Impact of D-dimer for pathologic differentiation on transesophageal echocardiography in embolic stroke of undetermined source: a single-center experience

Background Embolic stroke of undetermined source (ESUS) encompasses diverse embologenic mechanisms, which transesophageal echocardiography (TEE) is critical to detect. Specific markers related to each embolic source in ESUS is not fully studied. We focused on D-dimer levels, and explored the association of D-dimer with potential embolic sources (PES) identified on TEE in ESUS. Methods Consecutive patients with ESUS were included in this study. Clinical characteristics including D-dimer levels were compared between ESUS patients with and without TEE, and among none of, one, and at least two PES in ESUS patients undergoing TEE. Factors related to elevation of D-dimer were analyzed. Results A total of 211 patients (age, 69.3 +/- 13.2 years; 149 males) with ESUS were enrolled. Of these, 115 received TEE, displaying significantly younger age and lower D-dimer levels than patients without TEE (P < 0.05), and 20 (17%), 61 (53%), and 34 (30%) patients were classified into none of, one, and >= two PES, respectively. On multiple logistic regression analysis, D-dimer levels were related to one PES (odds ratio [OR]: 9.01; 95% confidence interval [CI]: 1.00-81.51; P = 0.050) and PES >= two (OR: 9.76; 95% CI: 1.07-88.97; P = 0.043). Right-to-left shunt (RLS) with deep venous thrombosis (DVT)(OR: 13.94; 95% CI: 1.77-109.99; P = 0.012) and without DVT (OR: 3.90; 95% CI: 1.20-12.70; P = 0.024) were associated with elevation of D-dimer. Conclusions D-dimer levels were higher in patients with PES. Among PES, RLS, with and without DVT, were associated with increase of D-dimer in ESUS.