NYMC Faculty Publications

Dispelling the Saddle Pulmonary Embolism Myth (From a Comparison of Saddle versus Non-Saddle Pulmonary Embolism)

Author Type(s)

Resident/Fellow, Faculty

Journal Title

The American Journal of Cardiology

First Page


Last Page


Document Type


Publication Date




Second Department



In the setting of an acute pulmonary embolism (PE), there is often an assumed association between a saddle PE (SPE) and increased clinical severity. We aimed to determine the magnitude of SPE proximal pulmonary artery (PA) flow obstruction and its impact on right ventricular (RV) function in the setting of acute PE in a single-center series. From 2005 to 2022, patients with acute PE presenting with acute RV dysfunction requiring intervention were classified as SPE and non-SPE based on presenting computed tomography (CT) scans. SPE flow obstruction was determined by the ratio of the orthogonal cross-sectional surface area measurements of clot and native PA at the location of maximum clot burden in the right PA and left PA. Presenting RV function based on clinical and imaging parameters (CT and transthoracic echocardiography) were compared between SPE and non-SPE cohorts. A total of 174 patients were identified (SPE 92 [52.9%] and non-SPE 82 [47.1%]). Demographics and co-morbidities were similar. In patients with SPE, there was a mean 25.9% total flow obstruction (right PA 26.9% and left PA 25.5%). Non-SPE had greater clinical RV dysfunction on presentation as reflected by more high-risk PE (43.9% vs 26.1%, p = 0.01), need for venoarterial extracorporeal membrane oxygenation (21.9% vs 10.9%, p = 0.05), and more preoperative cardiopulmonary resuscitation (16.7% vs 7.8%, p = 0.08). RV:left ventricular ratio (CT and transthoracic echocardiography) and RV fractional area change were statistically similar between groups. In-hospital mortality was statistically similar between cohorts (4.9% non-SPE vs 2.1% SPE, p = 0.32). In conclusion, in a single-center series of patients with acute PE with RV dysfunction, SPE did not cause proximal flow-limiting obstruction. Non-SPE was associated with more clinical RV dysfunction than SPE. Thus, it should not be assumed that a non-SPE is a marker of patient stability.