FormalPara Key Points

Both academic and non-academic healthcare institutions face numerous clinical, technical, and financial logistical challenges when starting an anti-amyloid monoclonal antibody (AMA) program.

We provide a framework for the institution of AMAs within a healthcare system.

Our experience has shown that the process was facilitated by having an enterprise-wide project manager, performing a financial analysis prior to starting, and leveraging electronic-medical-record-based tools.

1 Introduction

The US Food and Drug Administration (FDA) approval of anti-amyloid monoclonal antibodies (AMAs) for early-stage Alzheimer’s disease (AD) opens a new chapter in the management of neurodegenerative disease, but its implementation presents numerous clinical, technical, and financial logistical challenges for both academic and non-academic medical institutions hoping to administer this drug to the early AD patient community.

Inauguration of amyloid-clearing immunotherapy requires a thorough restructuring of the approach to early Alzheimer’s disease, involving staffing, clinic operations, electronic medical record (EMR)-based protocols and development of new clinical procedures. Whereas the “Appropriate Use” paper from Cummings et al. [1] serves as a valuable resource for effectively and safely treating the most appropriate patients, minimal resources are available to guide programmatic development for AMA readiness from the institutional level.

In this report, we describe our implementation approach and experience at an academic center with a regional referral clinic. This report will be most useful for clinicians working within academic or integrated healthcare systems that require the coordination of services from various systems stakeholders, including primary and specialty care, infusion centers, radiology, and pharmacy. It should be noted that we are presenting a preliminary report of AMA implementation and expect that, with greater experience and data, treatment strategies will continue to evolve. This report is organized into three parts: (1) feasibility assessment, (2) operations and going live, and (3) monitoring assessment. Furthermore, we are including content pertaining to the Epic tools that were developed at the University of Washington (UW) during the implementation process.

2 Setting

UW Medicine Memory and Brain Wellness Center (MBWC) is an academic multidisciplinary clinic based at Harborview Medical Center (HMC) in Seattle, WA. It evaluates and treats patients in whom AD is in the differential diagnosis. The provider staff consists of four behavioral neurologists, one neuropsychiatrist, two geriatric psychiatrists, two geriatricians, and two advanced practice providers. The center offers neuropsychological assessments, social work consultation, and a variety of services directed toward brain wellness (thememoryhub.org). The clinic has a volume of 1000 new patients and 4154 visits annually and serves as a tertiary center for the northwestern USA, receiving referrals from the Seattle metropolitan area, greater Washington, northern Montana, Wyoming, and Alaska.

3 Preliminary Work

Following the FDA approval of lecanemab, a broad-based project team was convened consisting of 50 members, representing the MBWC clinic, Radiology, Infusion Center, UW Medicine Strategy, Laboratory Services, Pharmacy, Information Technology Services (ITS), Finance, Revenue Cycle, and Communications. An enterprise-level project manager whose position was funded by the UW Strategy Division Project Management Office was recruited by the UW Medicine Neurosciences Institute (NSI) to organize the lecanemab start-up effort. The targeted timeframe for completing all tasks was 5 months (July–December 2023).

4 Phase 1: Feasibility Assessment

The feasibility assessment consisted of subgroup meetings to assess infusion, clinic, and radiological capacity. A separate group addressed the financial sustainability of administering AMAs.

4.1 Predicting Patient Volumes

The Alzheimer’s Facts and Figures 2023 data indicate that approximately 120,000 WA state residents had cognitive impairment in 2020 and that this number was expected to increase to 140,000 by 2025 [2]. Since it was estimated that 10% of this population had MCI, there were a predicted 12,000–14,000 persons with MCI residing in Washington state. MBWC providers reviewed their 50 most recent new referrals (350 in all) for eligibility under the CLARITY trial criteria, determining likelihood of amyloid positivity and receptivity to the lecanemab protocol. We estimated 60 potential new patients annually. We further noted that a Mayo Clinic study estimated that between 8% and 17% of their Aging Cohort would be considered candidates for lecanemab, when applying Appropriate Use criteria [3], converging with our estimate. We thus based our planning on a prediction of 40–80 patients initiating lecanemab within our practice at 1 year.

