Aduhelm: An Opportunity for Prescription Drug Reform
The FDA should approve more drugs faster, but we should stop forcing Medicare to pay for cost-ineffective drugs.
Executive Summary
American consumers bear the ever-increasing cost of prescription drugs. The problem is particularly urgent for those at or below median levels for income and wealth.
According to the Medical Expenditure Panel Survey, households at or below the median income level in 2020 spent an average of $1,605 on pharmaceuticals and related insurance claim payments per person. For seniors in the bottom half of income, prescription drug spending is especially high at more than $3,000 per person.
Put simply, we should lower the hurdles to FDA approval of new drugs like Aduhelm, while also liberalizing mandates that effectively force Medicare and Medicaid to pay for any new FDA-approved medicine, regardless of its price or value.
High prescription drug spending naturally results in greater strain on the public health programs that serve low-income Americans. While the Centers for Medicare and Medicaid Services (CMS) projects overall U.S. prescription drug spending will increase 62 percent in the coming decade, prescription drug spending for Medicare and Medicaid combined will likely increase by more than 87 percent over the same period.
While the United States leads the world in generic drug use to achieve savings, certain brand name drugs represent a disproportionate share of drug spending that is driving overall cost growth. Aduhelm (aducanumab), the new Alzheimer’s disease treatment from drug maker Biogen, is thought to be one such example: when the FDA approved the drug in June, 2021, treating the average Alzheimer’s patient would cost roughly $56,000.
To the consternation of many experts, the Food and Drug Administration (FDA) approved the treatment despite the FDA’s expert panel voting overwhelmingly against approval. The panel cited a lack of evidence for the drug’s effectiveness in slowing or reversing cognitive decline due to Alzheimer’s while also dramatically increasing the likelihood of side effects such as brain bleeding.
Numerous researchers, providers, and advocates protested Aduhelm’s initial price point, not only because of the drug’s questionable efficacy and safety profile, but because of the financial impact to Medicare patients generally, who experienced the largest increase in Part B premiums for 2022 in large part because of Aduhelm. In addition, patients at lower rungs of the socioeconomic ladder face higher risks of incidence and mortality from Alzheimer’s. While Biogen later lowered the list price by almost half to $28,200, the out-of-pocket cost under Medicare can still reach $5,640 annually.
Amid the controversy, CMS convened a six-month review to determine whether and how to cover Aduhelm. Following the review, CMS released a proposed coverage arrangement that would cover Aduhelm and any future treatments in Aduhelm’s drug class only for patients enrolled in follow-up clinical trials subject to CMS requirements.
While CMS’ coverage decision was praised by Aduhelm’s opponents as a common-sense restriction that effectively bars use of the drug, the decision represents the exception for Medicare coverage of physician-administered drugs, not the rule. In fact, Aduhelm presents a case study of the flaws in public insurance drug coverage and why the CMS decision likely will not end the controversy with Aduhelm or other high-cost drugs approved in the future.
In this paper, I propose a better path to address coverage for prescription drugs for Medicare and Medicaid. Put simply, we should lower the hurdles to FDA approval of new drugs like Aduhelm, while also liberalizing mandates that effectively force Medicare and Medicaid to pay for any new FDA-approved medicine, regardless of its price or value:
- Enact the Conditional Approval Act, which would enable the FDA to approve more drugs after compelling phase II clinical trial data, reducing the cost and risk of pharmaceutical R&D.
- Liberalize Medicare’s coverage mandates by liberalizing rules that force Medicare and Medicaid to pay for any FDA-approved drug regardless of its price or clinical value.
- Reform Medicare’s payment formulas to incentivize drug price competition.
By leveraging market-based tools, we can streamline CMS’ coverage decisions, and thus ensure Americans have access to drugs that provide clinical value at affordable prices.
A brief history of Alzheimer’s research
An estimated six million people aged 65 or older have Alzheimer’s dementia. As the U.S. population ages, these estimates may swell to 12.7 million by 2050. It is the sixth leading cause of mortality in the U.S., but ranked as high as the third leading cause of death among the elderly prior to the COVID-19 pandemic.
