a government negotiates a price, the other side is free to reject it. It’s not a price control. It would be a price control if the federal government said that no one–not an insurance company and not a patient–is allowed to pay more than $X for a drug. I don’t think that’s why Wyden is proposing; it’s certainly not the thinking of many of us want Medicare not to have an open-ended commitment to paying whatever a drug company charges.
For many years now, the Journal‘s editors have argued that any restriction on how much Medicare will pay amounts to a price control and they’ve repeated the argument many times. But repetition doesn’t make the argument stronger.
Here’s an earlier post where I criticized the editors on that point.
READER COMMENTS
Scott Sumner
Jun 30 2021 at 11:41am
Very good post. My position has been that the drug should be legal, but not subsidized at all. That would mean that Medicare would refuse to pay for it until more solid evidence of effectiveness was available. (I suspect the drug is not cost effective.)
Right now the “legalize” and “effectiveness” decisions are lumped together by the FDA, but logically they are very different decisions that should be based on a different set of criteria. (I’d legalize all drugs.)
In areas like health care and banking, conservative opinion that is basically pro-business is often disguised as being “pro-market”.
Garrett
Jun 30 2021 at 11:48am
It’s a shame how often the two are conflated. I consider myself to be pro-market but many of my opinions happen to be anti-business, anti the business of extracting economic rents in most cases.
Mark Z
Jun 30 2021 at 12:16pm
The argument against this (which I hear frustratingly often from scientists themselves against various new medical technologies) is that only rich people would be able to afford the cutting edge treatment. Of course if new treatments are uncertain in their effectiveness and thus not worth subsidizing, then it just means rich people are being voluntary guinea pigs for new drugs. I’m personally fine with that.
Thomas Lee Hutcheson
Jun 30 2021 at 12:06pm
In principle either the insurer (Medicaid, in this instance) or the FDA could do the benefit cost analysis. From what is written, it seems that it would fail. The price at which the drug would be supplied would of course enter the decision, so it’s possible that it would be worthwhile if the price were low enough.
BC
Jun 30 2021 at 9:58pm
Why shouldn’t the doctor and patient do the cost benefit analysis? The patient knows his or her risk preferences and opportunity costs better than anyone — no one else knows how much the patient values an x% probability of success vs. other things that the patient could buy with the money — and the doctor can supply the technical expertise. I am talking here about the patient spending his or her own money on the drug, not Medicare or insurance company money.
Thomas Lee Hutcheson
Jul 1 2021 at 9:49am
But doctor and patient are almost never spending the patient’s own money.
Jon Murphy
Jul 1 2021 at 12:35pm
Why are we assuming money is the only cost here? Indeed, if it is true that “doctor and patient are almost never spending the patient’s own money,” then we’re likely not in an equilibrium situation, which means monetary prices are pretty much useless as a proxy for costs and benefits.
Christophe Biocca
Jun 30 2021 at 12:40pm
There already exists an agency responsible for deciding what Medicare covers: the Centers for Medicare and Medicaid Services could chose not to cover the drug, or only cover the drug conditional on participation in a further medical trial meant to settle the question of whether the drug is actually effective.
They rarely exercise this power because anything that makes it past the FDA is usually worth paying for unconditionally.
Dylan
Jun 30 2021 at 1:45pm
We will get more information…but that doesn’t necessarily (or even likely) lead to more knowledge. I think we’re unlikely to have any clear cut answers on Aduhelm’s effectiveness, and it will be in Biogen’s best interest to make sure that’s the case. Anyone willing to bet on how long it takes Biogen to do the post-marketing trial that is required as part of the approval?
Jon Murphy
Jun 30 2021 at 6:05pm
Why wouldn’t it be in Biogen’s best interest to make sure their drug has clear cut answers?
Christophe Biocca
Jun 30 2021 at 7:27pm
The argument I’ve heard to that effect is that they effectively p-hacked their way to FDA approval of an ineffective drug and know it, so now it’s just a matter of keeping the money rolling in for as long as possible until evidence that the drug is ineffective becomes to clear to ignore. That theory predicts they’ll wait until the very end of their ~10 year deadline to hand in a report.
