ÌÇÐÄvlog

The US Food and Drug Administration’s June 2021 decision to approve aducanumab for treatment for Alzheimer dementia raised concerns that a drug with uncertain benefit and high cost could, in aggregate, threaten Medicare's solvency. In response to these concerns, Biogen recently announced a 50% annual drug price reduction from $56 000 to $28 200 per patient. Preliminary US spending estimates either used extrapolated Alzheimer dementia prevalence data from 2012 or did not explicitly quantify ancillary costs, such as additional diagnostic imaging to monitor the amyloid-associated imaging abnormalities (ARIAs) that occur in 41% of treated patients, and did not incorporate the recently announced price reduction.^1-3 We estimated upper bound and lower bound annualized Medicare costs for administering aducanumab to beneficiaries with the approved indications of mild cognitive impairment (MCI) or mild dementia, focusing on the degree to which associated ancillary health services affect spending.¹

For this cross-sectional study, we used validated cognitive measures from the 2016 Health and Retirement Study, a nationally representative survey of older adults, to identify Medicare Part B beneficiaries 65 years or older with MCI or mild dementia with either Medicare fee-for-service or Medicare Advantage coverage (Figure^4,5). Institutional review board approval was provided by the University of California, Los Angeles, and informed consent was waived because of the study's design as a secondary analysis of publicly available, deidentified data. We followed the Strengthening the Reporting of Observational Studies in Epidemiology () reporting guidelines for reporting cross-sectional studies.^4,5

The lower bound estimates of patients eligible for treatment with aducanumab assume prescribers would apply clinical trial inclusion criteria to adults with MCI or mild dementia with amyloid plaque on positron emission tomography imaging results. The upper bound estimate includes patients with MCI or mild dementia and plaque without age or comorbidity restrictions to reflect potential off-label prescribing.¹ Between 37% to 68% of patients with MCI or dementia have plaque according to population studies (we used 37% for the lower bound and 68% for the upper bound estimate).⁶

We quantified drug costs from patient weights and ancillary costs, such as additional magnetic resonance imaging scans, using amyloid-associated imaging abnormalities rates from clinical trials and US Food and Drug Administration recommendations (Table).^1,2 We multiplied annualized per-person costs by lower bound and upper bound population estimates of MCI or mild dementia prevalence. Medicare would pay 80%, and the remaining 20% coinsurance would be paid by beneficiaries, private supplemental plans, and/or state Medicaid programs. While cost sharing may vary for Medicare Advantage beneficiaries, the total costs will remain the same regardless of the cost-sharing rules. The eAppendix and eTables 1 to 3 in the Supplement provide further details on dementia identification, weight-based drug cost estimates, and cost analysis assumptions.

We analyzed data using SAS, version 9.4 (SAS Institute). We accounted for survey clustering and adjusted results by survey weights for the national representativeness and response rate.

We identified 8396 participants representing approximately 41.2 million (95% CI, 39.2-43.1 million) US adults 65 years or older with Medicare Part B coverage in 2016. Total annualized per-person weight-based drug costs equaled $27 759.36; ancillary costs equaled $6563.94. For the lower bound estimate, if 25% of the 1 066 670 (95% CI, 0.95-1.2 million) eligible patients with MCI or mild dementia and plaque with clinical trial age and comorbidity restrictions received treatment with aducanumab, Medicare would pay $7.0 billion (95% CI, $6.2-$7.8 billion) each year. For the upper bound estimate, if 25% of the expanded population of 5 715 983 (95% CI, 5.3-6.2 million) eligible patients with MCI or mild dementia and plaque received treatment with aducanumab, Medicare would pay $37.4 billion (95% CI, $34.4-$40.3 billion). Ancillary health services comprised 19.4% of total population spending estimates.

In this nationally representative analysis, we identified between 1.1 to 5.7 million Medicare beneficiaries who are potentially eligible to receive treatment with aducanumab. Ancillary health services comprised nearly 20% of total population spending estimates, suggesting that prior analyses underestimated the anticipated costs of aducanumab.³

This study has limitations. We used plaque rates from population studies rather than actual scans on Health and Retirement Study participants.⁶ Our validated approach of identifying dementia prevalence may have misclassified some cases.⁴ Surveys may be less reliable among participants with dementia, although participants’ proxies were frequently available.⁴ Furthermore, we did not account for potential future changes to drug acquisition costs or for societal costs, such as caregiver burden or health system capacity limitations.

The findings of this cross-sectional study suggest that the drug and ancillary costs of aducanumab could seriously strain Medicare’s budget, raising questions of how to maximize the value of public dollars, particularly for a drug with unknown benefit and known risk of patient harm.

Accepted for Publication: November 4, 2021.

Published: January 14, 2022. doi:10.1001/jamahealthforum.2021.4495

Open Access: This is an open access article distributed under the terms of the CC-BY License. © 2022 Mafi JN et al. JAMA Health Forum.

Correction: This article was corrected on February 18, 2022, to fix errors in the text and Figure.

Corresponding Author: John N. Mafi, MD, MPH, Division of General Medicine and Health Services Research, Department of Medicine, David Geffen School of Medicine at UCLA, 1100 Glendon Ave, #908, Los Angeles, CA 90024 (jmafi@mednet.ucla.edu).

Author Contributions: Dr Mafi (principal investigator) had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Concept and design: Mafi, Arbanas, Damberg, Sarkisian, Landon.

Acquisition, analysis, or interpretation of data: Mafi, Leng, Arbanas, Tseng.

Drafting of the manuscript: Mafi.

Critical revision of the manuscript for important intellectual content: All authors.

Statistical analysis: Mafi, Leng, Tseng.

Obtained funding: Mafi, Sarkisian.

Administrative, technical, or material support: Mafi, Arbanas.

Supervision: Mafi, Sarkisian, Landon.

Conflict of Interest Disclosures: Dr Mafi reported grants from the National Institute on Aging (NIA) during the conduct of the study, as well as nonfinancial support from Milliman MedInsight and grants from Arnold Ventures. Ms Arbanas reported grants from the NIA during the conduct of the study. Dr Sarkisian reported grants from the National Institutes of Health (NIH) during the conduct of the study. No other disclosures were reported.

Funding/Support: This work was supported by NIH/NIA award R01AG070017-01. Dr Mafi was also supported by a NIH/NIA Beeson Emerging Leaders in Aging Research Career Development Award (grant K76AG064392-01A1).

Role of the Funder/Sponsor: The funders had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

Additional Contributions: We thank Kenneth Langa, MD, PhD, University of Michigan, for his technical assistance in our methods to identify patients with mild cognitive impairment and mild, moderate, and severe dementia. He was compensated for his contributions.