Alzheimer’s disease is unique among the top causes of mortality because there is no known medical procedure or pharmaceutical agent that can prevent, slow, or cure the disease. Given the scope and scale of the problem, the pharmaceutical industry has invested billions in Alzheimer’s research and development for more than 30 years. Yet there are only a handful of FDA-approved drugs on the market, and none of them do more than reduce the severity of symptoms for a relatively short time.
Despite relatively few treatment options for patients, the drug development pipeline for Alzheimer’s disease is rich with candidates targeted at various mechanisms of action. In early 2021, prior to Aduhelm’s approval, there were 152 active clinical trials testing 126 drug treatments. The vast majority of these treatments are intended to be disease-modifying; that is, they seek to alter the underlying pathology of Alzheimer’s rather than simply relieve symptoms.
Aduhelm is the first FDA-approved drug that treats a suspected underlying cause of Alzheimer's: the formation of plaques in the brain composed of the protein amyloid beta.
The amyloid beta theory
In 1999, researchers at the biotech company Elan Corporation released a landmark study that showed immunization against the formation of protein plaque prevented the development of Alzheimer’s in the brains of mice before the onset of symptoms. Their research also showed that the clearance of such plaques from symptomatic mice reduced the extent and progression of the disease.
Since then, many in the Alzheimer’s research community have assumed that amyloid beta plaques are a defining hallmark of Alzheimer’s and, as a result, researchers and manufacturers developed numerous drug candidates to target them. Although some of these candidates have successfully targeted and cleared amyloid beta in humans, none have showed a clinical improvement in patients’ cognition.
It is possible that several biomarkers of Alzheimer’s, like rogue proteins or infectious diseases, may play a role in the disease. The presence of intraneuronal neurofibrillary tangles (NFTs), primarily composed of tau proteins appears to be one such biomarker, and researchers are increasingly interested in developing drugs to target tau. Perhaps the common thread among all Alzheimer’s pathologies — and many other neurological disorders — is simply inflammation in the brain. Drug candidates treat inflammation in different ways, including strengthening blood vessels in the brain, clearing bacterial infections, and inhibiting cytokines.
FDA’s controversial decision to approve Aduhelm
Despite the public health crisis posed by Alzheimer’s disease and patients’ desperation for new treatments, the elation from Alzheimer’s patients and advocacy groups following the FDA’s approval of Aduhelm quickly gave way to controversy and heated debate that continues today.
Prior to Aduhelm’s approval, scientists debated whether clearance of amyloid beta plaques slows or reverses Alzheimer’s.
Over the years, several drug candidates — including Pfizer’s/Johnson & Johnson’s bapineuzumab, Eli Lilly’s solanezumab, and Merck’s verubecestat, had failed to produce benefits in clinical trials. Biogen’s Aduhelm (aducanumab) showed promise of clinical benefit in early-stage trials, in part because the drug is highly effective at clearing amyloid beta plaques.
Biogen conducted two phase III clinical trials concurrently, named EMERGE and ENGAGE. Biogen ended both trials early for futility, meaning the trials indicated the drug would not produce a clinical benefit.
Following the abrupt ending of each trial, Biogen mined the data of the trials and discovered what they believed was evidence of clinical benefit — a slowing of cognitive decline — among participants in the EMERGE trial. Biogen argued that the EMERGE trial differed in that it administered a high dose of aducanumab to a subset of trial participants for a longer duration of the trial. In contrast, while some participants in the ENGAGE trial received a high dose, they did not receive it for as long as those in the EMERGE trial.
When digging deeper into the data, however, outside researchers have identified several inconsistencies that undercut Biogen’s putative causal relationship between a longer duration high-dose regimen and slowing of cognitive decline. For example, researchers noted that the rate of cognitive decline among patients on placebo were different in the two trials — a difference that not only could not be attributed to dosage levels of Aduhelm, but instead pointed to underlying differences in patients with mild cognitive impairment that could skew the trial results of the trial.
