What are clinical management algorithms (imaging thresholds and symptom criteria) for pausing or restarting anti‑amyloid therapy after ARIA is detected?
Executive summary
Clinical algorithms to pause or restart anti‑amyloid monoclonal antibody therapy after amyloid‑related imaging abnormalities (ARIA) hinge on two parallel axes: radiologic severity on standardized MRI sequences and the presence or absence of clinical symptoms; most trial-derived protocols recommend temporary suspension for imaging‑detected ARIA (particularly ARIA‑E) and permanent discontinuation in severe or complicated cases, with restart only after radiographic resolution and symptom improvement [1] [2] [3]. Guidance is antibody‑specific and informed by APOE genotype, baseline microhemorrhage burden and concomitant anticoagulant use, so real‑world algorithms combine imaging thresholds, symptom criteria, genotype/bleed risk and shared decision‑making [4] [5] [6].
1. What ARIA is and why algorithms matter
ARIA comprises two radiologic patterns—ARIA‑E (vasogenic edema/sulcal effusion) and ARIA‑H (micro/macrohemorrhage and superficial siderosis)—that can appear spontaneously but are more frequent with anti‑Aβ antibodies; because most ARIA are asymptomatic yet occasionally severe, standardized monitoring and clear dose‑hold/restart rules were embedded in trials and are now being translated into practice algorithms [7] [5] [8].
2. Detection: imaging protocols and baseline triage
Trials and expert reviews call for standardized MRI at baseline and serial scans during early dosing, using simple FLAIR and susceptibility‑sensitive sequences (GRE/SWI) that were sufficient in trials; baseline exclusion or caution thresholds often included >4 lobar microbleeds or imaging consistent with probable cerebral amyloid angiopathy (CAA) because these features predict higher ARIA risk [2] [5] [9].
3. Radiologic grading and the core imaging thresholds used in algorithms
Algorithms classify ARIA‑E by extent (focal vs. multifocal vs. large confluent edema) and ARIA‑H by number and newness of microhemorrhages or superficial siderosis; many trial tables guided action: small, asymptomatic focal ARIA‑E might allow continued dosing or brief interruption with close imaging, while multifocal or confluent ARIA‑E typically triggered treatment suspension until resolution [10] [11] [8].
4. Symptom criteria that trigger pausing immediately
Presence of new neurological symptoms—headache, focal deficits, confusion, seizures—led trials and reviews to classify ARIA as symptomatic and to pause or permanently stop therapy depending on severity; symptomatic ARIA historically prompted more aggressive management than isolated radiologic findings [8] [1] [7].
5. Imaging thresholds that trigger pausing even if asymptomatic
Even when asymptomatic, imaging‑based thresholds—new multifocal ARIA‑E, expansion of edema, or new ARIA‑H with multiple new microbleeds or superficial siderosis—were reasons to suspend dosing until MRI demonstrated radiographic improvement or stability; specific numeric cutoffs varied by protocol and drug, but the presence of baseline CAA features or >4 microbleeds commonly influenced initial eligibility and post‑ARIA decisions [1] [9] [5].
6. Criteria and timing for restarting therapy after ARIA
Trials generally permitted restarting only after radiographic resolution or clear improvement of ARIA and resolution (or control) of symptoms; timing depended on rate of radiographic resolution on follow‑up MRI and clinical recovery, with some protocols requiring complete resolution of ARIA‑E and absence of progressive ARIA‑H before resuming dosing [1] [3] [12].
7. When to permanently discontinue and role of risk modifiers
Permanent discontinuation was recommended for severe, progressive, or life‑threatening ARIA (for example extensive hemorrhage or persistent multifocal hemorrhagic transformation) and whenever new medical needs (e.g., initiation of anticoagulation) increased hemorrhagic risk; APOE‑ε4 homozygosity, baseline microbleeds, CAA features and concurrent antithrombotic therapy are explicit modifiers used to individualize decisions toward discontinuation [4] [11] [3].
8. Practical takeaways, uncertainty and gaps
The operative clinical algorithm in practice is a risk‑stratified, MRI‑driven workflow: baseline MRI and risk assessment (including APOE and microbleed count), scheduled MRIs during early dosing, pause for symptomatic ARIA or significant radiographic ARIA (especially multifocal ARIA‑E/expanding ARIA‑H), and restart only after radiographic and clinical resolution—yet precise numeric thresholds and uniform consensus remain evolving and antibody‑specific, and available literature stresses the need for standardized reporting and further real‑world data [5] [11] [4].