The Indiana University Brain Health Program to deliver amyloid‐targeted therapy to Alzheimer's disease patients

Abstract

INTRODUCTION

The Indiana University Brain Health Program was developed to support safe implementation of amyloid‐targeting therapies (ATTs) for early Alzheimer's disease (AD).

METHODS

We established Neurology Brain Health Navigators (neuroBHNs) to extend our Brain Health Navigator model into specialty care. NeuroBHNs pre‐screen patients for ATT eligibility, coordinate biomarker and imaging evaluation, provide structured education, and guide treatment initiation and monitoring. The program includes a dedicated social worker and insurance pre‐authorization specialist to streamline infusion access. Blood pressure is monitored at each visit, and infusions are deferred if readings exceed 140/90 mmHg.

RESULTS

Through December 16, 2025, 243 patients initiated lecanemab therapy. Apolipoprotein E (APOE) ε4 homozygotes were not excluded; 74% of treated patients carried at least one Apoε4 allele. The ARIA rate was 13.2%, and 40% of patients experienced infusion‐related reactions.

CONCLUSIONS

A navigator‐centered model can efficiently deliver ATT while supporting structured education, screening, and safety monitoring, potentially contributing to favorable safety outcomes.

Highlights

• Indiana University established the Brain Health Program to receive referrals, screen patients, and deliver treatment with amyloid‐targeted therapies.

• Delivery of amyloid‐targeted therapies has followed appropriate use recommendations with the addition of diligent monitoring of hypertension and acute illnesses.

• Indiana University chose to consider apolipoprotein E (APOE) ε4 carriers for lecanemab treatment, both heterozygous and homozygous, and 75% of our treated patients are APOE ε4 carriers.

• Given our high numbers of APOE ε4 carriers, our amyloid‐related imaging abnormalities (ARIA) rate has been comparable published lecanemab clinical trial data, at 13.2% overall, but lower than expected given the high number of APOE ε4 carriers in our treated patient population.

1. BACKGROUND

The implementation and delivery of the new amyloid targeting therapies (ATTs) for Alzheimer's disease (AD) brought about great optimism but also significant challenges for a field that has not had to deliver such United States Food and Drug Administration (FDA) ‐approved infusions in the past. The delivery of frequent intravenous infusions, including monoclonal antibody treatments, is not a new concept and is used throughout medicine including rheumatology, gastroenterology, oncology, endocrinology, and even neuroimmunology. However, the ATTs require management of infusion related symptoms, close monitoring of neurologic symptoms that could correlate with amyloid‐related imaging abnormalities (ARIA), frequent magnetic resonance imaging (MRI) scans to observe for ARIA and monitor ARIA resolution, enrollment and data tracking in Centers for Medicare and Medicaid Services (CMS) approved databases, and eventual decision‐making about amyloid status and maintenance therapy. Even the management of a single patient on ATT requires a significant amount of infrastructure. Upon the traditional FDA approval of lecanemab on July 6, 2023 ¹ our work began to develop and implement a safe and effective means for managing administration of the drug. The March 2023 lecanemab appropriate use recommendations (AUR), ² published in response to the accelerated approval lecanamab received from the FDA in January 2023, provides a framework for understanding the aspects and resources required for effective drug management (Figure 1 and 2). These recommendations have also been adopted by the Alzheimer's Association as a part of their lecanemab toolkit. ³ Using this as a basis for creating infrastructure within our medical system, we identified several challenges and solutions to begin safe delivery of treatment.

FIGURE 1.

Indiana University Referral Checklist based on AUR inclusion/exclusion criteria and internal decisions on appropriate candidates for treatment, referrals from internal and external neurologists. AUR, appropriate use recommendation.

FIGURE 2.

Indiana University Referral Checklist based on AUR inclusion/exclusion criteria and internal decisions on appropriate candidates for treatment referrals from geriatrics and primary care providers. AUR, appropriate use recommendation.

