Anti‐amyloid treatments: Why we think they are worth it

Abstract

Years of experience watching our patients progressively decline and die from complications of Alzheimer's disease (AD) has strongly motivated us to provide newly approved anti‐amyloid treatments to appropriate patients. Following detailed and personalized discussions of the potential risks and benefits of these treatments with patients and their families, almost 300 patients at our clinic have chosen to receive lecanemab infusions. We have found the frequency and severity of complications, including amyloid‐related imaging abnormalities (ARIA), to be manageable and as expected based on clinical trials. While the longer‐term benefits of these treatments are not yet clear, our patients and their families are accepting of even a modest slowing of disease progression. We have experienced the complexities, burdens, costs, and major logistical challenges associated with the treatment of AD with anti‐amyloid treatments. However, we also understand that for some of our current patients with early symptomatic AD, anti‐amyloid treatments are their best option for fighting this devastating disease, and we find it worthwhile to provide these treatments to our patients.

Highlights

• Many of our former patients have died from complications of AD.

• Our clinic now has nearly 300 patients receiving anti‐amyloid treatments.

• We have found the complications of anti‐amyloid treatments to be manageable.

• Despite the challenges, we find anti‐amyloid treatments worthwhile.

1.

Each of us has diagnosed and treated patients with Alzheimer's disease (AD) for many years. We do our best to provide accurate diagnoses, address exacerbating issues, eliminate sedating medications, and treat sleep and mood disorders. We recommend a healthy diet, exercise, cognitive stimulation, and acetylcholinesterase inhibitors and other medications as appropriate. However, persons with symptomatic AD inevitably decline, and, despite our best efforts, we have watched many of our patients progressively worsen and eventually die from complications of AD.

Because we understand the terrible toll of AD, we were encouraged when lecanemab (Leqembi), an anti‐amyloid monoclonal antibody, received traditional approval from the US Food and Drug Administration (FDA) for the treatment of early symptomatic AD in July of 2023. ¹ The subsequent coverage decision by the Center for Medicare and Medicaid Services made this therapy potentially available to millions of Americans. Our memory clinic at Washington University has been a site for multiple clinical trials of anti‐amyloid treatments, which provided experience that we have translated into clinical care. Nearly 300 of our patients are currently receiving lecanemab infusions.

The major side effects of anti‐amyloid antibodies are amyloid‐related imaging abnormalities (ARIA) with hemorrhage or hemosiderin deposits (ARIA‐H) or edema (ARIA‐E). ² In the CLARITY‐AD trial, ¹ ARIA‐H was found in 17.3% of lecanemab‐treated and 9.0% of placebo‐treated participants, while ARIA‐E was observed in 12.6% of lecanemab‐treated and 1.7% of placebo‐treated participants. ¹ Importantly, ARIA is a radiographic finding and is asymptomatic in most cases: in the CLARITY‐AD trial, only 31 of 898 treated participants (3.5%) had any symptoms from ARIA. ¹ Rates of ARIA in our clinic have been similar to those reported by CLARITY‐AD with most ARIA being asymptomatic and radiographically mild. ³ Even in patients with symptoms, ARIA has typically resolved within 4 to 8 weeks of stopping treatment, and steps have been taken to reduce the risk of severe complications. ⁴ Overall, the frequency and severity of side effects from lecanemab have been as expected and have been manageable.

Given our many years of caring for patients with AD, the published data, and our clinical experience, we have been disheartened by reports of anti‐amyloid treatments that do not put risks into context. An article in Science featured brain MRI images of severe ARIA and described complications of anti‐amyloid treatments as “brain swelling and bleeding, which in clinical trials affected up to about one‐third of patients and ranged from asymptomatic to fatal.” ⁵ Descriptions that lump together completely different outcomes may lead to a misperception that severe complications are common rather than only occurring in ∼1% to 2% of patients receiving anti‐amyloid treatments. Describing ARIA‐H as “brain bleeding” is particularly problematic because this term evokes a macrohemorrhage rather than typical ARIA‐H findings, such as one or two ∼1 mm microhemorrhages. Some articles describe smaller brain volumes in patients treated with anti‐amyloid antibodies as a mysterious and malignant effect ⁶ without mentioning reassuring possible explanations such as decreased inflammation. ⁷ Other articles state that death is a significant risk and recount a handful of cases reported over several years. ⁸ , ⁹ , ¹⁰ Again, we think it is essential to provide this information in context – serious complications related to ARIA can and do occur, but they are rare.

There is also controversy about whether the magnitude of the effects of anti‐amyloid treatments are meaningful, despite lecanemab meeting prespecified primary and secondary endpoints for slowing clinical progression in the CLARITY‐AD trial. ¹ Over the 18‐month study, lecanemab slowed clinical progression by 27% on the primary endpoint, the Clinical Dementia Rating Sum of Boxes (CDR‐SB), which measures both cognitive and functional features typically affected by AD. Some point to the relatively small difference (worsening on the CDR‐SB by 1.21 for lecanemab‐treated vs 1.66 for placebo‐treated participants), but because patients decline slowly, it is not possible to observe a large difference over just 18 months, especially because full amyloid clearance by lecanemab often does not occur until 12 months. As physicians, we understand that early interventions often have lasting effects, and we are very interested in learning whether anti‐amyloid treatments may improve long‐term outcomes that are important to patients and their families, such as additional months with independent function or delayed placement in a nursing facility. ¹¹ While we await these data, some of our current patients accept a 27% slowing of clinical progression as meaningful and want to be treated with lecanemab.

