Leqembi is a monoclonal antibody that lowers levels of accumulated amyloid beta in the brain. The FDA approved it following successful clinical trials that demonstrated treatment slowed cognitive decline in patients with amyloid plaques in their brain and with mild cognitive impairment (MCI) or mild dementia due to Alzheimer’s disease (AD). However, this drug has a black box warning label because it, like other drugs in its class, can cause side effects called Amyloid Related Imaging Abnormalities (ARIA) that are visible on brain MRI scans. There are two types of ARIA: ARIA-E, characterized by edema or brain swelling, and ARIA-H, characterized by microhemorrhages or bleeding in the brain. ARIA usually occurs shortly after antibody treatment, is more likely during the first few months, and, in most cases, is both asymptomatic and resolves itself. However, it can be serious and potentially life-threatening in some patients, particularly APOE4 carriers. Based on this, the drug label recommends increased vigilance for ARIA in all patients, which means patients must undergo repeated MRI scans while on the drug. Unfortunately, we still do not know why ARIA occurs or why some patients experience it, so all patients must be considered at risk and monitored accordingly.
Dr. Doraiswamy suggests that tracking blood-based biomarkers immediately after antibody treatment could complement the current approach to monitoring safety. He and his colleagues have set out to identify and test potential biomarkers of ARIA in this study. They have compiled a comprehensive list of candidate proteins related to non-AD conditions that have symptoms similar to ARIA, like acute brain injuries, inflammation, angiogenesis, and vascular injury. They are particularly interested in the promise of two biomarkers: GFAP and UCH-L1. Levels of both rise in blood serum shortly after mild traumatic brain injuries (TBI) and strokes. Notably, levels of GFAP were much higher in stroke patients who also had hemorrhages. This and other evidence suggest that blood GFAP levels may be a sensitive marker of ARIA-relevant brain pathologies not yet visible on MRI scans.
For this project, Dr. Doraiswamy proposed a real-world study in patients receiving Leqembi to evaluate if their blood-based biomarkers reliably predict ARIA. They are enrolling 30 participants with MCI or mild AD who are undergoing treatment with Leqembi as part of their routine clinical care at the Duke Alzheimer’s Disease Research Center. Blood is collected at baseline and after each of the first three infusions to analyze biomarkers (GFAP, UCH-L1, NfL, and pTau) and study how antibody infusions affect these protein levels and whether changes can predict ARIA. They are using the Alamar Nulisa CNS 100 panel, which includes these four markers as well as proteins linked to amyloid, inflammation, and vascular injury. They are also storing blood samples for future biomarker testing as new candidates emerge and available funding allows. In addition to assessing their candidate proteins, they are also using a commercial platform (SomaScan) that can quantify 1,500 proteins in each plasma sample. This expanded protein screening increases the odds of Dr. Doraiswamy’d team generating preliminary data for a promising set of biomarkers that can be used in pursuit of funding for further clinical validation and testing.
At the close of the first year of funding, the study is progressing well despite initial challenges. The clinical uptake of Leqembi was slower and smaller than initially expected. The first three months were focused on obtaining Institutional Review Board (IRB) approval, training staff, encouraging clinician referrals, and addressing logistical issues with scheduling, all of which delayed enrollment. Coordination with the infusion clinic presented additional challenges due to patient and staff availability, and adverse events related to Leqembi often caused delays in infusions and subsequent blood draws. However, Dr. Doraiswamy ‘s team has successfully adapted to these hurdles.
To date, 28 patients (25 white, 13 female) receiving Leqembi infusions have completed the study, and blood samples have been banked for all participants. Adverse events (e.g., ARIA) and clinical outcomes (e.g., cognition) are being followed up for 12 months. In the second year, the team plans to complete adverse event and cognitive outcome follow-ups and analyze blood samples for biomarkers, including a proteomics panel using SomaScan. They have also begun building a robust study database, with data entry completed for all enrolled subjects. Results from serial MRI scans and records of adverse events are being systematically entered as well. Additionally, the team has developed a novel symptom rating scale for ARIA, which is being used to collect data during follow-up phone calls with participants.
