Validation of the Boston Criteria Version 2.0 for Cerebral Amyloid Angiopathy in Patients Presenting With

Margaret H Downes1, Roshini Kalagara1, Christina P Rossitto1

  • 1Department of Neurosurgery, Icahn School of Medicine at Mount Sinai, New York, NY.

Neurology
|March 3, 2025
PubMed

Insights

The updated Boston criteria v2.0 improve diagnosis of cerebral amyloid angiopathy (CAA) by increasing sensitivity for probable CAA. This version effectively uses MRI biomarkers, including nonhemorrhagic markers like perivascular spaces, to enhance diagnostic accuracy.

Area of Science:

  • Neurology
  • Radiology
  • Pathology

Background:

  • Cerebral amyloid angiopathy (CAA) is a primary cause of lobar intracerebral hemorrhage (ICH) in older adults, leading to significant disability and recurrent events.
  • The Boston criteria, particularly version 2.0 (v2.0), have been updated with novel MRI biomarkers to enhance the diagnostic accuracy of CAA.

Purpose of the Study:

  • To validate the diagnostic performance of the Boston criteria v2.0 against version 1.5 (v1.5).
  • To compare the accuracy of v2.0 in diagnosing CAA in patients with spontaneous ICH who underwent surgical evacuation and brain biopsy.

Main Methods:

  • A retrospective single-center cohort study included 186 patients with spontaneous ICH who underwent surgical evacuation, brain biopsy, and preoperative MRI.
  • MRI markers assessed included hemorrhagic lesions (ICH, CMBs, cSS) and nonhemorrhagic markers (CSO-PVS, WMHs).
  • Pathologic confirmation of CAA was based on modified Vonsattel grading; diagnostic performance was evaluated using sensitivity, specificity, and predictive values.

Main Results:

  • Boston criteria v2.0 showed higher sensitivity for probable CAA (75% vs. 57%) with maintained specificity (96% vs. 99%) compared to v1.5.
  • Cortical siderosis (cSS) and lobar cerebral microbleeds (CMBs) were significantly associated with CAA.
  • Severe visible perivascular spaces in the centrum semiovale (CSO-PVS) were strongly associated with CAA, while multispot white matter hyperintensities (WMHs) were not.

Conclusions:

  • The Boston criteria v2.0 improve diagnostic sensitivity for probable CAA without sacrificing specificity, attributed to the inclusion of nonhemorrhagic markers like CSO-PVS.
  • Findings suggest v2.0 offers a more robust diagnostic framework for CAA.
  • Limitations include retrospective design and modest sample size.
Abstract