Limiting Coverage Based On Efficacy And Safety: A Path Forward For Medicare Regarding The Alzheimer’s Treatment Aducanumab
By Sean Dickson, Amy Killelea, Richard G. Frank
Today, the Food & Drug Administration (FDA) will announce a decision on whether to approve the biologic aducanumab, intended to alleviate symptoms in patients with Alzheimer’s Disease. Currently, more than 6 million Americans over the age of 65 are living with Alzheimer’s Disease. As the US population continues to age, the number of seniors with Alzheimer’s Disease is only expected to grow. An estimated 1 in 9 Americans over 65 have Alzheimer’s, yet the vexing fact is that there is no effective treatment for the condition.
Aducanumab, the much-anticipated biologic made by Biogen and Eisai, Co. Ltd, was supposed to be a game changer for Alzheimer’s treatment. Instead, it has become an ethical watershed moment for both the FDA and the Centers for Medicare and Medicaid Services (CMS). If the FDA approves aducanumab in light of significant efficacy and safety concerns and in the face of a resounding negative vote by an FDA Advisory Committee last November, all eyes will turn to CMS to launch a Medicare National Coverage Determination (NCD) process for aducanumab, including an independent assessment and evidentiary review. NCDs are the most powerful coverage tool that Medicare has and have generally been reserved for Medicare services that are costly or may be subject to variable local coverage decisions, or where there is a divide in the scientific community regarding efficacy among Medicare beneficiaries.
Aducanumab meets all three NCD conditions: 1) it is estimated to cost as much as $70,000 per year, but based on the divide of the drug’s clinical effectiveness, it would need to be priced between $2,500-$8,300 a year to be cost effective; 2) similarly expensive and controversial therapies have a history of differing local coverage decisions; and 3) the scientific community, as well as FDA’s own advisory board, have voiced serious concerns about coverage.
It will be up to CMS to assess potential limits on the use of aducanumab, ensuring access only for those patients who are most likely to see some clinical improvement. If it is now impossible to identify those patients, then CMS should deny coverage until those patients can safely be identified.
This post will provide a history of aducanumab development, highlight the safety concerns, then discuss the tools available to CMS to ensure that aducanumab is only used in patients where it is clinically appropriate, given the limited efficacy and serious health risks.
Background On Aducanumab
Aducanumab is a human monoclonal antibody and used for the treatment of Alzheimer’s Disease. It is an infusion therapy and works by binding to aggregated forms of β-amyloid and β-amyloid plaques in brain tissue. Developing therapies that target amyloid plaques has been a driving theory in Alzheimer’s research, but one that has produced limited clinical trial success to date and growing divides within the scientific community.
Similarly, the development of aducanumab has not been a straightforward success. Biogen and Eisai began Phase 3 clinical trials—EMERGE and ENGAGE—in 2015, enrolling 1,650 patients in each trial. Each trial had an identical design: Participants were randomly assigned to a placebo, a low dose of aducanumab, or a high dose, each to be given intravenously every 4 weeks over 78 weeks.
While results showed that aducanumab significantly reduced the presence of amyloid plaques, the effect on cognitive functioning was less clear. A futility analysis was conducted in 2019, testing the probability of a statistically significant difference between the high and low dose arms of the studies. Both trials failed this futility analysis. As a result, the trials were halted in March of 2019.
However, in the three months between the futility analysis and the trials being halted, a significant number of participants completed the trials, leading Biogen and the FDA to work together closely on an extensive post hoc analysis of the trials. As discussed below, even the creative ad hoc analysis (which has been the subject of much debate and criticism in the Alzheimer’s scientific community) yielded better but ultimately mixed results. Still, as a result of this analysis, in October 2019, Biogen announced it was applying for a Biologics License Application (BLA) for aducanumab with the FDA, claiming that another Phase 3 trial was unnecessary.
FDA Advisory Board Recommendation
In November 2020, the FDA’s Peripheral and Central Nervous System Drugs Advisory Committee convened to discuss the BLA for aducanumab. At that meeting, Biogen argued that the findings were far more positive if the three months of subsequent data following suspension of the trials were used, particularly in the EMERGE trial, which found 23 percent of patients in the high-dose group showed improvement compared with placebo. The ENGAGE trial study still showed futility even with the additional three months of data.
However, in their combined briefing book, Biogen and FDA staff presented a nuanced argument for interpreting the divergent results of the two studies. That argument relies on stepping away from the randomized design and examining individual cases that in turn are used to make judgements about the “best” comparisons. Those judgments led the FDA and Biogen to argue that one of the reasons that the ENGAGE trial did not demonstrate efficacy in the high-dose arm was the presence of a disproportionate number of “rapid progressors,” who were skewing the results. If those rapid progressors were removed, the data showed a much more favorable impact of aducanumab.
Members of the Advisory Committee and the FDA’s own statistician questioned the methodology behind removing rapid progressors, particularly since these individuals were not identified at the outset. It is also unclear how rapid progressors (for whom the treatment is arguably ineffective) would be sorted out in clinical settings and how a “goldilocks” cohort made up of those with symptomatic but not yet advanced Alzheimer’s would be identified.
The Advisory Committee gave a resounding “no” vote on whether the analysis across the ENGAGE, EMERGE, and a third Phase 2 study (PRIME) demonstrated effectiveness of aducanumab for the treatment of Alzheimer’s Disease—the vote total was: 0 yes, 10 no, and 1 uncertain. More importantly, serious safety concerns were found, namely the presence of brain swelling, particularly at higher doses, raising red flags about the relative efficacy, if any, of aducanumab in light of the risks to patients. Even with these safety concerns and lack of support from the advisory committee, most analysts expect that the FDA will approve aducanumab.