As a healthcare system, we considered several different models for treatment delivery. Our focus was on minimizing time from diagnosis to treatment, knowing that earlier treatment likely has better benefits. ⁴ The simplest model would be to place ATTs on a formulary and allow providers to manage their own patients. Systematic “guardrails” could be used to create uniform requirements/“best‐practices”, and standardized institutional literature. We were concerned that this implementation may come with significant risk given the rapidly changing landscape of ATTs and the unknown risk with introducing them into commercial treatment. Another model that was considered was to have a few qualified providers designated to manage patients on ATTs, shedding their other clinical responsibilities. This model has been successfully implemented in ATT programs known to the authors using neurologists and/or advanced practice providers (APPs) with neurologist oversight. ⁵ , ⁶ , ⁷ We recognize that this approach enhances control over patients, adds uniformity to the selection and management, and could potentially reduce adverse outcomes. However, we wanted to pre‐emptively avoid bottlenecks in the referral of patients and unnecessary delays from diagnosis to first infusion within our system. The final approach that we considered and eventually implemented is a hybrid model where only a qualified prescribing team (deemed so by an internal neurology council) and an internal support team composed of registered nurses (neurology Brain Health Navigators; neuroBHN), pharmacy team members, and social workers, would manage patients on ATTs and other providers would “refer” to the treatment team for ATT management. This hub‐and‐spoke model was felt to be scalable and allow for more flexibility and team‐based decision‐making. This paper describes our innovative Brain Health Program (BHP) to manage patients on ATTs using a team‐based approach to optimize efficiency of cost, sustainability, and scalability.

2. METHODS

2.1. Program design and setting

The BHP was developed within a large academic health system to facilitate the safe and efficient delivery of ATTs for patients with early AD. The program was designed to streamline and standardize patient identification, eligibility determination, treatment initiation, infusion management, and longitudinal safety monitoring while minimizing delays related to subspecialty access.

2.2. Patient selection and eligibility determination

Appropriate identification of treatment candidates was considered foundational to program implementation. Given widespread public interest in ATTs, structured eligibility screening was required to ensure appropriate use. Estimates suggest that fewer than 20% of individuals diagnosed with AD meet eligibility criteria for currently approved ATTs, largely due to disease stage, biomarker requirements, and exclusionary comorbidities. ⁸ , ⁹

Eligibility criteria were derived from published AURs and were aligned with inclusion and exclusion criteria from phase 3 clinical trials of lecanemab and donanemab. ² , ¹⁰ Based on these criteria, two structured referral checklists were developed to assist referring providers in identifying potentially eligible patients. These checklists were designed to distill core eligibility requirements into a simplified format to reduce referral barriers and improve workflow efficiency.

Neurologists were requested to obtain confirmatory biomarker testing (e.g., amyloid positron emission tomography [PET] imaging or cerebrospinal fluid biomarkers) prior to referral (Figure 1). For non‐neurology providers (e.g., geriatrics, internal medicine, family medicine), referral required a positive or indeterminant blood‐based biomarker (BBM) prior to BHP evaluation (Figure 2). To facilitate this process in primary care settings, a primary care Brain Health Navigator (pcBHN) model was implemented to assist with workup and BBM ordering before referral. ¹¹ Primary care patients meeting preliminary checklist criteria were referred directly to the BHP rather than to general neurology or behavioral neurology clinics, thereby reducing delays in subspecialty access.

RESEARCH IN CONTEXT

1. Systematic review: Amyloid‐targeted therapies (ATTs) received United States Food and Drug Administration (FDA) approval in 2023. Careful screening and monitoring is required due to the risk and development of amyloid‐related imaging abnormalities (ARIA). Indiana University established the Brain Health Program to receive referrals, screen patients, and deliver ATTs.

2. Interpretation: Our Brain Health Program has been developed to be scalable and reproducible by other healthcare systems. We successfully treated 243 individuals with lecanemab from September 2023 through December 2025. Diligence around hypertension and acute illness was implemented to minimize risk of ARIA development. 75% of our patients are apolipoprotein E (APOE) ε4 carriers, and 12% are homozygotes. We have maintained an overall ARIA rate of 13.1% in our lecanemab population.

3. Future directions: We continue to receive referrals and treat patients with ATTs. Patients are evaluated every 6 months and will examine cognitive outcomes related to ATTs as we accumulate longitudinal data that has sufficient power for a real‐world analysis.

2.3. neuroBHN model

To ensure continuity and coordination of care, a neuroBHN role was developed. This role is based on other successful models of nurse navigators in facilitating patient care. ¹² , ¹³ The neuroBHN serves as the single point of contact for patients and care partners throughout eligibility confirmation, education, treatment initiation, and ongoing monitoring every 6 months.

