Medicare ‘Coverage With Evidence Development’ For Aducanumab? How Might It Work?
In the wake of the Food and Drug Administration’s (FDA’s) controversial decision to approve aducanumab for Alzheimer’s disease, how will, and how should, Medicare respond? We propose a coverage with evidence development (CED) option for the Centers for Medicare and Medicaid Services (CMS), and we examine how the agency’s past decisions on high-price technology may offer a guide. While others have written generally about Medicare’s options for aducanumab, including possible action from CMS’s Center for Medicare and Medicaid Innovation, we elaborate on the design of CED studies, drawing on lessons from historical CMS decisions.
In judging whether and how to cover aducanumab, CMS confronts several issues: One of the therapy’s Phase 3 studies, ENGAGE, failed to meet its primary endpoint of reduced cognitive decline from baseline on the Clinical Dementia Rating-Sum of Boxes; the identically designed EMERGE trial showed statistically significant but clinically modest results. The FDA granted aducanumab accelerated approval based on a surrogate endpoint, reduction in amyloid beta plaques (sticky protein build ups that form in the spaces between nerve cells), although the publicly available data linking this surrogate to improvements in cognition are limited. The drug’s manufacturer, Biogen, is required to perform a clinical study to confirm that the drug does in fact slow cognitive decline. However, the company has seven to nine years to submit results from this confirmatory trial.
Although the clinical trials focused on patients with mild cognitive impairment due to Alzheimer’s disease and mild Alzheimer’s disease dementia, the FDA granted the product a broad label, raising questions about which patients are most likely to benefit from this therapy. Aducanumab’s potential side effects, including brain swelling and bleeding, also elicit questions about its safety. Finally, there are questions about the product’s cost-effectiveness, given its announced price of $56,000, several-fold higher than the Institute for Clinical and Economic Review’s initial estimate of the upper bound of $8,300 for aducanumab to represent reasonable value for money. Adding to the drug’s costs are separate expenses associated with infusion, imaging tests to identify eligible patients, and costs related to the monitoring of side effects.
Because roughly 80 percent of patients eligible for aducanumab are Medicare beneficiaries, the drug will likely have a profound impact on program expenses and on patient’s out-of-pocket costs. Because of Medicare copayments, there are serious concerns about the financial impact of this therapy on beneficiaries and the potential for substantial inequities in patient access.
The FDA approval does not guarantee Medicare coverage. Rather, CMS has the authority to determine independently whether any treatment is “reasonable and necessary” for Medicare beneficiaries. Over the years, CMS has modified its approach for judging clinical evidence underlying diagnostics and treatments; the agency has required that interventions demonstrate that they improve patient health outcomes, not merely surrogate endpoints, and that they do so in patient populations that are similar to Medicare beneficiaries. Other than for preventive care, CMS does not consider the cost-effectiveness of new technologies when adjudicating coverage (although it tends to scrutinize clinical evidence more closely when budget impacts are large), and Medicare does not have the authority to negotiate drug prices.
Historically, CMS has automatically paid for most FDA-approved drugs for their labelled indications, even when it restricted coverage for off-label uses. But given the significant uncertainties regarding safety and effectiveness surrounding aducanumab and the potentially enormous budget implications, Medicare will carefully consider which patients should have access to the drug, and with what if any conditions. CMS faces two broad alternatives with critical decisions on the horizon.
1. Leave The Decision To Medicare Regional Contractors And Medicare Advantage Plans
CMS could defer decision making to its 12 regional Medicare administrative contractors (MACs), which in the absence of a national coverage determination (NCD), issue their own coverage decisions. Regional MACs generally cover Part B drugs (physician-administered drugs) with few restrictions. For Alzheimer’s disease patients enrolled in Medicare Advantage (MA) plans (roughly 2.6 million individuals, assuming a 10 percent disease prevalence rate), access would be determined by each individual’s health plan.
This path would avoid the need for CMS to issue a single national determination, but it has downsides. It might produce wide regional variation in coverage by MACs, as well as patient and clinician concerns that the availability of the drug is determined by where one happens to live. It could be costly given likely patient demand for treatment and perverse financial incentives for physicians to prescribe aducanumab. (Since aducanumab is a Part B drug, Medicare pays prescribing physicians 6 percent of the drug’s average sales price, or potentially as much as $3,360 per annual prescription.) Moreover, MACs and MA plans have even less political leverage than CMS’s central office to conclude that evidence deemed sufficient for FDA approval is inadequate for Medicare coverage.
2. Issue A National Coverage Determination
The shortcomings of acceding to MACs mean that CMS will likely consider an NCD. The Medicare program generally reserves this approach for interventions considered particularly controversial or expected to have a significant financial or health impact. NCDs are binding across traditional Medicare and MA plans and thus would standardize beneficiary access to aducanumab. NCDs also provide CMS a powerful tool with which to fine-tune coverage, thus not only reducing unwarranted variation in patient access but limiting therapy to patients most likely to benefit. CMS has turned to NCDs intermittently over the years, focusing on big-ticket items and on preventive care, diagnostic imaging, and health education/behavioral therapy interventions. CMS has not generally employed NCDs for drugs, although there are some exceptions.
If it does decide to pursue an NCD, CMS will closely scrutinize aducanumab’s clinical evidence and issue one of three verdicts.
Decide Not To Cover
In theory, CMS could judge that the evidence is insufficient to show that reducing amyloid plaque positively affects Medicare beneficiaries’ health outcomes, such as cognition, quality of life, or functional activities. In other words, CMS could rule that despite FDA approval, the evidence available for aducanumab does not meet Medicare’s “reasonable and necessary” standard. Given the fanfare surrounding the drug’s approval and the large, pent up demand for Alzheimer’s therapies, this path seems unlikely and indeed would be unprecedented. While NCDs addressing medical devices have resulted in non-coverage decisions in roughly 20 percent of cases, NCDs addressing FDA-approved drugs have always resulted in at least some coverage.