Members of the FDA Advisory Committee subsequently described their opinions on the trials in a JAMA viewpoint. While they believe that the sponsor of aducanumab deserves recognition for the attempt to address a disease that poses a significant disease burden, there is no strong evidence that aducanumab would be effective in lightening the burden of Alzheimer’s Disease.
Independent CMS Review: Coverage With Evidence Development
CMS and the FDA play different and potentially complementary roles in drug approval, access, and coverage. FDA approval focuses on determining “reasonable assurance of safety and effectiveness for the drug or biologic” whereas CMS must consider whether an item or service is “reasonable and necessary” for insurance coverage and payment purposes. FDA approval does not (and should not) translate into automatic CMS coverage; indeed, the fundamental questions for each agency are fairly distinct.
Unlike FDA approval, Medicare NCD decisions are not an all or nothing proposition. In addition to deciding to cover or not cover a drug or biologic, CMS can either limit coverage or delay a full coverage determination following collection of more evidence. This latter “Coverage with Evidence Development” or “CED” option allows for Medicare coverage of a service only for those participating in a CMS-approved clinical trial, study, or registry aimed at collecting more information to guide broader coverage approval decisions. This is an important, but underutilized CMS coverage tool that could be important in the context of aducanumab, particularly given the controversy surrounding the ad hoc analysis of the clinical trial data.
Lessons From The ‘CART-T’ Process
The recent Medicare NCD process for the blockbuster cancer therapies known as chimeric antigen receptor T-cell therapy, or “CAR-T,” shed some light on how CMS review could proceed for aducanumab. In 2017, Kymriah (made by Novartis) and Yescarta (made by Kite Pharma/Gilead) became the first CAR-T drugs approved by the FDA and were priced at $475,000 and $373,000 respectively for a one-time treatment. Both were approved with Risk Evaluation and Mitigation Strategies, a patient safety mechanism the FDA uses to monitor approved drugs and biologics. Specifically, FDA required special certification for health care facilities to recognize and mitigate the risks of cytokine release syndrome and neurological toxicities. The FDA also required companies to further evaluate the long-term safety of the two products through post-marketing observational studies.
CMS received a formal request for a NCD from UnitedHealthcare, which administers Medicare Advantage plans. The impetus for the request was a concern that, given the high cost of CAR-T therapies, unless there was an NCD there would likely be variation in local coverage decisions as well as Medicare Advantage decisions. The request cited the complexity of the therapy, the potential for severe side effects, and the high cost of the therapy as factors that necessitated an NCD.
CMS convened a Medicare Coverage Advisory Committee (MEDCAC), an expert panel tasked with reviewing evidence and advising CMS on coverage decisions. The MEDCAC panel was specifically asked to review evidence on how patient-reported outcomes are assessed in clinical trials and which patient-reported outcome tools were most beneficial. The panel also assessed clinical effectiveness evidence and public testimony.
At the heart of the CAR-T NCD analysis was whether CAR-T could be provided safely to the Medicare population given the dangerous side effects. CMS paused on this question in its draft decision memo, not outright denying coverage for CAR-T, but proposing coverage of CAR-T with a CED requirement. CMS asserted additional collection of evidence for CAR-T was needed to “deepen CMS’s understanding of how patients in Medicare respond to it, so the agency can ensure that it is paying for CAR T-cell therapy for cases in which the benefits outweigh the risks.” The draft NCD for CAR-T outlined conditions of participation for healthcare facilities and limited coverage to those who participated in a CMS-approved registry or clinical study, in which patients would be monitored for at least two years post-treatment.
Following vociferous opposition from industry groups, CMS ultimately removed the CED requirement in its final NCD decision for CAR-T. However, the process here is still instructive. CMS has the ability to tailor its inquiry to weak points in clinical trial data, particularly gaps in real world implications for Medicare beneficiaries. In aducanumab’s case, CMS could use the NCD process and MEDCAC to tailor its inquiry to open questions posed by clinical trial data gaps, including how the “goldilocks” patients would be identified for aducanumab; how potential side effects would be monitored and avoided; and what threshold of efficacy, and for what segment of the population, would meet the CMS test of “reasonable and necessary.”
CMS should preemptively generate an internal request for an NCD regarding aducanumab to best protect Medicare beneficiaries from receiving a therapy where the risks likely outweigh the benefits. Given the public attention to aducanumab and the proportion of Medicare beneficiaries who could be eligible for treatment, a clear, evidence-based policy would protect beneficiaries and taxpayers and avoid differential policies across the nation when aducanumab launches.
The internally generated NCD policy route would have several advantages. First, it would allow CMS to undergo a transparent and public process to review the evidence base and determine the benefits of aducanumab coverage for beneficiaries. This transparency would be especially important in light of the controversy surrounding FDA approval for aducanumab and differing views in the scientific community about how clinical trial data was interpreted. Second, the NCD process would allow for more targeted review of the evidence through a slightly different statutory lens than FDA approval and could reveal gaps in knowledge and areas where more research and data re needed. And third, an NCD would set a national standard for coverage, avoiding differing access and local coverage decisions across Medicare Administrative Contractors and Medicare Advantage products.
Patients with Alzheimer’s Disease and their families are clamoring for, and deserve, effective treatments. Yet we must not let this push for treatment lead to unnecessary prescribing of a drug that hardly works and overly harms. CMS has a clear path to limit aducanumab usage to only the most appropriate cases, and, should aducanumab be approved, it should not hesitate to use this authority.