NeuroBHNs are Registered Nurses (RNs) with specialized training in dementia care, cognitive assessment, and ATT protocols. Although not prescribers, they operate under structured nursing protocols developed to facilitate rapid clinical decision‐making and adverse event management. NeuroBHNs are trained to administer standardized cognitive assessments including the Montreal Cognitive Assessment (MoCA), ¹⁴ Clinical Dementia Rating (CDR), ¹⁵ , ¹⁶ and Mini‐Mental State Examination (MMSE). ¹⁷ After receiving the referral, the neuroBHN reviews eligibility criteria, including cognitive testing results, biomarker confirmation, and neuroimaging. Updated cognitive testing is performed if prior assessments are outdated. MRI studies, biomarker reports, and neuroradiology reports are reviewed by prescribing clinicians prior to treatment authorization. Importantly, each staff member administering the MoCA, CDR, or other necessary cognitive testing, has been trained and certified for each tool used. When considering real‐world implementation of treatment, we believe that certification and appropriate training is a reasonable way to ensure appropriate administration of cognitive tests.

2.4. Patient education and informed decision‐making

After eligibility confirmation, patients and care partners participate in a structured education visit conducted by the neuroBHN using teach‐back methodology. ¹⁸ These visits last approximately 60–90 min and may occur individually or in small groups (approximately five patient–care partner dyads) depending on neuroBHP and patient availability. We require that participants have a care partner present to ensure all of their questions and concerns are addressed and that they are aware of the cadence of infusions, MRIs, and, importantly, are aware of the symptoms of ARIA.

Education includes:

• Treatment mechanism of action.

• Expected clinical outcomes (slowing of decline rather than cognitive improvement).

• Infusion frequency and duration (i.e., the patient “journey”).

• Safety MRI requirements.

• Risk of ARIA.

• Potential infusion‐related reactions.

• Recognition and reporting of adverse events.

• Discontinuation criteria and patient ability to discontinue at any time.

Teach‐back methodology is used to assess comprehension. Patients are provided with written materials, infusion schedules, MRI schedules, blood pressure monitoring logs, and wallet cards indicating contraindication to thrombolytics (e.g., tPA, tenecteplase) and certain anticoagulants while on ATT are provided to both the patient and care partner. Emergency department (ED) evaluation is encouraged for stroke‐like symptoms, with emphasis on mechanical thrombectomy as a potential option when appropriate. Patients electing to proceed with therapy undergo final prescriber evaluation prior to initiation.

2.5. Behavioral neurology oversight

To offset the long lead time to access neurologists, ¹⁹ , ²⁰ including in our own system, behavioral neurology involvement is structured as a targeted evaluation following neuroBHN‐led qualification to minimize bottlenecks. After completion of required assessments (e.g., cognitive testing, MRI, amyloid PET, APOE genotyping), documentation is entered into the electronic health record (EHR), enabling a focused virtual or in‐person evaluation by an ATT‐approved behavioral neurologist using dedicated appointment slots reserved for ATT evaluations. Prescribers wishing to be a part of the Indiana University's (IU) BHP must be approved to prescribe through our program by the IU neurology counsel, and to understand the screening, monitoring, and safety measures, we implement. During these visits, the prescriber:

• Reviews eligibility documentation.

• Discusses clinical trial data and expected outcomes.

• Reviews ARIA incidence, including APOE‐stratified risk based on available clinical trial data.

• Provides final counseling and answers patient questions.

• Prescribes therapy if appropriate.

Patients with APOE ε4/ε4 genotype were not excluded but received additional risk counseling. Cases involving borderline cerebrovascular disease, uncontrolled vascular risk factors, or high‐risk comorbidities were reviewed with heightened scrutiny and through consensus amongst prescribers. Patients were instructed to defer infusions during acute illness, surgery, or inflammatory states and notify the neuroBHN if any signs or symptoms of such conditions developed. Weekly multidisciplinary meetings involving prescribers, neuroBHNs, a pharmacist, and program leadership are conducted to review complex cases and program metrics.

2.6. Imaging and safety monitoring

For those referred patients without a recent 3T MRI with appropriate sequences within 1 year, we order a baseline MRI scan, which includes the same imaging protocol we use for our safety monitoring. We require 3T MRI for all APOE ε4/ε4 candidates at baseline and for monitoring scans, and attempt to have 3T MRI for all participants but allow flexibility for non‐homozygotes who have accessibility concerns or implanted devices (pacemakers, implantable cardioverter‐defribrilators [ICDs], INSPIRE devices, etc). Safety monitoring MRI scans are performed at regular intervals according to the appropriate use guidelines (following the 4th, 6th, and 13th infusion (and following the 25th infusion for APOEε4 carriers) on the same physical scanner where possible. For all scans, we perform T1, T2, fluid‐attenuated inversion recovery (FLAIR), and susceptibility weighted imaging (SWI) sequences Siemens 3T scanners are available at large IU Health (IUH) facilities including The Neuroscience Center, Methodist Hospital, IUH North in Carmel, IN, IUH Fishers, IUH Bloomington, and IUH Fort Wayne. All 3T scanners are manufactured by Siemens and scans are read by a board‐certified neuroradiologists. MRI reports use templates derived from American Society of Neuroradiology [ASNR] guidance ²¹ ). Amyloid PET reports include both visual interpretation and centiloid quantification. If these disagree, we defer to the visual interpretation and discuss with the nuclear‐medicine radiology team.

