Cardiovascular

Dynamic imaging markers of incident microhemorrhages and superficial siderosis in the A4 study: A longitudinal person-interval analysis.

TL;DR

Both current microhemorrhage burden and recent microhemorrhage accumulation were independently associated with increased short-term risk of incident hemorrhagic MRI abnormalities, supporting 'a dynamic model of hemorrhagic risk in Alzheimer's disease.'

Key Findings

Current microhemorrhage burden was independently associated with increased risk of hemorrhagic MRI events in both time-to-event and all-event models.

  • Time-to-event model: OR 1.37, 95% CI 1.09–1.73
  • All-event model: OR 1.24, 95% CI 1.04–1.49
  • Analysis used a discrete person-interval dataset constructed from longitudinal MRI data from the A4 study
  • 1069 participants contributed 3647 intervals to the analysis

Recent microhemorrhage accumulation was independently associated with increased short-term risk of hemorrhagic MRI events.

  • Time-to-event model: OR 1.83, 95% CI 1.01–3.30
  • All-event model: OR 1.43, 95% CI 1.00–2.04
  • Recent accumulation was modeled as a dynamic variable alongside current burden and baseline covariates
  • Both current burden and recent accumulation were independently associated with increased risk

Among 1069 participants across 3647 intervals, 171 hemorrhagic events occurred during the study period.

  • The study used longitudinal MRI data from the A4 clinical trial (NCT02008357), a trial of solanezumab for older individuals at risk for memory loss
  • The outcome was defined as two or more new microhemorrhages or one or more new superficial siderosis between consecutive MRI scans
  • A discrete person-interval analytical framework was used to model short-term hemorrhagic risk
  • Models incorporated both static baseline covariates and dynamic temporal imaging variables

Temporal imaging variables may provide independent prognostic information beyond baseline risk factors for hemorrhagic MRI abnormalities.

  • The authors state that temporal imaging variables 'may provide independent prognostic information beyond baseline risk'
  • This supports what the authors describe as 'a dynamic model of hemorrhagic risk in Alzheimer's disease'
  • The findings are relevant to ARIA-H (amyloid-related imaging abnormalities with hemosiderin deposition) risk stratification
  • Evidence exploring short-term dynamic predictors of incident microhemorrhages and superficial siderosis was described as 'limited' prior to this study

What This Means

This research examined brain imaging data from over 1,000 older adults enrolled in the A4 clinical trial to understand what factors predict the short-term development of small brain bleeds (microhemorrhages) and a related finding called superficial siderosis — both of which are types of hemorrhagic abnormalities seen on MRI scans. The study tracked participants across nearly 3,650 time intervals and found 171 instances where these hemorrhagic events occurred. Importantly, the researchers looked not just at fixed risk factors measured at the start of the study, but also at changing ('dynamic') imaging features over time, such as how many microhemorrhages a person currently had and whether they had been accumulating new ones recently. The study found that both the current number of microhemorrhages and the recent rate at which new ones were appearing were each independently linked to a higher risk of developing more hemorrhagic abnormalities in the short term. In other words, someone who had been accumulating microhemorrhages recently was at greater risk going forward, above and beyond what their baseline risk profile would suggest. This supports the idea that brain hemorrhagic risk in Alzheimer's disease is not static but evolves dynamically over time. These findings matter because microhemorrhages and superficial siderosis are directly relevant to a serious side effect called ARIA-H that can occur with certain Alzheimer's treatments, including amyloid-targeting therapies. This research suggests that monitoring how these imaging findings change over time — rather than relying solely on initial scans — could improve how clinicians assess and manage a patient's ongoing risk during treatment.

Have a question about this study?

Citation

Hill C, Morgan H, Michopoulou S, Niranjan M, Kipps C. (2026). Dynamic imaging markers of incident microhemorrhages and superficial siderosis in the A4 study: A longitudinal person-interval analysis.. Alzheimer's & dementia : the journal of the Alzheimer's Association. https://doi.org/10.1002/alz.71792