Importance of saline contrast transthoracic echocardiography for evaluating large right-to-left shunt in patent foramen ovale associated with cryptogenic stroke

• Published in: The International Journal of Cardiovascular Imaging Vol. 38; no. 3; pp. 515 - 520
• Main Authors: Takaya, Yoichi, Nakayama, Rie, Akagi, Teiji, Yokohama, Fumi, Miki, Takashi, Nakagawa, Koji, Toh, Norihisa, Ito, Hiroshi
• Format: Journal Article
• Language: English
• Published: Dordrecht Springer Netherlands 01-03-2022; Springer Nature B.V
• Subjects: Accuracy; Cardiac Imaging; Cardiology; Catheterization; Echocardiography; Echocardiography, Transesophageal; Evaluation; Foramen Ovale, Patent - complications; Foramen Ovale, Patent - diagnostic imaging; Foramen Ovale, Patent - therapy; Humans; Imaging; Ischemic Stroke; Medicine; Medicine & Public Health; Original Paper; Predictive Value of Tests; Radiology; Stroke; Stroke - complications; Stroke - etiology; Cryptogenic stroke; Patent foramen ovale; Right-to-left shunt; Transthoracic echocardiography
• ISSN: 1569-5794; 1573-0743; 1875-8312
• DOI: 10.1007/s10554-021-02418-6

Transcatheter closure of patent foramen ovale (PFO) is an effective therapy for preventing recurrent stroke in very specific patient cohorts, such as cryptogenic stroke (CS). The identification of high-risk PFO, which is more likely to be linked to CS, is essential. This study aimed to assess the accuracy of saline contrast transthoracic echocardiography (TTE) for evaluating large right-to-left (RL) shunt. We enrolled 119 patients with or without CS who were confirmed to have PFO by transesophageal echocardiography (TEE) or catheterization. The severity of RL shunt evaluated by TTE and TEE was classified as follows: small (< 10 microbubbles), moderate (10–20 microbubbles), and large (> 20 microbubbles). With TTE, large RL shunt was observed in 94 (79%) of 119 patients, including 66 of 74 with CS and 28 of 45 without CS. With TEE, large RL shunt was observed in 33 (28 %) patients, including 26 with CS and 7 without CS. TTE showed large RL shunt more frequently than TEE (p < 0.01). Large RL shunt evaluated by TTE had a sensitivity of 89 % and an accuracy of 70 % for the association with CS, whereas large RL shunt evaluated by TEE had a sensitivity of 35% and an accuracy of 56 %. Accuracy was significantly greater in TTE than in TEE (p = 0.02). In conclusion, TTE identified large RL shunt associated with CS with higher sensitivity and accuracy compared to TEE. Our findings suggest that the decision for device closure should be made based on the severity of RL shunt by TTE.