More importantly, both trials showed a large and statistically significant effect in clearing amyloid beta from the brain. One showed a clinically meaningful difference in slowing cognitive decline, and the other did not. Therefore, from a statistical standpoint, there’s no way to tell which trial gives the true signal of the drug’s effectiveness.
It was with this inconclusive data that the FDA’s Peripheral and Central Nervous System Drug Advisory Panel (PCNS) met in November 2020, voting 10–0 (with one vote “uncertain”) that the trial evidence was insufficient to recommend approval of Aduhelm. Given the PCNS vote against approval, many industry analysts believed the drug’s fate was sealed. In spite of the advisory panel’s negative vote, Biogen continued to meet with the FDA through a back-channel campaign — dubbed by internal corporate documents as “Project Onyx” — to highlight Aduhelm’s positive results in the EMERGE trial as justification for the drug’s approval through the FDA’s accelerated approval program (AAP).
On June 8, the FDA shocked the industry by announcing Aduhelm’s approval through the AAP. In exchange, the FDA gave Biogen nine years to prove the drug’s efficacy against cognitive decline through confirmatory trials.
CMS’ coverage determination
Shortly following Aduhelm’s accelerated approval, public attention turned to the $56,000 annual cost of the drug and its impact on federal and state budgets. An estimated 85 percent of Alzheimer’s patients are in the Medicare program. In addition, Medicare drug policy guarantees that virtually every FDA-approved drug is covered, with reimbursement rules favorable to pharmaceutical companies looking to drive larger profits.
The impact from Aduhelm’s approval and initial price point was especially felt by low-income seniors, even before a single Medicare beneficiary received the treatment. The drug’s potential to treat a large number of seniors at a high price prompted The Centers for Medicare and Medicaid Services to announce a 14.5 percent increase to Medicare Part B premiums for the 2022 plan year — the largest increase ever — with half of the increase made as a contingency to cover Aduhelm.
Shortly after approval of Aduhelm, CMS initiated a six-month review to determine whether and how to pay for the drug, known as a National Coverage Determination (NCD). On January 11, CMS released the draft NCD on Aduhelm. In the draft, CMS allowed the drug to be covered under a particularly restrictive version of the “coverage with evidence development” (CED) scheme. Coverage would be limited only to Medicare patients who took part in randomized clinical trials under CMS direction. Furthermore, the decision would not only apply to Aduhelm but to and all future anti-amyloid monoclonal antibody treatments for Alzheimer’s disease.
Future sales of Aduhelm rest largely on the Medicare decision, which is scheduled to be finalized in April 2022. Since Aduhelm’s approval under the accelerated pathway last June, the drug has treated no more than a few hundred patients as providers and insurers alike rejected the drug over efficacy and safety concerns. The drug was also refused overseas: the European Medicines Agency declined to approve the sale of Aduhelm throughout the European Union.
The move by CMS to restrict coverage of Aduhelm is unprecedented given the agency’s history of covering practically any drug approved by the FDA. The decision to limit coverage to a relative few in clinical trials is therefore considered by many analysts to be a near death blow to the drug’s prospects.
The majority of those who have submitted public comments to CMS ahead of a finalized determination in April show broad support for the proposed NCD. Despite the agency’s laudable intent to protect taxpayer money and ensure Aduhelm’s effectiveness, the proposed coverage announcement has also exposed several weaknesses in the NCD process.