4.2 Collaboration with Radiology

Discussions were held with the UW Harborview Department of Radiology. Based on the Appropriate Use criteria for lecanemab, surveillance brain magnetic resonance imaging (MRI) is required before the 5th, 7th, 14th, and 26th infusions in addition to the baseline MRI. We require baseline MRIs within 1 year of ordering lecanemab treatment. Surveillance for amyloid-related imaging abnormalities (ARIA) requires dedicated, sensitive, and consistent sequences sensitive to microhemorrhage, so we require the baseline and surveillance MRIs to be 3T scans with susceptibility-weighted (SWI) sequences. In addition to the scheduled surveillance scans, it was estimated an additional 36 MRIs might be required to follow-up discovered ARIA, based on an ARIA rate of 21.5% [4], the predicted number of 60 subjects per year, and the likelihood of needing three additional scans for each case with amyloid-related imaging abnormalities (ARIA).

Conversations with the neuroradiology scheduling staff centered around the process of scheduling individual MRIs versus all four surveillance MRIs. The teams concluded that placing an order for all MRIs through a Smartset order was the most efficient approach and facilitated the flexibility to coordinate surveillance scans to immediately follow infusions when appropriate.

4.3 Collaboration with Infusion Center

The UW Infusion center determined it would be capable of absorbing the increased volume of infusion referrals. Discussions centered instead around coordinating medical decision-making related to the surveillance MRI results with infusion center operations, i.e., to ensure infusions are held until the MRI is cleared by the provider team. A solution was developed leveraging the electronic medical record (EMR) and is discussed below under phase 2.

Safety-related conversations with the infusion center further included the nursing staff’s lack of familiarity with treating patients with monoclonal anti-amyloid antibodies. Given that 26% of the CLARITY-AD patients [5] experienced infusion reactions with lecanemab, we created a protocol that was previously used in lecanemab clinical trials that graded the severity of the infusion reactions from 1 (mild) to 4 (severe). Depending on the severity of the reaction, treatment recommendations included treating with anti-inflammatories and antihistamines in addition to placing the infusion on hold. A set of questions were developed for the infusion nursing staff to screen for safety issues as well as the possibility of interval development of ARIA (online resource #1). We decided against premedicating with Benadryl due to the drug’s detrimental effects in the elderly, but some medical centers have used cetirizine prophylactically with Tylenol.

4.4 Collaboration with Pharmacy

Pharmacy leadership was engaged following the approval of lecanemab by the FDA on 6 July 2023. Our Pharmacy and Therapeutics Committee recommended adding lecanemab to the formulary, restricting its use to the outpatient setting in early-stage AD patients with insurance coverage and proven amyloid pathology. All prescribers were required to be clinicians within the UW-MBWC, and all lecanemab infusions were to be administered at HMC to ensure safety and quality control. Lecanemab prescribing was limited to dementia subspecialists due to the specialized expertise required to safely administer the medication, the ability of UW-MBWC clinicians to provide cross-coverage for each other, the goal of confining treatment solely to a targeted early-stage AD population based on specific inclusion/exclusion criteria, and the expectation that best practice would be rapidly evolving.

4.5 Biomarker Assessment

Due to the initial uncertain insurance coverage for amyloid positron emission tomography (PET) imaging, the AMA committee decided to assume cerebrospinal fluid (CSF) biomarkers would be utilized initially for the purpose of confirming Alzheimer’s disease pathophysiology in patients with qualifying clinical syndromes. The UW neurology department has an established lumbar puncture clinic. All samples are sent to a Clinical Laboratory Improvement Amendments (CLIA)-certified lab for testing (Roche Elecsys platform). After discussions with the medical director of the lumbar puncture clinic, two additional advanced practice providers were hired to facilitate timely scheduling of lumbar punctures within 2 weeks of orders.

Following the decision of the Centers for Medicare & Medicaid Services (CMS) to roll back the national coverage decision on amyloid PET scan, to a permit clinical reimbursement at the decision of the Medicare Administrative Contractors (MACs), we also explored a model of obtaining baseline amyloid PET scans with our radiology group and subsequently started using amyloid PET imaging as an alternative to CSF biomarkers. Notably, amyloid PET played an important role in the phase 3 trial of donanemab (TRAILBLAZER), and should donanemab gain full FDA approval, best practice may evolve quickly.

