D-dimer for prediction of long-term outcome in cryptogenic stroke patients with patent foramen ovale.

Patent foramen ovale (PFO) is a potential cause of cryptogenic stroke, given the possibility of paradoxical embolism from venous to systemic circulation. D-dimer level is used to screen venous thrombosis. We investigated the risk of embolism and mortality according to the presence of PFO and D-dimer levels in cryptogenic stroke patients. A total of 570 first-ever cryptogenic stroke patients who underwent transesophageal echocardiography were included in this study. D-dimer was assessed using latex agglutination assay during admission. The association of long-term outcomes with the presence of PFO and D-dimer levels was investigated. PFO was detected in 241 patients (42.3 %). During a mean 34.0 ± 22.8 months of follow-up, all-cause death occurred in 58 (10.2 %) patients, ischaemic stroke in 33 (5.8 %), and pulmonary thromboembolism in 6 (1.1 %). Multivariate Cox regression analysis showed that a D-dimer level of > 1,000 ng/ml was an independent predictor for recurrent ischaemic stroke in patients with PFO (hazard ratio 5.341, 95 % confidence interval 1.648-17.309, p=0.005), but not in those without PFO. However, in patients without PFO, a D-dimer level of > 1,000 ng/ml was independently related with all-cause mortality. The risk of pulmonary thromboembolism tended to be high in patients with high D-dimer levels, regardless of PFO. Elevated D-dimer levels in cryptogenic stroke were predictive of the long-term outcome, which differed according to the presence of PFO. The coexistence of PFO and a high D-dimer level increased the risk of recurrent ischaemic stroke. The D-dimer test in cryptogenic stroke patients may be useful for predicting outcomes and deciding treatment strategy.