Introduction: Hyperacute injury marker (HARM) is extrahematomal CNS barrier disruption in ICH marked by enhancement on post-gadolinium FLAIR sequences. We sought to identify factors associated with HARM development over time in a prospective ICH pilot study.
Methods: We conducted a longitudinal MRI-based study of primary ICH with images at baseline and 90 days. After excluding subjects without post-gadolinium FLAIR sequences and those without usable images, we evaluated HARM formation along with other MRI markers of vasculopathy, including microbleeds, lacunes, white matter disease, cortical superficial siderosis, and enlarged perivascular spaces. HARM severity was scored as 0 (absent), 1 (1 to 3 sulci affected across fewer than 10 slices), 2 (dispersed involvement; more than 3 sulci), or 3 (widespread, confluent involvement over the hemisphere). Imaging review was blinded to clinical variables. Univariate analyses were performed at both time points.
Results: Of 29 subjects with usable baseline sequences, 11 (37.9%) had HARM, compared to 12/18 (66.7%) at 90 days. Twelve subjects had usable sequences at both time points, and 4 (33%) of these had HARM at both time points. HARM was most frequently noted in the temporal and occipital lobes. At baseline, HARM severity was 1 in 7 subjects, 2 in 2 subjects, and 3 in 2 subjects. At 90 days, HARM severity was 1 in 10 subjects, 2 in 1 subject, and 3 in 1 subject. HARM was distinct from parenchymal T1 enhancement on visual inspection. We did not identify statistically significant associations between HARM formation and subject demographics, vascular risk factors, hematoma characteristics, or MRI markers of vasculopathy.
Conclusions: In this pilot study we found a high prevalence of HARM at baseline and 90 days. Future studies will examine whether HARM is associated with the formation of new CNS lesions or poor outcomes. HARM appears to be a second type of CNS barrier disruption and may represent a therapeutic target.