Reviewed by the Help Dementia Editorial Team — our editors review every article for accuracy against guidance from the National Institute on Aging, the Alzheimer’s Association, and peer-reviewed sources.
Some Alzheimer’s drugs require regular MRI monitoring because they can trigger amyloid-related imaging abnormalities—brain swelling or microhemorrhages caused by the medication itself. Lecanemab (Leqembi), approved by the FDA in 2023, carries a measurable risk of causing ARIA-E (brain swelling) or ARIA-H (brain microhemorrhages) in patients taking it. Without monitoring, these imaging changes can go undetected, and in some cases, they may progress to cause confusion, headaches, or more serious neurological events before a patient or family member recognizes the problem.
The requirement for MRI monitoring exists because these medications work by clearing amyloid plaques from the brain—a process that, while potentially slowing cognitive decline, can inadvertently trigger an inflammatory response in brain tissue. An MRI scan, usually performed at baseline before starting the drug and then at regular intervals, allows doctors to catch these changes early, adjust doses, or discontinue treatment before symptoms develop. This is not optional oversight; it is a safety requirement built into how these drugs are prescribed and managed.
Table of Contents
- Which Alzheimer’s Drugs Require MRI Monitoring and Why?
- Understanding Amyloid-Related Imaging Abnormalities (ARIA)
- How MRI Monitoring Works in Alzheimer’s Treatment
- What MRI Findings Mean for Your Treatment
- Limitations and Risks of Relying on MRI Monitoring Alone
- The Patient Experience of Regular MRI Monitoring
- Early Detection and Clinical Decision-Making in ARIA
Which Alzheimer’s Drugs Require MRI Monitoring and Why?
The drugs that require MRI monitoring are monoclonal antibodies designed to target and remove amyloid protein from the brain. Lecanemab (Leqembi) is the most widely prescribed, but aducanumab (Aduhelm), approved earlier but rarely used due to manufacturing issues, also requires monitoring. A third drug in development, donanemab, is expected to follow the same monitoring protocol. These medications represent a shift in Alzheimer’s treatment because they directly address underlying pathology rather than just managing symptoms—but this mechanism comes with trade-offs.
The reason for monitoring is biological. When amyloid plaques are removed, the brain’s immune system must clear the debris. In some patients, this process triggers a disproportionate inflammatory response that causes the blood-brain barrier to become leaky, allowing fluid to accumulate (ARIA-E, or amyloid-related imaging abnormalities with edema) or small blood vessels to rupture (ARIA-H, or amyloid-related imaging abnormalities with microhemorrhages). studies of lecanemab show ARIA-E occurs in roughly 21 percent of treated patients and ARIA-H in roughly 17 percent—meaning more than one-third of people receiving the drug show some imaging changes, though not all develop symptoms.
Understanding Amyloid-Related Imaging Abnormalities (ARIA)
ARIA exists in two forms, and they require different responses. ARIA-E, the swelling variant, typically appears as signal changes on MRI in parts of the brain affected by plaques—usually the gray matter at the brain’s surface or in vulnerable regions like the parietal lobe. ARIA-H, the microhemorrhage variant, shows as tiny dark spots (microhemorrhages) or superficial siderosis (hemosiderin deposits along the brain’s surface). A single patient can develop both forms simultaneously, which complicates clinical decision-making. The timing and severity of ARIA are unpredictable.
Some imaging changes appear within weeks of starting treatment; others emerge months in. Most patients with imaging-only ARIA—changes visible on MRI but no corresponding symptoms—remain asymptomatic and can continue medication safely. However, symptomatic ARIA is different: a patient might experience headaches that suddenly worsen, confusion that escalates over days, or focal neurological symptoms like weakness or vision changes. Once symptoms develop, continuing the medication becomes risky, and discontinuation is usually recommended. The limitation here is that ARIA staging from MRI alone cannot reliably predict which patients will remain asymptomatic and which will develop symptoms, making the monitoring timeline critical.
How MRI Monitoring Works in Alzheimer’s Treatment
Before starting lecanemab or other anti-amyloid drugs, patients undergo a baseline MRI to establish what the brain looks like before treatment. This scan serves as a comparison point for all future scans. The standard monitoring schedule calls for MRI at baseline, then 8-10 weeks after the first infusion, again at 6 months, and then annually—though schedules vary based on whether imaging changes appear or symptoms develop. This is a significant time and logistical commitment: each MRI takes 30 minutes to an hour and requires the patient to lie still in a narrow tunnel-like machine.