I feel like such a move would be playing with fire, especially for a company that has more than one approved drug already. They’re already racking up Congressional attention before there’s evidence of them having done anything that nasty and I don’t think they’re so stupid as not to have predicted that would happen.
It think it’s more likely they actually believe their drug works for at least some patients, in which case the study will start soon-ish (within a year or two). Time will tell.
Dylan
Jul 1 2021 at 7:56am
Christophe is right that the belief is that the drug is very likely ineffective, while carrying increased risks to patients and I don’t think Biogen will be in a hurry to prove that unless external forces like Medicare decline to pay for the drug unless those studies are completed.
It’s not like this strategy is unusual, the last really controversial approval (Sarepta’s Exondys for Duchenne muscular dystrophy) had a required postmarketing study that didn’t start for years after it was supposed to, and still hasn’t completed. And, they got a 2nd controversial (read, likely ineffective) drug approved now for the same disease in the meantime.
Jon Murphy
Jul 1 2021 at 12:34pm
This is a response to both Christophe and Dylan.
I’m not really sure that accusations or suspicions of perfidy regarding statistical trials gets us to the conclusion that they would want to suppress information about the effectiveness of the treatment.
If they were truly being perfidious, why even bring the drug up for review at all? If you’re hiding in the dark, shining a light on yourself is not a wise course of action.
If they were not being perfidious, then they would want to clarify questions regarding their treatment for exactly the reason you state: so Medicare would pay for it. Also, so that other competitive firms couldn’t undermine their market share by releasing something better.
Christophe Biocca
Jul 1 2021 at 2:36pm
Are we asking about the trials that were part of the original FDA approval which they sought (and got)? Or about the postmarketing clinical trial?
If the former, that’s the only way to be able to legally sell the drug at all. Legally selling the drug has many advantages compared to trying to sell an unapproved cure for Alzheimer’s on Silk Road or whatever the current alternative is, including not going to jail if caught, being able to get private and possibly public insurance plans to pay for it, getting doctors to recommend it without risking losing their medical license, etc.
But I suspect you’re asking why they’d do a postmarketing clinical trial. In which case the answer is “because they were required to as part of the approval they got”. The degree of freedom they have on that front is that they can either start the trial early-ish, or delay it until close to the very end of the ~10 year period they have. Which one they choose is a decent indicator of what results they expect to have.
Jon Murphy
Jul 1 2021 at 2:54pm
Actually, I meant the FDA approval process. If you’re jucing the stats, you wouldn’t seek approval.
Dylan
Jul 1 2021 at 3:58pm
Jon,
I feel like I’m missing something in your logic. I don’t mean to suggest underhandedness by Biogen. The fact is, most drugs don’t work, and that goes triple (at least!) for Alzheimer’s drugs. If Biogen does another full trial for Aduhelm, failure is the most likely result. That’s just the way it is with every trial.
As to why go to the FDA at all with the data package they have, I think it was considered a calculated gamble, and one that has paid off for them. There’s a ton of pressure to approve anything for Alzheimer’s and they bet that they could get the drug over the finish line even with data that was pretty lacking. So now the question is, do you gamble again and take a big risk that it fails, and approval is rescinded. Or, do you go with what you have and see how many billions of dollars will roll in? My guess (and the guess of a lot of others who know pharma) is they do the latter. I’m willing to entertain anyone who wants to bet the other way though. (And I say this as someone who is generally pretty reluctant to bet on future states of the world)
Dylan
Jul 1 2021 at 4:10pm
A couple other things worth noting:
As of fairly recently, only about half of all required post-marketing studies were done 5 to 6 years after approval. And 1 in 5 hadn’t even started by then. The expectation for Biogen is no way outside of the norm.
A recent study (which I’ve misplaced the link to at the moment, I’ll try and find it again and post in a reply) looked at around 600 non-required post-marketing studies that pharma voluntarily ran, and found that most of them were underpowered, open label, and didn’t have a comparator. These are generally considered as marketing studies, and the standard is that the doctors keep the patients on the drug after the study ends.
Jon Murphy
Jul 1 2021 at 4:37pm
This is a response to Dylan’s comment of 3:58, but for the sake of organization, I am responding here.