Blanket assessments by some clinicians that anti‐amyloid treatments are not “worth it” for any patient and regulatory decisions including by the Australian Therapeutic Goods Administration that do not permit ordering of these treatments for any patient within a jurisdiction ¹² are inconsistent with the principles of evidence‐based medicine, personalized medicine, and patient autonomy. Clinical trials have already assessed the balance of risks and benefits of treatment and found that patients treated with lecanemab have better cognitive outcomes than those treated with a placebo. ¹ Clinicians can minimize adverse events by following appropriate use recommendations and selecting patients at lower risk based on APOE genotype and other factors. ¹³ If clinicians do not provide anti‐amyloid treatments, they should consider referring potentially eligible patients seeking these treatments to clinics that offer this option.

Although the adage of “first do no harm” may be cited as a rationale for not providing anti‐amyloid treatments, nearly all medical and surgical treatments are associated with risks, some of which may be life‐threatening. Immunosuppressive treatments for a wide variety of diseases including multiple sclerosis increase the risk of serious infections. Anticoagulants and antiplatelet agents are widely used for thromboembolic disorders but increase the risk of major bleeding. Chemotherapy and radiation for cancer often cause significant side effects and may damage organs. Clinicians use these treatments despite sometimes substantial risks and modest benefits because the untreated disease is expected to result in even worse outcomes. While the incorrect and pervasive belief that “senility” is normal in older individuals may reduce broad acceptance of aggressive treatment for AD, many patients understand that AD is a disease that causes progressive disability and death, and they are willing to accept risks to mitigate their disease course.

We advocate for a comprehensive and personalized discussion of the risks and benefits of anti‐amyloid treatments for eligible patients and shared decision‐making on whether to initiate therapy. For example, it is important to describe the range of ARIA severity to patients and explain the likelihood of these conditions based on their APOE genotype, rather than vaguely warning of “brain bleeding and swelling,” which can be frightening and misleading. To make a well‐informed decision, patients and clinicians must have an accurate and personalized assessment of the risks and benefits of treatment.

Concerns about whether anti‐amyloid treatments are “worth it” to patients should also be separated from the financial and staffing issues that affect clinics. Providing biomarker testing and anti‐amyloid treatments is logistically complex and challenging, which may affect the interest and enthusiasm of clinicians and staff for offering these treatments. At our center, a dedicated team is needed to monitor patients, work with insurance, schedule infusions and imaging, and answer questions. We understand that while there is a major upfront cost and burden in developing treatment clinics, the era of AD treatment has arrived, and we have a responsibility to provide appropriate patients with these new treatments.

Healthcare systems may also have non‐medical considerations that impact access to treatments. Anti‐amyloid treatments are expensive and may not be cost‐effective for healthcare systems in the short term, ¹⁴ although they could potentially be cost‐effective longer term if they reduce caregiving needs or delay placement in nursing facilities. Somewhat conflicting decisions regarding lecanemab in the United Kingdom (UK) reflect the complexities of balancing the interests of individual patients and healthcare systems. After an assessment of the risks and benefits, the UK Medicines and Healthcare Products Regulatory Agency approved lecanemab for early symptomatic AD, although it specified that use was contraindicated in patients who were APOE homozygotes or taking anticoagulation. ¹⁵ However, then the government‐funded UK National Health Service (NHS) decided that it would not provide lecanemab until its cost‐effectiveness could be demonstrated. ¹⁶ Although this outcome is undoubtedly distressing for UK patients who cannot afford anti‐amyloid treatments, the UK NHS was at least transparent that their decision not to offer lecanemab was based primarily on financial calculations.

Notably, a second anti‐amyloid monoclonal antibody, donanemab (Kisunla), received traditional FDA approval in July of 2024 after demonstrating that it slowed clinical progression, and it is now clinically available, ¹⁷ further confirming that the era of AD treatment has arrived. Although providing anti‐amyloid treatments requires considerable effort by clinicians and the support of governments and payers, we have found that it is possible. While we strive to develop even better therapies, we have concluded that for some of our current patients, anti‐amyloid treatments are their best option in fighting a devastating and often deadly disease. Therefore, we find it worthwhile to provide anti‐amyloid treatments to our patients.

CONFLICT OF INTEREST STATEMENT

S.E.S. has served on scientific advisory boards on biomarker testing and education for Eisai and Novo Nordisk and has received speaking fees for presentations on biomarker testing from Eisai, Eli Lilly, and Novo Nordisk. E.S.M.’s laboratory previously received research funding from Eisai for animal studies related to sleep. J.C.M. has no disclosures. Author disclosures are available in the Supporting Information.

Supporting information

Supporting Information

Schindler SE, Musiek ES, Morris JC. Anti‐amyloid treatments: Why we think they are worth it. Alzheimer's Dement. 2025;11:e70055. 10.1002/trc2.70055