Role of cine-fluoroscopy, transthoracic, and transesophageal echocardiography in patients with suspected prosthetic heart valve thrombosis.

Prosthetic heart valve thrombosis (PVT) is a rare but potentially life-threatening complication of heart valve replacement. An effective, quick, and easy diagnostic method is highly desirable. We evaluated the diagnostic efficacy of cine-fluoroscopy (CF), transthoracic (TTE), and transesophageal (TEE) echocardiography in 82 consecutive patients with mechanical valves and suspected PVT. Criteria for PVT were: leaflet(s) motion restriction at CF, increased Doppler gradients at TTE, and evidence of thrombi at TEE. Patients were divided in 4 groups (A, B, C, and D) according to results of CF and TTE. Group A was composed of 24 patients with positive CF and TTE. Thrombi were detected by TEE in all cases, suggesting that when both are positive, CF and TTE correctly identified PVT in all patients so that TEE may be deferred. Group B was composed of 12 patients with positive CF and negative TTE; TEE showed PVT in 4 patients (33%). These patients had very slight leaflet motion restriction as in the case of initial PVT. This suggests that CF compared with Doppler may identify patients with "hemodynamically significant" PVT. The remaining 8 patients in this group had monocuspid prostheses with negative TEE, suggesting that abnormal leaflet motion at CF may be due to functional changes. Therefore, TEE should always be performed in case of monocuspid prostheses with isolated CF abnormalities. Group C was composed of 18 asymptomatic patients with small-sized aortic prostheses and very high Doppler gradients on routine TTE. CF showed normal leaflet motion and TEE ruled out PVT in all cases outlining the diagnostic role of CF in this particular subset. Finally, group D was composed of 28 patients with negative CF and TTE. TEE did not show thrombi in 24 of 28 patients (86%), confirming that, when both yield negative results, CF and TTE are reliable methods to rule out valve thrombosis in most cases. However, in 4 of 28 patients (14%) TEE showed "nonobstructive" prosthetic thrombosis: these patients had mitral prostheses, chronic atrial fibrillation, and 3 of 4 had systemic embolisms. Thus, TEE should be performed in selected patients despite negative CF and TTE results. Sensitivity, specificity, and positive and negative predictive values were 87%, 78%, 80%, and 91% for CF and 75%, 64%, 57%, and 78% for TTE, respectively. CF and TTE correctly identified PVT in 70 of 82 patients (85%). TEE was actually required in 15% of the cases. Thus, CF and TTE are quick, effective, and complementary diagnostic tools to diagnose PVT in most patients. TEE still remains the gold standard technique in selected cases.