Dylan, you said:
My mistake. I had assumed you were agreeing with Christophe’s comment:
P-hacking is a huge no-no. It’s not quite fraud, but it’s about as close as you can come. So, my comment about perfidy was a response to the claim about p-hacking. If Biogen was truly and intentionally p-hacking then they would not want to go up for approval as it would invite the very skepticism and attention one would want to avoid.
Dylan
Jul 1 2021 at 5:59pm
What Biogen has done is posthoc subset analysis, which is a form of P-hacking, one that is incredibly common in pharma. Best practice is of course to follow up with a trial and confirm the results. Sometimes though a drug company will try to get approval on the basis of subset analysis, it doesn’t normally work, but sometimes it does.
I don’t think this is particularly malicious, I’m sure that there are people at Biogen who are convinced the drug offers some benefit. However, I will be shocked* if the people in charge are convinced enough to take a chance of running a trial that could prove them wrong. And I’m willing to put money on it.
*This has gotten enough publicity and is a big enough budget hit to Medicare, that there could be a situation where they won’t cover it until there’s more data. I didn’t know this was even possible until reading Christophe’s link. I think it is unlikely that they would fully deny coverage, and I expect that Biogen will think that even partial coverage and the revenue they get from it will be too much to risk by running the post-marketing trial. I hope to be proved wrong though.
Alan Goldhammer
Jul 1 2021 at 8:21am
This approval was a total joke. Biogen proved nothing at all in their submission to the FDA and dared the Agency to not follow the Advisory Committee. The leadership at the FDA blinked and approved it with the requirement for a post-marketing efficacy study. Biogen have nine years to do this study while they can make a lot of $$$$. The management at Eli Lilly must be regretting their decision to pull the plug on their Alzheimer’s program which was far better funded and innovative than Biogen’s.
My bet is that we will never know whether the Biogen drug works or not until we have an objective endpoint correlation rather than subjective evaluation of memory and cognitive decline. The same issue occurs with SSRI anti-depressants. We know they sometimes work but it’s all subjective evaluation. There has been a huge industry/government funded effort to look at neural imaging in Alzheimer’s patients but I’ve not seen any results that point to resolution of this issue.
The WSJ editors are wrong. My big question is whether Medicare can sue Biogen and claw back money when/if it is shown the drug does not work. I can’t wait to see the advertising program that Biogen launches for this drug. I also hope they have a good pharmacovigilance program as this drug does have some very unpleasant side effects from the clinical trials.
Floccina
Jun 30 2021 at 2:33pm
This gets to the whole issue of “death panels” which are really refuse to waste taxpayers money on treatments that are not cost effective panels.
The GOPers acted like crazy socialists on that issue to try to scare old people away from the Democrats.
BC
Jun 30 2021 at 10:20pm
“When a government negotiates a price, the other side is free to reject it. It’s not a price control. It would be a price control if the federal government said that no one–not an insurance company and not a patient–is allowed to pay more than $X for a drug..”
Agree for the most part. However, Medicare is not voluntary. It’s possible that people would prefer to pay more than $X, or buy an insurance policy that would pay more than $X, for a drug but they are precluded from doing so because they were forced to spend the money that they would have used to do so to pay for Medicare instead. From what I understand, Medicare is a huge buyer in the healthcare market, mainly because participation is mandatory. “Price control” might be the wrong term. But, the mandatory nature of Medicare does distort prices whenever Medicare decides to cover, or not cover, a particular drug at a particular price.
Similar comment on whether Medicare should “subsidize” a potentially ineffective drug. Given its mandatory nature, all Medicare patients are not equal. Some are subsidizers while others are subsidizees. So, in some cases, Medicare is subsidizing a drug, while in others it’s just returning a person’s money back to them. One really has to look at the specific patient’s lifetime Medicare taxes vs. their actuarily fair premiums at the time that they paid those taxes. Even though insurance companies routinely make these types of calculations, for some strange unknown reason, Medicare doesn’t seem to try to collect that data.
Alan Goldhammer
Jul 1 2021 at 8:27am
There are different types of programs in Medicare. Even if one if enrolled, pharmaceutical coverage can be provided outside of Medicare. We get our coverage through my former employer and it’s better than Medicare Part D. It’s complicated in the case of the Biogen drug as it is covered under Part B that deals with infused drugs. Medicare is not forced to cover every patient and can make a coverage decision about which patients are eligible to receive the drug. Also if one is in a Medicare Advantage plan, there can be further restrictions placed on the use of the drug.