4.6 Financial Analysis

In phase 1, a financial analyst working with the team created a pro forma summary estimating the costs and revenue associated with integrating AMAs into clinical practice. The pro forma predicted downstream revenue related to the intravenous (IV) infusions when comparing the number of expected patients relative to the expense of offering the treatment. Calculations were performed adjusted for 40 and 60 patients beginning treatment per year. The amount of revenue was directly proportional to the number of patients infused.

Sources of revenue included (1) new patient visits (2 visits for every 1 new patient seen to meet criteria for treatment), (2) lumbar puncture testing (2 visits for every 1 new patient treated seen to identify patients positive for AD biomarkers), (3) laboratory visits, (4) MRI evaluations (estimated as a ratio of 4 scans for every 1 new patient seen to assess treatment progress plus 36 additional MRI scans for ARIA per 60 new patients), and (5) AMA infusions (volume estimated as a ratio of 36 visits for every 1 new patient seen, with the average patient having 16 treatments in the first year and 20 visits in the second year.

Calculated expenses included direct expenses for a medical assistant and medical staff expenses. Even when using the more conservative estimate of 40 patients treated per year, the rate of revenue exceeded the expense of offering lecanemab treatment. The reimbursement rate, which was estimated using the payor mix of the initial target patient population, ranged from 30.21 to 31.84 on the basis of whether 40 versus 60 patients were treated annually with IV lecanemab infusions.

5 Phase 2: Operations and Going Live

5.1 Clinical Practice

Patient Pathway: The evaluation of cognitive impairment has consisted of routine gathering of patient/care partners history, cognitive screening, laboratory studies for reversible causes of dementia (B12, thyroid stimulating hormone, and rapid plasma reagin in select populations), and neuroimaging according to American Academy of Neurology (AAN) guidelines [6]. Depending on the healthcare institution’s resources, neuropsychological testing has also been an important diagnostic component of the dementia workup. The integration of AMAs into this clinical practice additionally requires (1) documentation of the presence of AD biomarkers in either the cerebrospinal fluid or on amyloid-based molecular PET imaging, (2) risk assessment for ARIA, and (3) evaluation of financial feasibility. Whereas the traditional evaluation for dementia has required a total of two visits, one for diagnostic purposes and another for diagnostic impression/treatment plan, the introduction of an AMA requires an additional one to two visits to review results from biomarker and APOE genotype studies.

Costs associated with lecanemab treatments and associated diagnostic tests are discussed during initial conversations about the potential of this drug treatment with families. Patients with Medicare coverage are informed about the potential for being responsible for 20% of the associated $26,500 in drug costs. Furthermore, those individuals with private medical insurance are informed about the possibility of coverage denial given that half of major private insurers have decided not to cover drug costs. Any patient unable to meet the financial requirements for either the co-pay or full drug cost are referred to a clinic nurse responsible for corresponding with Eisai financial support (https://www.leqembi.com/eisai-patient-support) to find drug access solutions.

5.1.1 Discussions of AMA Benefits and Risks

Our group discussed the role of obtaining informed consent for AMA administration, and ultimately decided that documenting in the EMR a discussion with the patient and family about benefits and risks was sufficient. In neurology, there are a few precedents for using informed consent, such as the use of thrombolytics in acute stroke; it was felt that the administration of an AMA would not incur the same degree of risk, and so no consent form was used.