- Distorts CMS’ role in drug evaluation. Notably, the FDA and CMS have a different sense of what it means for patients to have “access” to drugs. While the FDA grants access to drugs based on whether they are effective and safe, CMS considers coverage for drugs that are “reasonable and necessary.” This distinction becomes apparent when considering past NCDs. CMS typically required a manufacturer to enroll patients in registries to track outcomes while deferring to the FDA’s requirements that the manufacturer completes confirmatory clinical trials. CMS abandoned this distinction, however, when it ordered clinical trials to prove Aduhelm’s efficacy as part of its coverage restriction, and set forth requirements for how those trials should be conducted. While CMS intentions may be justified, its primary charge is to organize payment for health care services, not to lead, organize, or evaluate clinical trials. CMS’ lack of experience in this regard is evident from the terms of the proposed NCD. For example, CMS did not stipulate or allude to possible benchmarks the drug must reach to meet its standard that using Aduhelm results in a “clinically meaningful difference in decline in cognition and function.”
- Leaves CMS open to lobbying influence. The NCD process follows a traditional regulatory framework, in which proposed agency decisions are open to public comment before a final decision is issued. As the swell of public comments to CMS has shown, issuing an NCD with a restrictive CED on a drug as scrutinized as Aduhelm leaves CMS open to substantial outside influence from stakeholders. Biogen and sympathetic stakeholders have time to exert influence on which clinical benchmarks will result in coverage, or to simply convince CMS to broaden coverage before trials are completed. Such an outcome is not without precedent: In 2019, CMS initially proposed a NCD requiring a CED for a class of drugs known as chimeric antigen receptor (CAR) T-cell therapies to treat cancer. The CED required placing patients on registries and conducting studies that included two years of patient follow up. But months later, CMS reversed its decision and largely approved payment for the drug class.
- Downstream effects on Medicaid. The proposed NCD notes that CMS will pay for the cost of the drug in clinical trials; an outcome that is unnecessary because Biogen was paying for confirmatory trials required by the terms of the FDA’s accelerated approval. But an even larger concern is that the NCD does not apply to the other public insurance program CMS operates: Medicaid. While Medicaid employs preferred drug lists to extract pricing concessions from drug makers, it is statutorily required to cover any drug from manufacturers that participate in Medicaid’s drug rebate program, including Biogen’s Aduhelm.
Financial consequences to Medicaid
Unlike Medicare, Medicaid — the public health insurance program for the poor — does not have the NCD process as a tool to limit or circumscribe coverage. John Coster, director of pharmacy in CMS’ Center for Medicaid and CHIP Services, told state Medicaid directors last fall that Medicaid is required to cover Aduhelm because of Medicaid’s drug rebate program coverage requirement.
Still, some analysts believe CMS has the tools to limit or deny Medicaid’s coverage of Aduhelm in a similar way to the NCD process in Medicare, speculating that it will use the exceptional case of Aduhelm to issue waivers to states to deny coverage. However, it is unclear how or under what statutory authority CMS could issue such waivers. And CMS has not indicated a willingness to do so, perhaps given the unusual precedent such action would set.
One avenue that may create an opening for states to deny coverage of Aduhelm may be the Section 1115 Demonstration Waiver authority. Such waivers give states latitude to design Medicaid programs outside the rigid confines of the Medicaid State Plan, the standard array of benefits and program features states are required to administer. Tennessee is the only state that has a closed drug formulary, allowing the state to exclude drugs entirely through its Tenncare III waiver, while still using a Medicaid drug rebate program to extract price reductions from drug makers.
But the 1115 waiver approach has downsides as well. The waiver process is time-consuming for states and often requires concessions to CMS in exchange for the flexibilities states desire. In Tennessee’s case, the Trump administration approved the Tenncare III waiver in its final days, likely on the condition that the state would adopt a block-grant model for cost savings. There is no word yet whether the Biden administration would allow states to operate similar waivers, including those with closed drug formularies. CMS could allow states to set medical necessity criteria for Aduhelm’s use, including requiring preauthorization. Still, the data suggest the cost to Medicaid — including state budgets — could be substantial.
A portion of Medicare enrollees are also enrolled in Medicaid to receive services Medicare will not pay for, including services that assist with activities of daily living like eating, bathing, and dressing. Normally coverage of Aduhelm for these individuals, known as “dual-eligible” beneficiaries, would first be paid by Medicare. However, because of the narrow NCD issued by CMS, state Medicaid directors worried the NCD would force Medicaid to pick up the tab for Aduhelm.