Charley Hooper
Jul 1 2021 at 6:00pm
There are a number of things to consider here.
(1) Most Americans who at risk of Alzheimer’s or dementia, or who are in the early stages of decline, are on Medicare because of their age.
(2) Aduhelm is reimbursed under Medicare Part B because it is physician administered.
(3) Medicare Part B is partly funded through premiums; taxpayers aren’t completely on the hook. Part B premiums might rise due to Aduhelm.
(4) Hospitals will game the system by getting deserved and undeserved discounts for Aduhelm under the 340B program.
(5) Medicare has a mechanism to limit the reimbursement of Aduhelm. Medicare can provide a national coverage determination (NCD) and specify for which types of patients Aduhelm will be covered. Absent an NCD, Aduhelm will be reimbursed under local coverage determination (LCD), meaning patients here and there will get it while others don’t. Coverage will be spotty.
(6) Even if Aduhelm is given at essentially no cost to patients, patients will still be the hook for the large costs of diagnostic tests such as PET scans to show that they are eligible. Paying $3,000 for a PET scan will dissuade many patients and limit demand.
(7) If Aduhelm works, it will be worth the cost Biogen is charging. If it doesn’t work, it’s not worth much.
(8) No one really knows if it works. Clinical trials are often underpowered, too short, conducted in artificial settings, etc.
(9) It takes a really long time to determine is a drug prevents Alzheimer’s disease. What we’re seeing here is a limitation of the current approach to drug approvals: it’s really hard to show true efficacy in a reasonable amount of time in a reasonable number of patients.
(10) If insurance, here Medicare Part B, operated more like normal insurance and it covered only a portion of Aduhelm’s cost, then those patients who found it appealing could pay for it while those who didn’t wouldn’t.
(11) Approval is based on a centralized federal government agency. Coverage is based on a centralized federal government agency. Because of this, we miss the power of individual decision making and experimentation.
(12) Biogen has a monopoly on Aduhelm, but not a monopoly on all Alzheimer’s therapies. Because so many potential customers are covered by Medicare, CMS (the agency that runs Medicare) is a monopsonist. What is the proper price and quantity for a product in a monopolist/monopsonist market?
Note that I did some consulting work for Biogen years ago, but nothing related to Aduhelm.
Ken P
Jul 1 2021 at 7:42pm
Legality
For starters, we need to separate approval from payment. On that point, I agree with Scott Sumner above that the drug should be legal: “Our bodies. Our choice.”
Cost
I don’t like using the scary $56k retail price. Analysts expect the actual drug cost to between $10k and 20k. This is an important difference in price. Insurance companies/Medicare don’t pay list price.
It’s also important to know that this cost results from the many treatments/visits required to complete the course. So even though someone else pays, the end user is paying with time and inconvenience. If doing cost benefit analysis, Medicare/insurance would need to compare to the reduction expeceted in in-home healthcare or nursing home care (how many months average reduction to equal $10-20k?). In many cases, the incentive might be off on a time frame. If you treat a person in their 50s to save costs when they’re in their 60s, it could be insurance paying to avoid medicare costs. I’m not sure what the expected treatment ages are, though.
Thomas Strenge
Jul 2 2021 at 12:26pm
One thing that is missing is a discussion of the alternatives. Does Aduhelm prevent, delay or reduce the need for institutionalization of Alzheimer patients? A year of in patient care in a long term care facility can run between 50k to 100k per year. Medicare does not cover these costs. Many can’t afford this. This forces many families to have a member drop out of the workforce to provide at home care. In the end, there is no free lunch.
Dylan
Jul 2 2021 at 2:53pm
An Alzheimer’s therapy that worked would almost certainly be a bargain by reducing spend in other areas, and that’s before considering the human cost of the disease. However, we have little reason to suspect that Aduhelm does anything to change the progression of the disease. Indeed, the Ph III trials were stopped for futility, after seeing that it was having no impact on the designated endpoints of the trial.
Comments are closed.