Clinicians begin with a discussion about the significance of CSF AD biomarkers and how they are indicative of a 60% risk of progression in 3 years or, in the case of mild AD, confirm the presence of the disease itself. Conversations then center around the 27% slowing of the decline in Clinical Dementia Rating (CDR) relative to placebo, based on the CLARITY AD trial, as well as the fact that the drug successfully results in amyloid clearance [1]. This information is presented in a more patient-friendly manner by explaining that this slowing translated into an additional 4−5 months over the 18-month period [7]. Individuals are informed that AMAs do not necessarily lead to symptomatic improvement. All patients are counseled about the 21.5% risk of ARIA with lecanemab and the fact that fewer than 3% of ARIA cases are symptomatic. Patients are informed that ARIA is most likely to occur during the first 3 months of AMA treatment. We explain how steps were taken to prevent and assess risk prior to the first infusion, including obtaining a baseline brain MRI to ensure the absence of clinical features predisposing one to ARIA, excluding patients on anticoagulation, checking an APOE genotype in advance (cost is covered at our center), and drawing labs for coagulation factors. We chose to provide patients homozygous for APOE e4 the opportunity to start lecanemab, but also inform them that the risk for ARIA is increased 4−6 times compared with non-carriers (e.g., E3/E3) and that there is a 9% risk for symptomatic ARIA. In addition, the favorable effects of lecanemab on the CDR in this population may be reduced compared with other groups according to the CLARITY-AD subgroup analysis [5]. Those individuals who were either heterozygous for APOE4 or were non-carriers are educated about their relatively lower risk. Results of APOE testing and impact on disease/drug risk are disclosed by the treating clinician to the patient and family during a virtual or in-person visit. UW, as with many other institutions, lacks the resources to offer genetic counseling services to all families. Drug- and risk-related information is outlined on the UW lecanemab resource page: https://depts.washington.edu/mbwc/resources/lecanemab-leqembi-update.

Patients are followed up with every 3 months by their neurology clinician, at which time they are assessed for adverse effects, and surveillance imaging findings were shared with them.

Any patients with baseline risk factors for ARIA (e.g., having baseline microhemorrhages or a history of lobar hemorrhage) are excluded based on the inclusion/exclusion criteria (online resource #2).

5.1.2 Creation of an AMA Approval Board

Despite there being a predefined Appropriate Use criteria for lecanemab, the committee concluded that referrals meeting these criteria but with associated high risk for complications or those failing to meet Appropriate Use criteria could benefit from a formal adjudication process. The approval board consisted of eight dementia experts, a neuropsychologist, two radiologists, and a nuclear medicine radiologist. The approval board was scheduled to meet monthly and required at least two clinical dementia experts and one radiologist for guidance about whether or not to treat prospective infusion patients with complex presentations. After an initial few cycles of review that served to level set the clinicians and establish a review culture, any cases squarely within Appropriate Use parameters did not require the AMA board to formally review.

5.1.3 Neuroradiology Reporting

The UW Memory and Brain Wellness Center and neuroradiology agreed to tailor any surveillance MRI reports to address the presence and absence of ARIA-E/H as well as the severity of the findings. Any cases of ARIA are communicated urgently to the ordering provider by paging them.

5.1.4 Clinic Call Coverage

The transition of AD patient management from relatively well-tolerated oral medications such as donepezil and memantine to an IV-administered biologic where approximately a quarter of patients experience infusion reactions and a fifth experience a combination of ARIA-E/H symptoms represents a profound change in clinical practice for the dementia expert. A day call schedule was created where a provider of the day is appointed to address lecanemab-related queries within and outside of the organization. Any severe catastrophic adverse event related to lecanemab resulting in malignant cerebral edema or macrohemorrhage is referred to the UW stroke physician on call. Outside of working hours between 8 a.m. and 5 p.m., urgent patient issues related to lecanemab are addressed by the on-call neurology resident. All patients are given a nurse-staffed emergency line should they develop symptoms in the setting of lecanemab infusions.

5.1.5 ARIA Monitoring and Management

The monitoring strategy for ARIA was influenced by the CLARITY-AD protocol [8], the FDA drug label (https://www.leqembi.com/-/media/Files/Leqembi/Prescribing-Information.pdf?hash=77aa4a86-b786-457a-b894-01de37199024), and the Appropriate Use recommendations by Cummings et al [1]. As specified in these publications, brain MRIs are conducted at baseline and prior to the 5th, 7th, and 14th infusions and at 52 weeks (i.e., prior to the 26th infusion). Any additional infusions following the MRIs performed related to the 5th, 7th, 14th, and 26th infusions are placed on hold until the ordering clinician has reviewed MRI results and cleared the patient for further treatments with the HMC infusion center.

The approach to management of ARIA is discussed in Cummings et al [1] and is based on the radiological grading of ARIA severity as well as the presence of symptoms associated with the radiological findings. Any patients who developed radiological ARIA and associated symptoms required assessment in clinic within 24–48 h depending on the presentation.