In response to emerging data identifying hypertension as a risk factor for ARIA. ²² a strict blood pressure protocol was implemented. Blood pressure is measured upon infusion center arrival. If the patient's blood pressure is >140/90 mmHg, the measurement is repeated after 15 min of rest. The infusion is deferred and coordination with primary care is initiated for blood pressure optimization if the persistent elevation of the blood pressure continues upon repeat measurement deferral and coordination with primary care for blood pressure optimization. Patients are instructed to maintain home blood pressure logs at their education visit and we suggest they meet with their primary care team for more aggressive management if they are borderline.

2.7. Infusion management

ATT infusions were initially administered solely at the IU neuro‐oncology infusion center to leverage nursing expertise in neurologic adverse event management. Currently all participants receive their first four to six infusions at this location and then move to an infusion center within the IU system closer to their home, or use an outside infusion partner for areas without an IU location that is convenient. Pharmacy teams developed standardized order sets and infusion reaction protocols. Since Lecanemab dosing is weight‐based, patients are weighed prior to each infusion, and dose adjustments are made if body weight changes by ≥10% from baseline.

Patients are monitored following AUR post‐infusion guidance:

• 3 h after first infusion.

• 2 h after second and third infusions.

• 30 min for subsequent infusions.

Infusion reactions are treated per protocol (e.g., meperidine for rigors, ondansetron for nausea). Premedication protocols have evolved over time; acetaminophen and famotidine are now required for the first four infusions. Diphenhydramine is avoided when possible, due to sedation concerns. Patients with recurrent reactions received methylprednisolone premedication for four subsequent infusions and a trial off is then attempted. Missed infusions trigger coordinated rescheduling of both infusions and required safety MRIs. Patients re tracked in an encrypted database separate from the EHR. Weekly safety reports, including ARIA events, discontinuations, and new starts, are monitored by the team.

2.8. ED coordination and acute event management

Patients and their caregivers are provided with a wallet card that they are asked to present if they are admitted to the ED or visit an urgent care facility (Figure 3A). Institutional stroke protocols were modified to identify ATT use as a relative contraindication to thrombolytic therapy. EHR‐based alerts were implemented to notify providers whenever thrombolytics or anticoagulants were ordered for patients receiving ATTs. ED clinicians were encouraged to contact the on‐call BHP team for risk–benefit discussions. Educational resources regarding recognition and management of ARIA were distributed to ED providers. Communication pathways were established to facilitate rapid coordination in urgent clinical scenarios.

FIGURE 3.

(A) Wallet card. (B) Illustration summarizing the brain health program team and the patient workflow through the ATT screening process. ATT, amyloid targeting therapies.

2.9. Care management and billing infrastructure

Given the longitudinal and intensive nature of ATT monitoring, the program incorporated Principal Care Management (PCM) billing under CMS guidelines. ²³ NeuroBHNs provide qualifying services including education, diagnostic coordination, treatment monitoring, and community resource linkage. Services are billed under PCM codes (e.g., 99426, 99427) through the supervising prescribing neurologist. Documentation supporting billing requirements is maintained in the EHR.

2.10. Program adaptation and continuous oversight

Program leadership continuously monitors emerging evidence, safety signals, and regulatory updates. Modifications to protocols are implemented in real time after consensus by the clinical and leadership teams The structured neuroBHN‐centered model was designed to optimize behavioral neurologist time, reduce referral‐to‐treatment intervals, and provide scalable infrastructure for ATT delivery.

