Authors:
Martina Göldlin (Bern | CH)
Madlaine Mueller (Bern | CH)
Bernhard Siepen (Bern | CH)
Waldo Valenzuela Pinilla (Bern | CH)
Piotr Radojewski (Bern | CH)
Arsany Hakim (Bern | CH)
Johannes Kaesmacher (Bern | CH)
Thomas Meinel (Bern | CH)
Leander Clénin (Bern | CH)
Mattia Branca (Bern | CH)
Davide Strambo (Lausanne | CH)
Tim Fischer (St. Gallen | CH)
Friedrich Medlin (Fribourg | CH)
Nils Peters (Zurich | CH)
Emmanuel Carrera (Geneva | CH)
Karl-Olof Lövblad (Geneva | CH)
Grzegorz Karwacki (Lucerne | CH)
Carlo Cereda (Lugano | CH)
Julien Niederhaeuser (Nyon | CH)
Marie-Luise Mono (Zurich | CH)
Achim Mueller (Zurich | CH)
Susanne Wegener (Zurich | CH)
Sabine Sartoretti-Schefer (Winterthur | CH)
Alexandros A. Polymeris (Basel | CH)
Valerian Altersberger (Basel | CH)
Marios-Nikos Psychogios (Basel | CH)
Rolf Sturzenegger (Chur | CH)
Michael Schaerer (Solothurn | CH)
Susanne Renaud (Neuchâtel | CH)
Werner J. Z'Graggen (Bern | CH)
David Bervini (Bern | CH)
Leo Bonati (Rheinfelden | CH)
Marcel Arnold (Bern | CH)
Urs Fischer (Basel/Bern | CH)
David Werring (London | GB)
David Seiffge (Bern | CH)
Background: Cerebral small vessel disease (SVD) is the major cause of intracerebral hemorrhage (ICH), but there is no comprehensive classification of ICH subtypes according to different SVD phaenotypes. We aimed to develop an MRI-based classification for SVD-related ICH and assess association with outcomes.
Methods: We performed a retrospective study of prospectively collected data of patients enrolled in the national, multicenter Swiss Stroke Registry from 2013-2019. We included consecutive patients with non-traumatic, SVD-related ICH and available MRI. Patients were classified according to a novel MRI-based classification using haemorrhagic (microbleeds, cortical siderosis) and non-haemorrhagic (white matter hyperintensities, lacunes, perivascular spaces) MRI markers and haematoma location as cerebral amyloid angiopathy (CAA), deep perforator arteriopathy (DPA), mixed SVD or undetermined SVD. The primary clinical outcomes were recurrent ICH or ischemic stroke at 3 months. We performed a Firth penalized logistic regression and competing risk analysis according to Fine and Gray.
Results: We enrolled 1180 patients (age (IQR) 73 (62-80) years, baseline NIHSS 6 (2-12), 538/1180 (45.6%) lobar hematoma location, systolic blood pressure on admission 166 (145-185) mmHg). During follow-up 57 patients had 58 events (29 ICH, 29 ischemic strokes, event rate 4.8%). The rate of recurrent ICH and ischaemic stroke at 3 months was 4%/1% for CAA, 0.7%/2.6% for DPA, 2.7%/2.9% for mixed SVD and 3.8%/3.4% for undetermined SVD, respectively, without a statistically significant difference between groups.
Discussion: This new MRI-based SVD-ICH phaenotype classification based on objective radiological markers is feasible and may improve classification and reporting of ICH subtypes in clinical practice and research.