5.2 Documents

For the purpose of standardizing lecanemab administration throughout the practice, the implementation group created a variety of documents to provide guidance to clinic support staff and clinicians with respect to operational workflow, treatment, and management of adverse drug effects.

5.2.1 AMA Protocol

A formalized protocol or standard operating procedure (SOP) was created to ensure consistent care quality and safety when administering the infusion. The protocol addressed the following topics: (1) development of inclusion/exclusion criteria for patients receiving lecanemab (online resource #2), (2) creation of an AMA approval board, (3) daytime and after-hours coverage for AMA-related queries, (4) treatment procedures for diagnostic workup and risk assessment, (5) monitoring and management of ARIA, (6) monitoring drug efficacy, and (7) stopping criteria. This document provides standardized criteria that could always be referenced throughout the monoclonal antibody infusion process.

5.2.2 Clinic Flowchart

Information related to clinic operations was consolidated to a Visio chart describing each the patient pathway that includes clinic visits, laboratory studies, neuroimaging, and infusion treatments for an 18-month treatment period (Fig. 1). The Visio chart ultimately served to standardize the patient experience for all clinic referrals and provide a visual summary of the lecanemab protocol. Due to lack of available clinical trial data or guidelines, we did not create any specific protocols for treatment beyond the infusion at 18 months.

Fig. 1
figure 1

Vizio flowchart for amyloid monoclonal antibody treatment and monitoring

5.2.3 Lecanemab Checklist for Providers

A checklist was created to ensure that all treating clinicians were reminded of the new and more critical steps in the lecanemab pathway (Figs. 2, 3).

Fig. 2
figure 2

Lecanemab checklist for memory and brain wellness provider(s)

Fig. 3
figure 3

Epic flowsheet example

5.2.4 ARIA Flowchart

Recommendations from Cummings et al [1] and the UW lecanemab protocol provided radiographic criteria for determining whether cerebral ARIA-H/E changes were mild, moderate, or severe (Fig. 4). The severity of radiological ARIA has practical implications for medical decision-making, as mild changes may not necessarily impact treatment, whereas moderate/severe changes may prompt suspension of infusions. Furthermore, any symptoms accompanying MRI changes may also have implications in terms of halting therapies.

Fig. 4
figure 4

ARIA management protocol

5.3 EMR-Based Tools

5.3.1 Clinic Flowsheet

The last task for the lecanemab committee was to create Epic-based tools to guide prescribing clinicians (Fig. 3). The working group created an Epic-based flowsheet checklist to guide clinicians through the patient pathway. This work required collaboration between a MBWC physician, a project manager, and members of the EMR (Epic) build/design team. The flowsheet was divided into three phases: There was the clinician-driven diagnostic workup section (phase # 1) that included the cognitive assessment, laboratory testing, imaging, and biomarkers to confirm the presence of AD pathology. There was the clinic support-staff driven lecanemab treatment planning (phase #2) that included information relating to the scheduling of brain MRI and clinic visits. The final portion, infusion therapy (phase #3), related to the completion and review of MRI findings and required participation from both the clinician and the infusion nurse (infusion nurse portion is highlighted red). All infusions would be placed on hold until the MRIs are read and cleared by the treating clinician. The clinician is responsible for denoting whether an MRI was completed, whether the hold could be lifted from the order, and whether the patient could proceed with treatment. This portion of the flowsheet was created to prevent patients with ARIA from receiving infusions, and serves as the primary indicator of treatment status between the MBWC and the infusion center. Clinicians complete the flowsheet and then send an inbox message to the clinic staff, which in turn would communicate with the infusion center. This flowsheet was created for the duration of 18 months, as described in the CLARITY AD trial [5].

5.3.2 Epic Smartset

The EMR-based Smartset was created (by A.F. and D.M.) to facilitate the workup for both dementia and lecanemab candidacy (Fig. 5). This EMR tool is inclusive of baseline labs and MRI to establish a diagnosis of dementia as well as AD-confirmatory biomarkers (CSF and amyloid PET imaging). Furthermore, the Smartset includes orders to ensure safety: APOE genetic screening, coagulation labs, and surveillance brain MRIs before the 5th, 7th, 14th, and 26th infusions.

