CADMUS: MRI Classification System Feasible for Small Vessel Disease-Related ICH

A novel magnetic resonance imaging (MRI)-based classification system, CADMUS, for cerebral small vessel disease (SVD)-associated intracerebral hemorrhage (ICH) is feasible and reproducible, according to study findings published in Neurology.

ICH is a complex disease that can be caused by multiple underlying conditions, including cerebral amyloid angiopathy (CAA), deep perforator arteriopathy (DPA), and a combination of both. The currently available classification systems for ICH were developed using single-center data or are too complex for use in routine practice and do not require MRI data.

Researchers designed this study to develop a classification system, the CADMUS (Cerebral Amyloid angiopathy, Deep perforator arteriopathy, Mixed CAA-DPA, or Undetermined SVD using hemorrhagic and nonhemorrhagic MRI markers). To do so, they included data from 2 independent cohorts, the prospective multicenter Swiss Stroke Registry (SSR) and the prospective Stroke InvestiGation in North And central London (SIGNAL) cohorts. Patients (N=1180; mean age, 73; 44.5% women) with ICH who had available MRI data obtained within 3 months were classified using the CADMUS.

The primary outcome was inter-rater reliability between the CADMUS and 2 experts.

Among the patients, 76.9% had hypertension, 50.9% hyperlipidemia, and 11.4% had a history of ICH.

The researchers classified patients as having CAA (13.1%), DPA (6.3%), mixed CAA-DPA (63.6%), and undetermined (17.2%).

Stratified by CADMUS phenotype, patient groups differed by:

• age,
• hypertension prevalence,
• systolic and diastolic blood pressures at presentation,
• hematoma epicenter,
• National Institutes of Health Stroke Scale (NIHSS) score at admission, and Structural vascular lesions, Medication, Amyloid angiopathy, Systemic disease, Hypertension, or Undetermined (SMASH-U) classification (all P <.001).

The 2 raters agreed on 38 of the 50 CADMUS phenotypes for the SSR cohort and 35 of the 44 from the SIGNAL cohort, indicating an inter-rater reliability of 0.69 (95% CI, 0.53-0.85) for SSR and 0.74 (95% CI, 0.58-0.90) for SIGNAL.

Compared with SMASH-U classification, the CADMUS phenotypes were in agreement for 12.2% of patients.

Follow-up data were available for 94.3% of patients. A total of 57 events occurred among 5.0% of patients, including ischemic stroke (2.8%) and recurrent ICH (2.5%).

Outcomes at 3-months were associated with female gender, CAA, and previous ICH but not with CADMUS phenotype. In the competing risk analysis, DPA associated with decreased risk for recurrent ICH relative to mixed CAA-DPA (subhazard ratio [sHR], 7.06×10^-7; 95% CI, 3.94×10^-7-1.03×10^-6).

In a sensitivity analysis, which excluded patients with a history of ICH, CAA was associated with ICH recurrence (adjusted odds ratio [aOR], 4.47; 95% CI, 1.54-12.97).

This study was limited, as severe ICH was likely underrepresented in this study.

“ICH can be classified into several CADMUS phenotypes, which are associated with different risk profiles and short-term outcomes,” the researchers concluded.

Disclosures: Some study authors declared affiliations with biotech, pharmaceutical, and/or device companies. Please see the original reference for a full list of disclosures.