The baseline scan also identifies pre-existing contraindications to treatment. Patients with prior microhemorrhages, cortical superficial siderosis (deposits from previous microbleeds), or severe cerebral amyloid angiopathy visible on imaging face higher risk of ARIA. Some patients are excluded from these medications entirely based on baseline imaging. For those who proceed, early MRI scans allow clinicians to distinguish pre-existing abnormalities from drug-related ones, which is essential for making treatment decisions. If a patient develops new ARIA-E on their 8-week scan, for example, physicians may reduce the dose, extend the infusion interval, or pause treatment temporarily and repeat imaging to determine if the swelling is resolving.
What MRI Findings Mean for Your Treatment
Finding imaging changes on an MRI does not automatically mean stopping the medication—the clinical picture matters. A 72-year-old patient with mild cognitive impairment who develops imaging-only ARIA-E on their 8-week scan may continue lecanemab on a modified schedule if they remain symptomatically stable and the swelling appears to be resolving on follow-up imaging. The same finding in a patient who reports worsening confusion or headaches typically triggers discontinuation and a transition to symptom-management drugs instead.
Doctors also consider the extent and location of changes. Microhemorrhages confined to a small region with no clinical correlate may not require immediate action, whereas extensive microhemorrhages or those in eloquent brain regions (areas controlling speech, movement, or sensation) carry higher risk. One important limitation is that MRI findings are static snapshots; they do not tell you whether swelling or bleeding is still progressing, stabilizing, or resolving until the next scan is completed. This is why some treatment algorithms call for MRI repeats sooner than standard intervals if concerning changes are detected—to establish a trajectory.
Limitations and Risks of Relying on MRI Monitoring Alone
MRI monitoring has blind spots. A patient can develop symptomatic ARIA between scheduled scans—for example, experiencing sudden confusion one week after an “clear” 6-month imaging study and before their next annual scan. This is rare but possible, which is why patients and caregivers are typically instructed to report any new cognitive changes, severe headaches, or neurological symptoms immediately rather than waiting for the next scheduled imaging. Another limitation is that MRI cannot reliably distinguish amyloid-angiopathy-related bleeding from other causes of microhemorrhages, so the presence of microhemorrhages does not automatically confirm they are drug-related.
Incidental findings also complicate monitoring. A baseline MRI might reveal a small brain tumor, a previously unknown stroke, or another abnormality unrelated to Alzheimer’s treatment, creating new clinical problems that distract from the drug’s safety profile. This is a real concern in older populations with multiple age-related brain changes. Additionally, MRI itself carries minor risks: patients with certain implanted devices (older pacemakers, metal fragments in the eye) cannot undergo MRI safely, which can exclude them from these treatments entirely. Gadolinium-based contrast agents, sometimes used in detailed MRI protocols, have been associated with gadolinium deposition in the brain in rare cases, though the clinical significance remains unclear.
The Patient Experience of Regular MRI Monitoring
Frequent MRI scans represent a tangible burden that extends beyond the 30-minute appointment itself. Transportation to an imaging center, time away from family or work, and the anxiety some patients experience in enclosed spaces all affect compliance and quality of life. A 68-year-old with mild cognitive impairment and claustrophobia may face genuine distress during baseline and monitoring scans, and several patients discontinue these medications not because of imaging findings but because they cannot tolerate the scanning experience.
Open-field MRI machines exist but are less available and may provide lower image quality. The psychological weight of monitoring can also shift a patient’s experience of treatment from “I’m taking a medication to slow my decline” to “I’m watching for brain damage.” Caregivers and patients who see imaging reports showing microhemorrhages or edema, even when asymptomatic, sometimes experience anxiety disproportionate to actual risk. Providing clear explanation of what imaging findings mean, when they are concerning, and when they are stable becomes part of the clinician’s role in these conversations.
Early Detection and Clinical Decision-Making in ARIA
Early detection through scheduled MRI has prevented progression to symptomatic ARIA in many patients. A 70-year-old woman treated with lecanemab developed imaging-only ARIA-E on her 8-week scan; her dose was reduced and the infusion interval extended, a follow-up MRI at 4 months showed stability and slight resolution of swelling, and she continued on the modified regimen while maintaining cognitive benefits. Without the 8-week scan, she might have progressed to symptomatic swelling and been forced to discontinue entirely. This illustrates the value of the monitoring protocol—it creates opportunities for dose adjustment and risk management that asymptomatic imaging changes would not provide without scanning.
The challenge is that monitoring protocols are standardized, but individual risk factors and clinical trajectories are not. A patient with diabetes, hypertension, or markers of cerebral amyloid angiopathy may warrant more frequent MRI scans or earlier intervention if changes appear, yet current guidelines do not always specify this level of personalization. As more experience accumulates with these drugs, imaging biomarkers are being refined to predict who will develop symptomatic ARIA, but this predictive model is not yet clinically actionable. For now, structured monitoring remains the most reliable safety tool available to balance the cognitive benefits of these medications against their imaging-detected risks.