Fig. 5
figure 5

Sample Smartset for amyloid monoclonal antibody evaluation

5.3.3 Epic Therapy Plan

To create a treatment order that generated lecanemab infusions every 2 weeks for a period of 18 months, an EMR specialist (M.M.) developed an Epic-based therapy plan specifically for lecanemab (online resource #3). The template that was used for the lecanemab infusions was based on what had been previously used for multiple sclerosis treatments. By having this arrangement, the clinician places a single order as opposed to orders every 2 weeks. This work required collaboration between a MBWC physician, a project manager, and members of the EMR build/design team. The process of creating a therapy plan resembled that used for IV infusions used to treat multiple sclerosis. Clinicians go through the process of ordering the infusion as documented (online resource #3). There is, further, a process that required the clinician to place the therapy plan on hold in the case of ARIA or other adverse drug effects.

5.4 Oral Dementia Treatments in the Setting of AMAs

Similar to the patient demographic of the CLARITY AD trial, a large portion of the UW-MBWC early-stage AD cohort was receiving treatment with cholinesterase inhibitors and/or memantine. Whether or not a patient was taking or had recently started on an oral dementia medication had no influence on the timing or dosing of the AMA. All patients diagnosed with early onset AD were offered cholinesterase inhibitors prior to starting AMAs. In an effort to avoid confusing drug-related side effects, patients were required to be stable on a cholinesterase inhibitor for 1 month. However, we do not see any clear contraindication to starting both agents simultaneously. As was customary in our practice, memantine is not presented as an option for patients in the early stages of AD.

5.5 Internal and External Education

5.5.1 Internal Education

In-service sessions were provided to the frontline and nursing staff to educate them regarding potential concerning symptoms in patients receiving monoclonal antibodies and how to appropriately triage these patients. Furthermore, the infusion nursing team were provided guidance regarding questions to gauge patient tolerance of lecanemab (online resource #1).

An educational session was also provided to the full AMA project team at its first meeting; it was focused on the nature of the therapeutic advance and the specific changes in approach to evaluation, diagnosis, and management that would need to be supported by the project.

All dementia clinicians received training from the Epic team related to utilization of EMR-based tools and Smartsets.

An in-service session was held with the neurology resident and stroke services about approaching symptoms related to ARIA. In addition, we discussed how thrombolytics were contraindicated in those patients receiving lecanemab due to the increased risk of intracranial hemorrhage. In healthcare organizations with emergency rooms lacking 24/7 in-person acute neurology coverage, it is recommended that similar educational sessions be held with the emergency and hospitalist attendings to avoid treatment of lecanemab patients with thrombolytics or anticoagulation.

5.5.2 Community Education

The MBWC clinicians partnered with a UW communications specialist to produce a webinar providing a broad overview of lecanemab, its mechanism of action, drug efficacy, and adverse side effects such as ARIA. The webinar (https://www.youtube.com/watch?v=FMDxn80fPXI&t=1s) was advertised on social media platforms, including Facebook and Twitter. It was also shared with the clinic support staff, including nursing, for training purposes.

The working group further created a list of questions commonly asked by patients related to anti-amyloid monoclonal antibody treatments and posted the document on the clinic website. This resource was used for the purpose of patient education (https://depts.washington.edu/mbwc/resources/lecanemab-leqembi-update).

6 Phase 3: Monitoring Assessment

6.1 Outcome Data

All patients receiving lecanemab received an assessment to address cognition [Montreal Cognitive Assessment (MOCA)], function [functional assessment questionnaire (FAQ)], mood [Geriatric Anxiety Disorder (GAD)-7; Patient Health (PHQ)-9], and behavioral symptoms [Neuropsychiatric Inventory Questionnaire (NPI-Q)] at baseline and every 6 months. These batteries were chosen on the basis of clinic familiarity with the assessments and the ability to sustainably administer them without disruption of clinic workflow.