Despite these concerns, Medicaid will not cover Aduhelm for dual-eligible beneficiaries not enrolled in clinical trials. Because the NCD prohibits coverage of the drug for such individuals, Medicare regulation reclassifies Aduhelm from a Part B drug to a Part D drug, regardless of whether Aduhelm is administered in a physician setting or whether the individual is enrolled in a Part D plan. And by law, Medicaid is prohibited from covering the cost of any Part D drug.
However, early-onset Alzheimer’s disease occurs in individuals younger than 65, before they become eligible for Medicare. It is possible that some of these individuals are eligible now or in the future for Medicaid, and therefore would gain coverage for Aduhelm as long as they meet prior authorization requirements imposed by states.
Based on an estimated prevalence of 200,000 Americans with early-onset Alzheimer’s under age 65, and the percentage of those enrolled in or possibly eligible for Medicaid, I estimate up to 50,000 people with early-onset Alzheimer’s under the age of 65 could gain coverage of Aduhelm through Medicaid.
The amount Medicaid would reimburse for each Aduhelm patient would vary based on the outpatient drug reimbursement formulas used in each state. In general, states reimburse for the cost of the drug ingredients based on the lowest cost option among a selection of prices such as the wholesale acquisition cost (WAC), the average sales price paid on the commercial market (ASP), or the “usual and customary” price (defined as the price a cash-paying person would pay at the point of sale). In addition, a professional dispensing fee, typically around $10–$15, is added to the ingredient cost. Finally, the manufacturer is required to pay back a rebate to Medicaid based on a statutory formula in order to be covered.
It is important to note that, because Aduhelm has seldom been sold in the United States, there is little to no reliable data on how much Aduhelm sells for after discounts and rebates in the commercial market, preventing states from calculating the drug’s ASP. In addition, Biogen has little incentive to negotiate with private insurance in order to keep prices high in Medicare should the drug eventually be covered. The net result is most state Medicaid agencies could end up paying at or near the list price for the drug while also capping the size of the rebate.
Take California’s Medicaid program: If Biogen decided to charge its list price for Aduhelm, the ingredient price California would pay for the average Alzheimer’s patient under its reimbursement rules is $2,171.40 per treatment. After adding in the state’s $13.20 dispensing fee and subtracting for the drug’s estimated rebate, the total reimbursement comes to $1,683 per treatment. Since a patient would receive around 13 infusions per year, the annual cost of treatment is $21,879; a discount of only 22.4 percent relative to the list price.
With these figures in hand, it is easy to see how profitable Medicaid coverage of Aduhelm would be for Biogen. Assuming half of the aforementioned 50,000 Medicaid enrollees indicated for Aduhelm take the drug, and the national average reimbursement for each treatment matches the rate for California, Biogen would receive nearly $550 million in net annual Medicaid payments — among the highest drugs paid by Medicaid.
In the future, the impact of Aduhelm on Medicaid spending could be even more substantial, oddly because of Biogen’s decision to reduce the drug’s list price in half last December. Based on current Medicaid drug rebate laws, single-source brand name drugs would normally have to pay an additional rebate if the manufacturer raises the price of the drug faster than inflation. To calculate the rebate, the difference in the drug’s current and original price is compared to the inflation rate over the same period. Given that Biogen cut the price of Aduhelm in half from its original price of $56,000, it could raise the price up to this original amount and pay no additional rebate. Ironically, Aduhelm could be excluded from payment by almost the entire market in perpetuity — including private insurance and Medicare — and still become a billion-dollar blockbuster drug only by treating a relatively small population in Medicaid.
Policy recommendations
The FDA’s approval of Aduhelm is a striking example of the need to reform prescription drug coverage for Medicare and Medicaid. Outside the limited use of the NCD process, virtually all FDA-approved drugs are covered in Medicare Part B regardless of the drug’s cost. Though Medicaid uses preferred drug lists to incentivize manufacturers to offer rebates, Medicaid must cover FDA-approved drugs as long as manufacturers participate in the rebate program.
