Amyloidosis and long-term risk of stroke and major bleeding: a nation-wide matched cohort study
N Noory, J Petersen, O Westin, E Havers-Borgersen, L Koeber, M S Maurer, K Rossing, F Gustafsson, E FosboelAbstract
Background
Cardiac amyloidosis (CA) is frequently complicated by atrial fibrillation and is considered a high thromboembolic condition. Anticoagulation is often initiated early, yet long-term comparative data on stroke risk are limited, and bleeding risk is a major concern.
Purpose
To evaluate the 5-year risk of stroke or transient ischemic attack (TIA) and major bleeding in patients with cardiac amyloidosis compared with matched controls from the background population.
Methods
In a nationwide registry-based cohort study, patients aged >60 years with incident amyloidosis (1996–2023) were identified. CA was defined as amyloidosis plus ≥1 of the following: cardiomyopathy, atrial fibrillation, chronic heart failure, or pacemaker implantation. Patients were matched to controls from the background population on cardiomyopathy status, atrial fibrillation, pacemaker implantation, age, sex, and index year, in a 1:3 ratio. Patients with CA entered follow-up at the time of diagnosis, and matched controls were assigned an index date corresponding to the matched case's diagnosis date. The primary outcome was stroke or TIA. The secondary outcome was major bleeding. Cumulative incidences were estimated using competing-risk methods, and hazard ratios (HR) were derived from Cox models adjusted for comorbidities and antithrombotic therapy.
Results
A total of 735 patients with CA and 2,205 matched controls were included (median age 79 years; 75% male). Baseline antithrombotic therapy use was similar between groups. At 5 years, stroke or TIA occurred in 5.4% of patients with CA and 6.2% of controls (adjusted HR 1.10, 95% CI 0.67–1.81), with comparable results in analyses stratified by atrial fibrillation status. Major bleeding occurred in 21.4% and 15.8%, respectively (adjusted HR 2.09, 95% CI 1.65–2.64). Results were consistent in sensitivity analyses restricted to patients surviving beyond 6 months after the index and in analyses excluding individuals receiving antithrombotic therapy.
Conclusions
CA was not associated with increased long-term risk of stroke/TIA compared with matched controls but was associated with a higher risk of major bleeding, supporting individualized antithrombotic strategies in this population.