6.2 Lecanemab Registries

For lecanemab treatment to be reimbursed by CMS, treatment centers must participate in an FDA-approved registry for the purpose of gathering additional evidence about the benefits and risks related to anti-amyloid monoclonal antibodies. CMS offers a registry (qualitynet.cms.gov/alzheimers-ced-registry) that requires providers to enter a National Provider Number (NPI) and a Medicare Beneficiary ID (MBI). This registry captures basic information related to demographics, diagnosis, AD biomarker results, MOCA scores, FAQ scores, use of anticoagulants, and the incidence of ARIA. Relative to other available registry options, the CMS data capture is relatively lean, providing minimal disruption to clinic workflow, but at the same time inclusive of only the essential datapoints related to efficacy and safety. Alzheimer’s Network for Treatment and Diagnostics (ALZ-NET), which is based on the National Stroke Registry model, is a national network sponsored by the Alzheimer’s Association designed to capture data that can be used to make conclusions about the safety and efficacy of AMAs. Such surveys in the future may help answer pending questions related to duration of AMA treatment, risk/benefit of AMAs in APOE4 homozygotes, and safety of AMAs in patients on anticoagulation.

6.3 Quality Assurance

Following the go-live start date, the AMA committee continued to meet biweekly with representatives from the MBWC, pharmacy, radiology, and the infusion center to address any challenges associated with the workflow for the initial five patients referred for lecanemab. Most issues that required modifications were related to further streamlining of EMR-based workflow.

6.4 AMA Knowledge Gaps

With the majority of clinical trials enrolling early AD cohorts that may not fully reflect “real life” clinic populations and limited to an 18-month period, physicians’ utilization AMA will need to exercise clinical judgement in situations where the risk and benefit is unknown. One common question that arises relates to whether patients should be treated with lecanemab beyond 18 months. Currently, there is a lack of guidance both from the CLARITY AD trial as well as the Appropriate Use paper from Cumming et al. [1] on this matter. Until the CLARITY AD open-label extension data are available, clinicians will need to evaluate the clinical, radiological, and financial status of each individual patient. Likewise, there may be clinical situations that fall into “a knowledge gap,” whether that be a genetic condition, an autoimmune disease, the use of immunomodulating treatments, etc., where minimal data exist describing the impact of these factors on the safety and efficacy of AMAs. Finally, the long-term clinical sequelae of ARIA-H has been poorly defined, and only with additional longitudinal studies will neurologists be able to inform patients about whether these microhemorrhages actually contribute to a future risk of intracranial hemorrhage as seen in amyloid angiopathy. This may be another situation where clinical judgement may vary from clinician to clinician. Whereas the Appropriate Use paper from Cummings et al. [1] states that lecanemab should be stopped with the development of “more than 10 microhemorrhages since treatment initiation,” some neurologists may choose a more conservative approach. As mentioned previously, registries such as ALZ-NET will be useful in providing objective data to answer these questions.

7 AMAs in Community Neurological Practices

The community neurologist lacking the healthcare system infrastructure will face unique additional hurdles when initiating AMAs in the AD population. More than likely, AMAs will be prescribed by general neurologists rather than behavioral neurologists, and consequently, continuing medical education will be necessary for generalists less familiar with managing dementia. Academic institutions fortunate enough to possess the subspecialty expertise and clinical infrastructure to offer AMAs to AD patients are encouraged to take on the responsibility of mentoring community neurologists in forums such as the Alzheimer’s Association Project ECHO (https://www.alz.org/professionals/professional-providers/the-alzheimers-and-dementia-care-echo). At our center, we have met virtually with smaller practices through the Outreach Program at the UW Neurosciences Institute. Other institutions where statewide access to neurological care is more limited have created both virtual and in-person training sessions for community neurologists. Such efforts will not only ensure that the more appropriate patients receive AMAs but also provide a buffer for academic systems that will inevitably reach their saturation point in meeting the growing demand for these new treatments.

Although a community neurologist in private practice has the ability to prescribe AMAs without the constraints of intra-institutional committees, they may experience difficulty with obtaining the biomarkers required to confirm the presence of AD. There are US states lacking access to amyloid PET scanners, and the ability to perform routine spinal taps for CSF panels may further be limited by the clinician’s practice. Currently, few solutions to this dilemma exist beyond outsourcing the CSF acquisition to local partners in interventional radiology. Recent data related to blood biomarkers, particularly plasma pTau217, suggest that the accuracy of this test is at least equivalent to that of CSF biomarkers [9]. pTau217, which is not FDA approved or covered by Medicare, is currently available at various commercial laboratories (https://www.mayocliniclabs.com/test-catalog/Overview/621635; https://precivityad.com/; https://www.certuitad.com/), and patients will incur an out-of-pocket expense that will range from $200–1200 depending on the company. There is evidence that medical factors such as elevated body mass index [10] and renal disease [11] may confound blood biomarker outcomes, making this option less ideal. However, recent advancement in AD blood-based biomarkers is expected to lead to accessible, earlier, and more affordable means of diagnosing this condition.