At the same time, CMS’ proposed NCD requirements for Aduhelm, in particular the clinical trial requirements, are imperfect solutions that place the agency at odds with the FDA. Thus, industry experts are concerned that the new precedent set by CMS’ Aduhelm decision could unnecessarily complicate approval decisions for future therapies.
Several reforms would reduce U.S. drug costs while ensuring the pharmaceutical industry continues to innovate:
- Reform the FDA’s Accelerated Approval Program. The FDA implemented the Accelerated Approval Program (AAP) to speed up access to drugs that treat serious conditions for which there is an unmet medical need. But for many in the scientific community, Aduhelm’s approval through the AAP has undermined the program by casting doubt in the FDA’s adherence to evidence-based standards. Congress has introduced legislation to help address this problem. The Conditional Approval Act, introduced in 2019 by Sen. Mike Braun (Ind.), would establish a sunset provision on accelerated approval, allowing a drug maker time to show a drug’s efficacy, renewable each year up to five years. The sunset provision would automatically withdraw a drug’s approval should the company not produce confirmatory evidence of the drug’s efficacy, shielding the FDA from outside pressure to keep unproven drugs on the market.
- Replace CMS’ National Coverage Determination process with a closed drug formulary for both Medicare and Medicaid. A true drug market cannot exist where buyers are forced to buy. Unlike the Veterans Health Administration and private health insurance plans, Medicare and Medicaid do not have the ability to exclude drugs from their formularies. Even Medicare Part D, Medicare’s prescription drug benefit and the most market-oriented of the public health plans, is unable to exclude drugs from certain protected classes like antidepressants and cancer drugs. While the NCD process may temporarily prevent payment of egregious examples of high-cost, low-value drugs like Aduhelm, it is rarely used; Medicare pays for numerous drugs that offer little additional benefit over previous treatments. And as previously noted, the NCD process does not affect coverage decisions in Medicaid. A simpler and more elegant solution is to eliminate the NCD in favor of a closed drug formulary across all Medicare and Medicaid drug plans. This reform would give Medicare and Medicaid the strongest tool to negotiate lower drug prices: the power to walk away.
- Reform payment rules for public payers. As we have argued previously, both Medicare and Medicaid employ reimbursement methods that incentivize higher prices and limit competition among drug manufacturers. In conjunction with a closed formulary, reforming each program’s payment rules would further harness the power of markets to deliver better value to patients. For example, Medicare Part B reimburses drugs based on the average sales price plus six percent (ASP+6) formula. The six-percent commission for physicians to administer such drugs results in a feedback loop: physicians are incentivized to prescribe the highest cost drug, which encourages drug companies to raise prices, and so on. Eliminating ASP+6 and replacing it with a flat fee, similar to dispensing fees in Medicaid, would eliminate physicians’ profit incentive while increasing price competition among manufacturers whose drugs have similar therapeutic effects. Another reform is leveraging aggressive drug price negotiation in Medicare Part C and D plans to deliver the Part B drug benefit. This reform could be further enhanced by automatically enrolling seniors in such plans. As for Medicaid, statutes limit the size of rebates for drug prices that have nonetheless increased substantially over time. The Fair Care Act of 2020 addresses this problem by raising the cap on the rebate, while also eliminating the cap completely when a manufacturer raises prices faster than inflation. Finally, Medicaid statute should be amended to permit states to directly negotiate discounts from manufacturers if such discounts would be larger than the statutory drug rebate formulas allow.
Private insurance companies routinely use closed drug formularies to limit or completely deny coverage for drugs like Aduhelm, because their prices do not justify their clinical benefits and risks. Extending the same tool to Medicare and Medicaid, in conjunction with drug payment reform, would make innovative drugs more affordable and accessible for those who truly need them, and help make these programs fiscally sustainable for generations to come.
