Prevalence and clinical impact of white matter hyperintensities, cerebral microbleeds, and medial temporal lobe atrophy in Alzheimer's disease

Abstract

Background: This study examined the prevalence and clinical impact of white matter hyperintensities (WMH), cerebral microbleeds (CMB) and medial temporal lobe atrophy (MTLA) in patients with Alzheimer's disease (AD). Methods: We analysed the data for 70 patients with AD and 42 age-matched cognitively normal controls recruited from Memory and Geriatric Medicine Clinics, Queen Mary Hospital. Controls had no history of dementia, stroke, Parkinson’s disease, head injury, seizures, cancers within 5 years, end-stage organ failure, excessive alcohol or drug use, or psychiatric disease. 3T MRI brain images were retrospectively examined for the severity and location of WMH (Fazekas Score 0-3, >1 abnormal), CMB (Microbleed Anatomical Rating Scale 0-3, >0 abnormal) and MTLA (Scheltens Score 0-4, >1 abnormal). Results: Mean (± s.d.) ages were similar for AD patients (76.5 ± 8.5 yrs) and controls (74.1 ± 6.6 yrs). Gender and history of diabetes, hypertension, hyperlipidaemia, and ischaemic heart disease were similar. AD patients were more likely to be dependent (p=.037), and live in nursing homes (p=.05). Controls were more likely to have atrial fibrillation (p=.007) and take antiplatelets (p=.02) or anticoagulants (p=.009). In the AD group, 19% of patients had a history of stroke or TIA, and the mean MMSE was 18.9 ± 4.9. Comparing AD vs. control groups, the prevalence of: a) abnormal periventricular WMH was 26% vs. 19% (p=.42); b) abnormal subcortical WMH was 40% vs. 74% (p=.001); c) CMB was 39% vs. 36% (p=.76); and d) abnormal MTLA was 61% vs. 33% (p=.004). Severity grades of MRI abnormalities were significantly different between AD and control groups for periventricular WMH (AD worse, p<0.001), subcortical WMH (controls worse, p=.002), and MTLA (AD worse, p=.026), but not CMB. Regression models confirmed that MMSE was independently predicted by nursing home residency (p=.03), but not age or vascular risk factor. Presence of WMH, CMB or MTLA did not provide additional independent predictive power. Conclusion: MRI evidence of small vessel disease and medial temporal lobe atrophy are prevalent amongst AD patients and age-matched cognitively normal controls. Higher rates of AF and antithrombotic use may have accounted for more subcortical WMH lesions amongst normal controls. Severity grades of MRI abnormalities differ significantly between AD and controls, but do not independently predict cognitive function. These findings may indicate complex overlaps of neurodegenerative pathologies in AD and cerebral aging. Further studies are needed.