Co-Author(s):
Robert Regenhardt, MD, PhD
Elif Gokcal, MD
Mitchell Horn, BSc
Nader Daoud, BA
Joshua Goldstein, MD, PhD
Anderson Christopher, MD
Patel Aman, MD
William Kimberly, MD, PhD
Viswanathan Anand, MD, PhD
lee Schwamm, MD
Jonathan Rosand, MD
Steven Greenberg, MD, PhD
Mahmut Gurol, MD
Massachusetts General Hospital
55 Fruit Street, Boston, MA 02114
Introduction | Objectives: Limited data exists on the predominant cerebral small vessel disease
(CSVD) type underlying oral anticoagulant-related intracerebral
hemorrhage (OAC-ICH), including non-vitamin K antagonist oral
anticoagulant-related ICH (NOAC-ICH) and vitamin K antagonist-related
intracerebral hemorrhage (VKA-ICH). Herein, we evaluated the two common
types of CSVD, hypertensive-CSVD (HTN-CSVD) and cerebral amyloid
angiopathy (CAA) in atrial fibrillation (AF) patients with NOAC-ICH and
VKA-ICH.
Methods: In a prospective database of consecutive non-traumatic ICH patients
admitted to a single referral center (April 2003 to December 2019),
brain MRI scans were reviewed to diagnose probable CAA (using modified
Boston criteria) or HTN-CSVD (the presence of a deep macrobleed with
strictly deep or no microbleeds [MBs]). The frequency of additional
magnetic resonance imaging (MRI) markers of high ICH risk including
lacunes and cortical superficial siderosis (cSS) were also assessed.
Results: 1841 patients were included in this study, of which 398 (21.6%) had
AF. In this cohort, 53 experienced a NOAC-ICH and 302 experienced a
VKA-ICH. The in-hospital case fatality was 39.6% for NOAC-ICH and
47.0% for VKA-ICH, both higher than non-OAC-ICH (35.3%, p< 0.05 for both comparisons). The prevalence of HTN-cSVD was
similar between NOAC-ICH and VKA-ICH (28.1% in NOAC-ICH and 27.8% in
VKA-ICH, p = 0.97). Probable CAA was diagnosed in 28.1% of
NOAC-ICH compared to 24.6% of VKA-ICH (p = 0.66). The frequency
of lacunes, cSS, and MBs was also similar between groups. Moreover,
96.9% of patients with NOAC-ICH had at least one neuroimaging marker of
high ICH risk compared to 86.4% of patients with sICH (p =
0.087).
Conclusions: CAA represents a substantial underlying pathology in both NOAC-ICH and
VKA-ICH. As 97% of NOAC-ICH patients had high ICH risk markers on MRI:
1) patients without these markers are unlikely to have ICH while on
NOACs, 2) nonpharmacological stroke prevention methods can be considered
in the presence of these markers.