Association of plasma D-dimer and B-type natriuretic peptide levels with acute ischemic stroke mechanism, severity, and outcome

Background: We sought to evaluate the utility of D-dimer and BNP in relation to stroke mechanism, severity, and outcome in acute ischemic stroke. Methods: We retrospectively analyzed 1335 patients from our prospective stroke registry. D-dimer and BNP levels were compared among stroke subtypes. The associations of D-dimer and BNP with initial NIHSS, early neurological deterioration (END), and 3-month modified Rankin Scale (mRS) were investigated. Results: D-dimer levels were significantly different among stroke subtypes. Cardioembolism (CE) had the highest median D-dimer levels (1.03 ㎍/mL (0.51-2.54)) followed by undetermined negative (0.66 (0.37-1.26)), large artery atherosclerosis (LAA) (0.59 (0.38-1.14)), and small vessel occlusion (SVO) (0.45 (0.30-0.76)). Among the LAA, LAA steno-occlusion subgroup had higher D-dimer levels than LAA branch occlusive disease subgroup (0.66 (0.41-1.32) vs. 0.45 (0.32-0.66), p<0.001). BNP levels were higher in CE than all the other subtypes (233.9 pg/mL (103.5-478.3) vs. 30.9 (14.1-70.2), p<0.001) and the C statistic of BNP for predicting CE was 0.889. Multivariable regression analysis showed that NIHSS was associated with log D-dimer (B=3.556, p<0.001) and log BNP (B=0.799, p=0.010) along with short onset-to-sample time, atrial fibrillation, and no history of coronary disease. END was associated with D-dimer (OR, 1.048; 95% CI, 1.008-1.089; p=0.018), short onset-to-sample time, NIHSS, diabetes mellitus, LAA, and SVO. The mRS was associated with D-dimer (OR 1.093, 95% CI, 1.037-1.151; p<0.001), age, NIHSS, history of previous stroke, LAA, and SVO. Conclusion: D-dimer levels were different among stroke subtypes and associated with stroke severity, END, and mRS. BNP was useful for predicting CE, and associated with stroke severity but not independently with END or mRS.