3. RESULTS

Consistent with the AUR, we treated patients within the age range suggested and who have no evidence of superficial siderosis, less than four pre‐existing microbleeds, and lack of severe white matter disease. Treated patients generally have cognitive evaluations and scores suggested in the AUR, with some treated patients having atypical presentations of AD (such as logopenic primary progressive aphasia, posterior cortical atrophy variant, etc.). Cognitive evaluations may be limited to a brief assessment, such as the MoCA, if the history taking and results are consistent with amnestic MCI or mild dementia, and biomarkers are consistent with AD. In cases where there is ambiguity in testing, co‐pathology is felt to be present, or in the opinion of the prescriber additional cognitive testing is needed, neuropsychological testing is pursued. We felt an obligation to offer a treatment choice to patients with homozygous APOE ε4 with appropriate counseling. We have had significant concern about APOE ε2/ ε4 carriers given the authors personal experience with individuals treated in various circumstances. With a lack of safety data on ATTs for patients with APOE ε2/ ε4, and after the concerns raised in the donanemab AUR, ¹⁰ we have chosen to not treat this population due to their propensity for extensive cerebrovascular involvement until there is more data in this population.

Our mean time from referral receipt to first infusion has been 106.4 days (± 57 days). We have found that our primary referral source has been our memory disorders clinic (73% of referrals), with 24% of referrals from general neurologists, and 2% from a geriatric dementia clinic. Our non–memory‐disorder referral sources have increased incrementally since the program began in 2023. We have also been able to increase our capacity for patients with workflow optimizations. We initiated treatment of 26 patients in 2023, 85 patients in 2024, and 132 patients in 2025. Over the current study period, we received 546 referrals into the BHP. Of these, 338 were determined to be eligible, and 243 choosing to receive lecanamab treatment.

Among the 243 patients treated with lecanamab, 32 (13.2%) developed ARIA (Table 1). Baseline demographic characteristics were similar between those with and without ARIA. Mean age did not differ significantly (72.0 vs. 74.1 years, p = 0.12), nor did sex distribution (59% vs. 54% female, p = 0.54), education level (p = 0.93), cognitive diagnosis (p = 0.09), or MoCA scores (21.5 vs. 21.4, p = 0.42) (Table 1). Race and ethnicity distributions were also comparable. Medication use, including antihypertensives, antidiabetic agents, antiplatelets, and low‐dose aspirin, did not differ significantly between groups. In contrast, APOE ε4 status was significantly associated with ARIA risk (p = 0.001) (Table 1).

TABLE 1.

Baseline characteristics of BHP patients.

Parameter	All patients (N = 243)	No ARIA (N = 211)	Any ARIA (N = 32)	p‐Value
Mean age (SD), y	73.8 (7.1)	74.1 (7.1)	72.0 (6.9)	0.12
Sex				0.54
Female	132 (54)	113 (54)	19 (59)
Male	111 (46)	98 (46)	13 (41)
Education level 1				0.93
≤12 y	39 (15)	34 (17)	5 (16)
13‐15 y	37 (14)	31 (16)	6 (19)
16 y	73 (29)	62 (32)	11 (34)
>16 y	81 (31)	71 (35)	10 (31)
Cognitive diagnosis				0.09
MCI	193 (79)	164 (78)	29 (91)
Mild dementia	50 (21)	47 (22)	3 (9)
MOCA score (SD) 2	21.4 (3.5)	21.4 (3.5)	21.5 (3.1)	0.42
Race and ethnicity
Asian	1 (0.4)	1 (0.5)	0 (0)	0.70
Black	4 (2)	4 (2)	0 (0)	0.43
Hispanic	5 (2)	5 (2)	0 (0)	0.38
White	236 (97)	204 (97)	32 (100)	0.30
APOE ε4 status				0.001 *
ε4 homozygote	30 (12)	20 (10)	10 (31)a
ε4 heterozygote	150 (62)	132 (63)	18 (56)b
ε2/ε4 genotype	1 (0.4)	1 (0.5)	0 (0)
ε3/ε4 genotype	149 (61)	131 (62)	18 (56)
ε4 non‐carrier	63 (26)	59 (28)	4 (13)b
ε2/ε3 genotype	3 (1)	3 (1)	0 (0)
ε3/ε3 genotype	60 (25)	56 (27)	4 (13)
Medication use
Antihypertensive	138 (57)	119 (56)	19 (59)	0.75
Anti‐diabetic	29 (12)	27 (13)	2 (6)	0.29
Antiplatelet	8 (9)	7 (3)	1 (3)	0.96
Aspirin low dose	57 (24)	50 (24)	7 (23)	0.88

Note: All numbers are listed as N (%) for categorical variables and mean (SD) for continuous variables.

¹ N = 230.

² N = 218. MOCA scores adjusted for educational level. Remaining cohort had lecanemab eligibility determined by MMSE.