Cover For Selected Patient Subgroups
Another option would be for CMS to limit coverage to beneficiaries for whom the evidence is most robust. The agency has turned to this option on occasion for medical devices or procedures. For example, in their NCD for bariatric surgery for the treatment of morbid obesity (which typically involves placement of a gastric band), CMS limited coverage to patients who meet three criteria: BMI ≥ 35 kg/m2; an obesity-related comorbidity; and failed medical treatment. CMS has issued few NCDs for drug therapies, and those have generally provided national coverage according to the FDA label. However, given uncertainties around aducanumab’s clinical data, CMS could use an NCD to restrict coverage for aducanumab, perhaps limiting it to patients who meet the inclusion criteria of the therapy’s Phase 3 studies.
Doing so would lessen aducanumab’s budget impact, although the costs to the program would likely remain high. Of greatest concern, an NCD would also leave unanswered questions about how well the therapy worked while the world awaits Biogen’s required new study on the question.
Coverage With Evidence Development
All of this leaves CED as an attractive avenue. Since 2003, CMS has used this pathway in a few cases for interventions ranging from amyloid positron emission tomography for clinical evaluation of Alzheimer’s disease to implantable cardioverter defibrillators. CED provides’ access to beneficiaries enrolled in a CMS-approved clinical study or a patient registry that accumulates data about an intervention’s real-world effectiveness and safety. The pathway thus provides a compromise, permitting a degree of access while research continues as a treatment is adopted in clinical practice.
A CED approach would provide the medical community, patients, caregivers, and payers with additional information long before the FDA’s required post-approval studies are completed. It would also ensure that data on every patient treated would add to the knowledge base about how aducanumab impacts patient outcomes such as cognition, function, and quality of life. A CED would also provide an opportunity to study aducanumab in a more diverse and representative patient population than the one enrolled in Biogen’s Phase 3 studies and set an evidence standard for future amyloid-targeted agents.
CED policies have their own challenges, for example, they rarely involve prospectively randomizing patients and can involve complex data collection requirements and be costly to implement. (Medicare would presumably pay the treatment costs but not research expenses.) But given the uncertainty and cost impact of aducanumab, CED offers an appealing option. Ideally, a large, prospective national registry of all Medicare beneficiaries treated with aducanumab would be established, preferably through an independent, nonprofit organization—such as a patient advocacy organization. The design of the study should be informed by the broad participation of key stakeholders, including patients, provider groups, and drug companies (including those with existing products and those with products under development). In addition, active collaboration between CMS and the FDA would be critical to ensure that the evidence generated would complement the evidence generated by the confirmatory trials required by the FDA and would help address the future policy needs of both agencies.
Medicare’s past experience with CED offers important lessons. For instance, the CED for transcatheter aortic valve replacement for the treatment of symptomatic aortic valve stenosis collected such detailed information that the registry required dedicated staff and substantial expenses, raising challenges for efficient data collection. Any CED for aducanumab should focus on a minimum set of essential data elements needed to answer the key questions, as this increases the chances that the critical data will be consistently acquired. It would be critical to leverage the capabilities of the health information technology ecosystem and the rapidly expanding sources of high-quality, real-world data, including information from claims, electronic health records, and laboratory and pharmacy information, as well as smartphones, wearables, and other sources, to take maximum advantage of routinely collected data. This would allow for the fastest and least burdensome approach to generating information to help decision makers. Once the registry is established, it would also be possible to conduct pragmatic clinical trials on the registry platform, should it be determined that such a design is essential to providing actionable evidence.
A key unresolved issue is how amyloid positron emission tomography (PET) imaging—which is important for aducanumab treatment (and all patients enrolled in the aducanumab trials had presence of amyloid demonstrated by PET)—would be covered and reimbursed. CMS issued an NCD in 2013 that provided coverage of amyloid PET imaging through a CED policy, and it is unclear whether CMS would need to create a separate CED study specifically for the use of PET imaging in the context of aducanumab.
A CED approach could also help commercial payers and Medicaid. Although those payers need to make their own coverage determinations for aducanumab (for example, Biogen has announced a value-based contract with Cigna), they will undoubtedly watch Medicare’s next move closely and can learn from Medicare’s data collection efforts.
Aducanumab’s approval has thrust Medicare into the unenviable position of determining appropriate access in the face of highly uncertain evidence and huge potential costs. Given questions regarding aducanumab’s efficacy, side effects, and budget impact, the Medicare program and its beneficiaries would be best served if CMS pursued an NCD for aducanumab. It would be wise for the agency to consider CED to collect data for all treated patients in a new clinical study or a patient registry. While this approach has its own challenges, it would help address concerns surrounding aducanumab by standardizing Medicare beneficiaries’ access to the drug while simultaneously collecting much needed real-world data on safety and effectiveness.
Should Medicare decide not to deploy CED, it will take many years more than necessary to generate the additional evidence needed to confirm whether patients with Alzheimer’s disease are helped by this therapy.
The authors are members of the Center for the Evaluation of Value and Risk in Health at the Institute for Clinical Research and Health Policy Studies at Tufts Medical Center. The center receives funding from government, private foundation, and pharmaceutical industry sources. Drs. Chambers, Lin, and Neumann have consulted with pharmaceutical companies including Biogen and others on issues related to health economics and outcomes research. Dr. Neumann served as an external reviewer on ICER’s recent evaluation of aducanumab. Dr. Tunis has consulted on clinical development and reimbursement strategy with several life sciences companies, including Biogen; he has served as a special government employee at the FDA but was not directly involved in any drug/biologic discussions, including aducanumab.