ARRS 2022 Abstracts

RETURN TO ABSTRACT LISTING


1443. Baseline Prevalence of Amyloid-Related Imaging Abnormalities in Early Alzheimer Disease
Authors * Denotes Presenting Author
  1. Saurabh Jindal *; Washington University
  2. Nelly Joseph-Mathurin; Washington University
  3. Farzaneh Rahmani; Washington University
  4. Lauren Koenig; Washington University
  5. Charlie Chen; Washington University
  6. Jorge Llibre Guerra; Washington University
  7. Tammie Benzinger; Washington University
Objective:
Aducanumab (Biogen Inc.) was recently approved by the United States Food and Drug Administration (FDA) for the treatment of mild cognitive impairment due to Alzheimer’s disease (MCI-AD) and early AD. However, it can cause amyloid-related imaging abnormalities (ARIAs), edema (ARIA-E), and/or hemosiderin (ARIA-H), which necessitates baseline and longitudinal magnetic resonance imaging (MRI) monitoring. Concerns from radiologists with this recommendation is that a patient’s treatment history is not often available at the time of image interpretation, and also that the cohorts included in the clinical trials might not represent actual patient populations. Therefore, we have investigated application of the FDA’s recommended ARIA scoring system to a clinical population undergoing dementia workup.

Materials and Methods:
We identified 194 patients referred by a dementia specialist for clinical brain MRI over a 2-year period. All of them had cognitive impairment or early dementia, as assessed by the clinical dementia rating score of 0 or 0.5. They were classified into AD, mixed AD/vascular, and others. Fifty of the 194 patient scans were assessed for ARIA-E, ARIA-H, and chronic small vessel disease (CSVD), with location and severity, recorded when present. ARIA-E, as visualized on fluid-attenuated inversion recovery (FLAIR) sequence represents vasogenic edema seen as cortical/subcortical white matter hyperintensities (WMH) and/or extravasated fluid phenomena seen as sulcal effusion. ARIA-E was dissociated from CSVD, which is seen as deep and/or periventricular WMH. ARIA-H as visualized on susceptibility-weighted imaging (SWI) presents as blooming foci attributable to parenchymal microhemorrhages (mH) and/or superficial siderosis.

Results:
In our preliminary dataset, the mean age of the cohort was 74 ± 8 years. None of the patients showed imaging findings consistent with ARIA-E. Overall, 70% were diagnosed with late-onset AD dementia, 10% with early-onset AD, 10% with mixed AD/vascular, and 10 % with atypical AD. Imaging findings suggestive of CSVD were seen in 80% of all patients and in 80% of those with mixed AD/vascular dementia. Fazekas score for WMH was significantly higher with older age (assessed using linear regression analysis, p = 0.007). Four of the patients, who all had AD dementia, showed mH not co-localized with WMH. None of the patients showed superficial siderosis.

Conclusion:
Baseline ARIA-E and ARIA-H are very rare in clinical practice, outside the setting of monoclonal antibody treatment for AD. There can be potential for imaging overlap between treatment-related ARIA and underlying cerebral amyloid angiopathy (CAA), which has a similar clinico-neuroimaging and CSF profile, and both entities are associated with ApoE e4 association; however, spontaneous CAA is rare. If patients with incomplete medical histories present for MRI, particularly in the emergency room setting, and if findings suggestive of ARIA are identified on imaging, radiologists should be alert to this, because cessation of therapy may be indicated.