^* p < 0.05 as per chi‐squared testing for categorical variables and t‐testing for continuous variables.

Abbreviations: APOE, apolipoprotein E; ARIA, amyloid‐related imaging abnormalities; BHN, Brain Health Navigator; MCI, mild cognitive impairment; MoCA, Montreal Cognitive Assessment; SD, standard deviation.

Among patients who developed ARIA, APOE ε4 status differed by ARIA subtype. In cases of isolated ARIA‐E (N = 10), most patients were ε4 heterozygotes (60%), followed by ε4 non‐carriers (30%) and ε4 homozygotes (10%) (Table 2). In isolated ARIA‐H without siderosis (N = 12), ε4 homozygotes comprised the largest proportion (50%), followed by ε4 heterozygotes (40%) and non‐carriers (10%) (Table 2). In contrast, among isolated ARIA‐H with siderosis, all cases occurred in ε4 heterozygotes. For patients who developed combined ARIA‐E and ARIA‐H (N = 10), ε4 homozygotes and heterozygotes were equally represented (50% each), and no cases occurred in ε4 non‐carriers (Table 2). Overall, ε4 homozygosity appeared more frequently in ARIA‐H and combined ARIA presentations, whereas ε4 heterozygosity predominated in isolated ARIA‐E, and ε4 non‐carriers were infrequently represented across ARIA subtypes. We did not observe a relationship between the numbers of pre‐existing microbleeds and ARIA incidence, although it is likely our relatively low number of patients developing ARIA is too small to draw any conclusions.

TABLE 2.

ARIA‐E and ‐H rates by APOE carrier status.

APOE ε4 status	Isolated ARIA‐E (N = 10)	Isolated ARIA‐H (N = 12)		ARIA‐E and H (N = 10)
		No siderosis	Siderosis	No siderosis	Siderosis
ε4 homozygote	1 (10)	5 (50)b	0 (0)	2 (33)	2 (50)
ε4 heterozygote	6 (60)	4 (40)a	2 (100)	4 (67)	2 (50)
ε4 non‐carrier	3 (30)	1 (10)a	0 (0)	0 (0)	0 (0)

Note: All numbers are listed as N (%). ^a,b indicate groups which are statistically significant different as per post‐hoc analyses.

Abbreviations: APOE, apolipoprotein E; ARIA‐E, amyloid‐related imaging abnormalities with edema/effusion; ARIA‐H, amyloid‐related imaging abnormalities with hemorrhage;

Among 243 treated patients, 95 (39%) experienced an infusion‐related reaction (IRR) (Table 3). Age, sex, cognitive diagnosis, MoCA score, race and ethnicity, and APOE ε4 status did not differ significantly between patients with and without IRRs (Table 3). Mean age was 73.1 ± 6.6 years in the IRR group and 74.3 ± 7.5 years in the no‐IRR group (p = 0.20). Sex distribution was similar (59% vs. 51% female; p = 0.25), as were rates of mild cognitive impairment (MCI) (79% vs. 80%; p = 0.88). APOE ε4 carriage was comparable between groups (p = 0.80). Anti‐diabetic medication use was significantly more common among patients with IRRs (21% vs. 6%; p < 0.001) (Table 3). Use of antihypertensives, antiplatelets, and low‐dose aspirin did not differ between groups. Most IRRs were associated with flu‐like symptoms (including chills and rigors) (55%) and headache (43%), while less common symptoms included nausea (17%), confusion (8%), dizziness (7%), trouble walking (4%), and seizure (2%). Of the IRRs, 56 (58%) were considered mild and treated at home with over‐the‐counter medication, 29 (31%) were considered moderate and occurred during the infusion visit requiring intravenous medication, and 10 (11%) were severe, requiring an ED visit and treatment discontinuation.

TABLE 3.

Baseline characteristics of individuals by IRR at initial administration.

