Stroke Risk and Antithrombotic Treatment During Follow-up of Patients With Ischemic Stroke and Cortical Superficial Siderosis

Background and Objectives In patients with ischemic stroke (IS) or transient ischemic attack (TIA) and cortical superficial siderosis (cSS), there are few data regarding the risk of future cerebrovascular events and also about the benefits and safety of antithrombotic drugs for secondary prevention. We investigated the associations of cSS and stroke risk in patients with recent IS or TIA. Methods We retrospectively analyzed the Microbleeds International Collaborative Network (MICON) database. We selected patients with IS or TIA from cohorts who had MRI-assessed cSS, available data on antithrombotic treatments, recurrent cerebrovascular events (intracranial hemorrhage [ICrH], IS, or any stroke [ICrH or IS]), and mortality. We calculated incidence rates (IRs) and performed univariable and multivariable Cox regression analyses. Results Of 12, 669 patients (mean age 70.4 ± 12.3 years, 57.3% men), cSS was detected in 273 (2.2%) patients. During a mean follow-up of 24 ± 17 months, IS was more frequent than ICrH in both cSS (IR 57.1 vs 14.6 per 1, 000 patient-years) and non-cSS (33.7 vs 6.3 per 1, 000 patient-years) groups. Compared with the non-cSS group, cSS was associated with any stroke on multivariable analysis {IR 83 vs 42 per 1, 000 patient-years, adjusted hazard ratio [HR] for cSS 1.62 (95% CI: 1.14–2.28; p = 0.006)}. This association was not significant in subgroups of patients treated with antiplatelet drugs (n = 6, 554) or with anticoagulants (n = 4, 044). Patients with cSS who were treated with both antiplatelet drugs and anticoagulants (n = 1, 569) had a higher incidence of ICrH (IR 107.5 vs 4.9 per 1, 000 patient-years, adjusted HR 13.26; 95% CI: 2.90–60.63; p = 0.001) and of any stroke (IR 198.8 vs 34.7 per 1, 000 patient-years, adjusted HR 5.03; 95% CI: 2.03–12.44; p < 0.001) compared with the non-cSS group. Discussion Patients with IS or TIA with cSS are at increased risk of stroke (ICrH or IS) during follow-up; the risk of IS exceeds that of ICrH for patients receiving antiplatelet or anticoagulant treatment alone, but the risk of ICrH exceeds that of IS in patients receiving both treatments. The findings suggest that either antiplatelet or anticoagulant treatment alone should not be avoided in patients with cSS, but combined antithrombotic therapy might be hazardous. Our findings need to be confirmed by randomized clinical trials.