Pharma is interested beginning CADASIL trials in the USA, yet no national or even large published USA cohorts exist. Cross-sectional and longitudinal observational studies to identify the optimal clinical outcomes and biomarkers for future treatment trials are needed in the USA.

To begin to characterize a CADASIL cohort at UCSF, we began to assess all potential cases at our center.  We queried our EPIC EMR for any patients evaluated at UCSF over the past 20 years with either the ICD-10 code for CADASIL (167.85) or any of the following three terms in their EMR notes “CADASIL,” “cerebral autosomal dominant arteriopathy subcortical infarcts and leukoencephalopathy” or “NOTCH3.”

The EPIC query yielded 750 potential CADASIL patients; review of the first 150 determined 13.3% had a NOTCH3 mutation and/or clinical CADASIL (based on family history and symptoms and/or diagnostic MRI), estimating our total cohort size at ~100. Preliminary detailed phenotypic characterization of the initial 35 patients identified several features. We have ~23 families representing 21 different NOTCH3 variants.  89% of patients were symptomatic (migraine alone not considered symptomatic).  On brain MRI, 91% had anterior temporal lobe (ATL) white matter disease (WMD), 23% had microbleeds, and none had cortical siderosis. The mean Fazekas white matter deep extent and periventricular scores were 2.5 and 2.6, respectively (range 0-3). 48% had migraine, half with aura, and 31% had headache-NOS.  Interestingly, as some non-carriers had migraines and mild WMD (Fazekas 1, sparing ATL), one should not presume that at-risk individuals with unknown NOTCH3 status with these features to be “symptomatic” and/or a carrier.