Parameter	No IRR (N = 148)	IRR (N = 95)	p‐Value
Mean age (SD), y	74.3 ± 7.5	73.1 ± 6.6	0.20
Sex			0.25
Female	76 (51)	56 (59)
Male	72 (49)	39 (41)
Education level 1
≤12 y	27 (19) a	12 (14) a	0.025 *
13‐15 y	15 (11) a	22 (25) b
16 y	50 (35) a	23 (26) a
>16 y	50 (35) a	31 (35) a
Cognitive diagnosis			0.88
MCI	118 (80)	75 (79)
Mild dementia	30 (20)	20 (21)
MOCA score (SD) 2	21.2 (3.5)	21.6 (3.4)	0.20
Race and ethnicity
Asian	1 (1)	0 (0)	0.42
Black	3 (2)	1 (1)	0.56
Hispanic	4 (3)	1 (1)	0.38
White	142 (96)	94 (99)	0.17
APOE ε4 status			0.80
ε4 homozygote	18 (12)	12 (13)
ε4 heterozygote	93 (63)	57 (60)
ε2/ε4 genotype	1 (1)	0 (0)
ε3/ε4 genotype	92 (62)	57 (60)
ε4 non‐carrier	37 (25)	26 (27)
ε2/ε3 genotype	1 (1)	2 (2)
ε3/ε3 genotype	36 (24)	24 (25)
Medication use
Antihypertensive	87 (59)	51 (54)	0.43
Anti‐diabetic	9 (6)	20 (21)	<0.001*
Antiplatelet	4 (3)	4 (4)	0.52
Aspirin low dose	38 (36)	19 (20)	0.32

Note: All numbers are listed as N (%) for categorical variables and mean (SD) for continuous variables.

^* p < 0.05 for chi‐squared testing for categorical variables with post‐hoc Bonferroni analyses or t‐testing for continuous variables.

¹ N = 230.

² N = 218. MOCA scores adjusted for educational level. Remaining cohort had lecanemab eligibility determined by MMSE.

^{a, b} indicate groups which are statistically significant different as per post‐hoc analyses.

Abbreviations: APOE, apolipoprotein E; IRR, infusion‐related reaction; MCI, mild cognitive impairment; MoCA, Montreal Cognitive Assessment; SD, standard deviation.

We had 10 patients who presented for urgent or emergency evaluation of infusion‐related reactions during treatment. Most were male (6/10) and in their mid‐to‐late 70s; seven were APOE ε4 heterozyotes, one was APOE ε4 homozygous, and two were non‐APOE ε4 carriers. Eight patients had MCI and two had dementia. Presenting symptoms included headache (n = 3), syncope and/or falls (n = 3), confusion (n = 2), breakthrough seizure (n = 1), and non‐neurologic complaints such as chest or epigastric pain (n = 1). Three presentations occurred on the day of infusion. Acute evaluation MRI within 24 h of symptom presentation demonstrated no new ARIA‐E (edema/effusion) or ARIA‐H (hemorrhage) in seven cases. Two patients were found to have ARIA on follow‐up imaging: one ε3/ε4 carrier with moderate ARIA‐E and severe ARIA‐H, and one ε4/ε4 carrier with mild ARIA‐E and moderate ARIA‐H. One ε3/ε3 patient discontinued treatment prior to follow‐up imaging. Overall, most acute presentations were not associated with new radiographic ARIA.

We have not observed any differences in systolic or diastolic blood pressures between those patients with and without ARIA, presumably due to our strict blood pressure protocols and modification of infusion schedules when blood pressures are elevated. We have delayed 20 infusions to date, 11 of which were due to elevated blood pressure at the infusion center, two were due to illness or hospitalization, five were due to personal scheduling conflicts, two were due to insurance or other issues.

4. DISCUSSION

Working with the leadership of our statewide medical system and all the healthcare units involved in the program, we built a streamlined, thorough, and patient‐focused program summarized in Figure 3B. The BHP has a broad focus, which involves making brain health a priority in all aspects of medical care in the state of Indiana. Establishment of the ATT program is just a single facet of the program's focus, and one that will be changing rapidly as new forms of delivery of ATTs evolve, and new treatments are developed.

A critical aspect of the BHP is the involvement of primary care physicians and leadership. The authors firmly believe that the future of AD treatment will involve earlier detection of disease which ultimately will be completed through biomarkers preceding the onset of measurable cognitive symptoms. This future involves primary care as the most important site for detection. Having clear evaluation and referral criteria, easy and accessible referral pathways, and established communication with primary care providers is essential to future success. Building infrastructure, education, and focus on managing patients with cognitive impairment in the current state is paramount so that the future state is feasible. Our approach with primary care providers thus far has involved a partnership with education about early detection, treatment, and the development of strategies to facilitate the referral and management of cognitively impaired individuals. ¹¹ A system‐wide collaboration, led by the IU Neuroscience Institute, has developed and implemented resources to promote patient‐centered early detection care pathways from primary care to specialty care, including the use of digital cognitive assessment tools and a primary care BHN. These collaborative clinical and research pathways, including the IU Center for Aging Research, have developed and are testing multiple programs across the state for early detection of cognitive impairment in primary care.

