ObjectivesThere are conflicting data concerning the prognostic significance of syncope in acute pulmonary embolism (PE). This study aimed to investigate the impact of syncope on clinical outcomes of acute PE, and determine the clinical phenotypes of PE patients with syncope and their correlation with prognosis. MethodsIn the ongoing, national, multicenter, registry study, the China pUlmonary thromboembolism REgistry Study (CURES) enrolling consecutive patients with acute PE, patients with and without syncope were investigated. Principal component analysis (PCA) was performed using nine variables relevant to syncope and PE, including age, sex, body mass index, history of cardiovascular disease, recent surgery or trauma, malignancy, pulse, systolic blood pressure, and respiratory rate. Patient classification was performed using cluster analysis based on the PCA-transformed data. The clinical presentation, disease severity and outcomes were compared among the phenotypes. ResultsIn 7,438 patients with acute PE, 777 (10.4%) had syncope, with younger age, more females and higher body mass index. Patients with syncope had higher frequency of precordial pain, palpitation, and elevated cardiac biomarkers, as well as higher D-Dimer level. In the syncope group, more patients had right ventricular/left ventricular ratio > 0.9 in ultrasonic cardiogram and these patients had higher estimated pulmonary arterial systolic pressure compared with patients without syncope. As the initial antithrombotic treatment, more patients with syncope received systemic thrombolysis. Despite a higher prevalence of hemodynamic instability (OR 7.626, 95% CI 2.960-19.644, P < 0.001), syncope did not increase in-hospital death. Principal component analysis revealed that four independent components accounted for 60.3% of variance. PE patients with syncope were classified into four phenotypes, in which patients with high pulse and respiratory rate had markedly higher all-cause mortality during admission. ConclusionSyncope was associated with hemodynamic instability and more application of thrombolysis, without increasing in-hospital deaths. Different clinical phenotypes existed in PE patients with syncope, which might be caused by various mechanisms and thus correlated with clinical outcomes.