Coordination of infusions and surveillance imaging may further present challenges to the community neurologist within a private practice. Partnerships with local infusion centers and imaging center will be necessary. In addition, ongoing dialogue between the neurologist and the reading radiologist will be necessary to ensure that the latter is providing accurate and prompt interpretations of MRI studies with respect to ARIA.

As observed in our experience, significant coordination will be necessary to track patients’ progress through a system, address insurance coverage, ensure entry into a CMS-approved registry, and schedule follow-up visits for surveillance. These additional steps in care may strain support staff resources and decisions will need to be made about whether to invest in additional clinical support. While the Epic tools provided may be helpful in practices using Epic as an EMR, various electronic treatment tracking systems are available on a subscription basis (https://medtraker.com/home/).

8 Discussion

Our experience in developing an AMA IV infusion program for early-stage AD at the UW-MBWC resulted in multiple important lessons. Firstly, the establishment of this new treatment as a routine clinical intervention required ongoing partnerships across a variety of departments for a period of 5 months.

The UW-MBWC was able to justify a need for AMAs to engage organizational leaders, who then assigned an enterprise-wide project manager to the undertaking. The project manager was able to efficiently engage each of the directors of the relevant departments and foster multidisciplinary collaboration across neurology, the infusion center, pharmacy, neuroradiology, finance, and the EMR technical staff. Although clinician champions remain engaged through the start-up process, the project manager was a key factor enabling our organization to effectively transform a well-formulated plan into a functional treatment program. Larger healthcare institutions lacking this constituent may find the process of instituting such an AMA program to be challenging.

Another important lesson was the role of financial analysis in determining the sustainability of an AMA program. Significant debate has been held about whether the statistically significant CDR outcomes from the CLARITY AD trial translate into clinical significance for the patient [12]. As a result, multiple private insurance companies have deferred on covering the costs for lecanemab. Nonetheless, our financial leadership performed a pro forma analysis showing that an infusion program was indeed sustainable, and our dementia program was able to obtain buy-in from organizational leadership for this effort.

We anticipate that implementing additional agents and/or evolutions in practice (e.g., amyloid PET as a treatment marker) will be greatly facilitated by the practice framework that has been implemented. For instance, this framework can be applied to donanemab, should the drug be approved by the FDA, with additional modifications such as incorporation of baseline tau-PET as a predictor of response as well as a surveillance amyloid PET scan at 24 weeks to confirm sufficient plaque clearance to justify pausing additional treatments.

The final lesson from our experience emphasized the importance of leveraging the EMR in simplifying management of the patient receiving AMAs. By developing Epic tools such as Smartsets and treatment plans, we consolidated the necessary steps for diagnostic workup, longitudinal safety evaluation, and arranging an 18-month therapy plan into readily accessible and personalized electronic menus that minimally impacted clinic workflow. Furthermore, we created flowsheets that enabled clinicians to readily identify where a given patient was in the treatment process. Our perception is that a robust clinic pathway tracking system and templated EMR order set are important in ensuring patients are managed safety and efficiently in a large healthcare system. We recognize the challenge of creating new Epic tools and have included our versions in this report for the purpose of assisting other institutions, both academic and non-academic, interested in adopting or revising them.

9 Conclusion

The launch of an AMA infusion program from a healthcare-systems level requires a multidisciplinary collaborative approach involving leaders from behavioral neurology, infusion medicine, pharmacy, neuroradiology, electronic medical records, coding, and billing. We have provided a framework for initiating AMA treatment from the medical-institutional level and have shared the tools developed during our operational development process while also recognizing that the proposed framework may not be feasible for all medical centers. Nevertheless, the experience during the start-up process has provided valuable lessons as dementia care enters the new era of AD biological therapy.