# Lecanemab over a two-year duration: Key insights from a regional specialty medical center

## TL;DR

This study examines the safety and outcomes of lecanemab treatment for Alzheimer's disease over two years, focusing on amyloid-related imaging abnormalities and patient characteristics.

## Contribution

The study provides real-world data on lecanemab safety, particularly the risk of ARIA in APOE ε4 homozygous patients.

## Key findings

- ARIA occurred in 22.3% of patients, with ε4 homozygotes having higher rates of ARIA compared to non-carriers.
- Infusion-related adverse events decreased after introducing a pre-infusion medication cocktail.
- Cognitive decline and ARIA progression were the main reasons for discontinuing lecanemab.

## Abstract

The anti-amyloid monoclonal antibody lecanemab (Leqembi®) treats patients with mild cognitive impairment (MCI) or mild dementia due to Alzheimer’s disease (AD). We sought to evaluate the incidence of amyloid-related imaging abnormalities) ARIA and other adverse events associated with lecanemab.

This retrospective and observational study features 187 patients who received at least one lecanemab infusion at our multidisciplinary Norton Neuroscience Institute Memory Center over a two-year duration (August 25, 2023-August 24, 2025).

A total of 109 (58.3 %) patients were diagnosed with MCI, and 78 (41.7 %) had mild dementia prior to starting lecanemab. The mean age at the initial infusion was 73 years (Range: 49-90 years). Most (127 [67.9 %]) patients were female, and the majority (181 [96.8 %]) were Caucasian. Of the 175 patients who underwent at least one surveillance brain MRI following lecanemab initiation, 39 (22.3 %) had evidence of ARIA (both ARIA-H and ARIA-E: 13 [33.3 %]; solitary ARIA-H: 17 [43.6 %]; and solitary ARIA-E: 9 [23.1 %]). Of these 39 patients, 20 (51.3 %) were ε4 heterozygous, 12 (30.8 %) were ε4 homozygous, and 7 (17.9 %) were ε4 non-carriers. Patients who were ε4 homozygous more frequently had evidence of any ARIA (p-value = 0.002), ARIA-E (p = 0.041), and ARIA-H (p = 0.004). Of the 25 patients who underwent at least one surveillance brain MRI and were ε4 homozygous, 12 (48.0 %) had ARIA detected. Five (12.8 %) patients with ARIA were symptomatic, requiring lecanemab suspension. Three of these symptomatic patients were ε4 homozygous, and two were ε4 heterozygous. The ARIA was most frequently detected on the surveillance brain MRI performed before the 5th infusion (29 [74.4 %] patients). All 39 cases of ARIA occurred before the 14th lecanemab infusion. Patients with more baseline microbleeds more frequently developed any ARIA (ARIA-H and ARIA-E) (p = 0.041) and solitary ARIA-H (p = 0.022). The presence of baseline microbleeds was associated with a higher frequency of solitary ARIA-H, though was only marginally statistically significant (p = 0.051). Sixty (32.1 %) patients experienced infusion-related adverse effects, with 54 (90.0 %) occurring after the first lecanemab infusion. Mild and transient headaches were most common, affecting 26 (48.1 %) of these patients after the first infusion. After initiating a pre-infusion oral cocktail of acetaminophen 650 mg, loratadine 10 mg, and famotidine 20 mg, the number of patients who experienced an infusion-related adverse event decreased from 45.2 % to 28.3 %. Thirty-two (17.1 %) patients discontinued lecanemab, primarily due to cognitive decline associated with progressive AD (10 [31.2 %]) and ARIA progression (9 [28.1 %]). Of the 73 patients who had MMSE scores performed at baseline and after 1 year post-lecanemab, 13 (17.8 %) had increased scores, 51 (69.9 %) had decreased scores, and the scores remained the same in 9 (12.3 %) patients.

Our findings suggest that ARIA is a significant concern especially in patients who are ε4 homozygous. Close monitoring of patients who are ε4 carriers is recommended to recognize any complications that may ensue.