The described overall ARIA rate using lecanemab is less than what was reported in clinical trial data. We attribute the lower rates to several factors. We have focused on the use of 3T MRI for screening purposes which may have led to screen failing of potential treatment candidates that were at higher risk and may have been included if only 1.5T was used. Although hypertension was not statistically significant in terms of ARIA risk in our cohort, this is likely influenced by encouraging the use of blood pressure logs, directing those with elevated blood pressures back to primary care for more aggressive management, and postponing infusions for individuals with a blood pressure greater than 140/90. At the neuroBHN education visit, potential infusion patients and caregivers are given information about infusion postponement for elevated blood pressure, which can lead to “wasted trips.” We have observed this educational information to be helpful in promoting engagement with blood pressure logs, adherence to anti‐hypertensive medications, and patient‐engagement with primary care providers for additional anti‐hypertensive medications when blood pressures are borderline. It is unclear what impact that has had due to mixed findings in literature, ⁵ , ²⁴ but theoretically we suspect a favorable impact from well controlled blood pressures given the importance of good vascular risk factor control and models of disease/vasculopathy. Our vigilance about co‐inflammatory processes increasing the risk of ARIA may also play a role in our significantly lower ARIA rate than predicted. We advise all participants to reschedule their infusion if they have a viral infection, are running a fever, or generally feeling sick. If our infusion nurses note an acute illness, or event that could be relevant, the team is alerted and in most instances patients have their infusion rescheduled for the following week or to skip the infusion until the next scheduled. Although no relevant increase in bleeding risk has been found with anti‐platelet medications, we counsel patients to discuss the need for regular use of anti‐platelet medications with the prescriber, suggesting caution with regular non‐steroidal anti‐inflammatory (NSAID) use, and we ensure that patients understand that anything that increases their bleeding risk could potentially increase their risk for developing ARIA.

Most of our IRRs have been mild‐to‐moderate in nature. The rate of infusion reactions has substantially decreased since our April 2025 institution of pre‐medications before the first four infusions. We have followed the AUR very carefully, but recently have given more flexibility for atypical presentations and concurrent medications that we feel may have lower risk (e.g., monoclonal antibody treatments for migraine, cholesterol), with careful counseling about unknown levels of risk, and discussion with the prescribing physician to ensure there is mutual agreement as to the plan of treatment

In conclusion, the creation of a successful BHP has required a multi‐disciplinary and system‐wide approach to care. Innovative roles in nursing care, the BHN, both in primary care and neurology departments, helps support the patient journey from primary care to ultimate treatment with ATTs. As the future of AD care likely will involve pre‐symptomatic treatments, we firmly believe that developing awareness, education, and the infrastructure now with ATTs is a logical step. Ultimately, we anticipate that the system we have established is one that will be sustainable and continue to provide a high level of safety and patient care as access to ATTs, and other future AD treatments is expanded.

CONFLICT OF INTEREST STATEMENT

Jared R Brosch clinical trial research support from ABBVIE, Alnylam Pharmaceuticals, Athira Pharma, Biogen, Eisai America Inc, Eli Lilly and Company, F. Hoffman‐La Roche Ltd, Washington University, Takeda Pharmaceuticals, University of Southern California; Internal Advisory Consulting for Eli Lilly and Company, Eisai America Inc, F. Hoffman‐La Roche LTD, Abbvie, Syndeio Biosciences; Shareholder/Founder for Monument Biosciences. Liana G. Apostolova is a consultant for Biogen, Prothena, IQVIA, Genentech, Siemens, Corium, Eli Lilly, GE Healthcare, Eisai, Roche Diagnostics, Alnylam, Otuska. Meredith Tobar internal advisory Eisai America Inc. Donna M Wilcock is an SAB member and paid consultant for Vigil Neurosciences Inc., SynapsDx, Longeveron Inc., Beren Therapeutics Inc. Other authors have no disclosures. Author disclosures are available in the Supporting Information.

CONSENT STATEMENT

Informed consent was not required for this study. The study utilized electronic health record information on clinical patients.

Supporting information

Supporting Information: trc270291‐supp‐0001‐icjme.pdf

Brosch JR, Wang S, Apostolova LG, et al. The Indiana University Brain Health Program to deliver amyloid‐targeted therapy to Alzheimer's disease patients. Alzheimer's Dement. 2026;12:e70291. 10.1002/trc2.70291