Dynamic contrast-enhanced MRI before and after transcatheter occlusion of patent foramen ovale

The purpose of this study was threefold: to evaluate the diagnostic accuracy of dynamic contrast-enhanced MRI compared with transesophageal echocardiography (TEE) in the detection of patent foramen ovale (PFO) and of residual shunts after occlusion of PFO, to define cutoff values for semiquantitative analysis of signal intensity-time curves, and to compare the diagnostic accuracy of visual detection with that of semiquantitative analysis.

Forty-three patients (18 women, 25 men; mean age, 51 +/- 14 years) who underwent TEE for suspicion of PFO (n = 26, 19 patients with and seven without PFO) or for routine assessment for residual shunt after transcatheter PFO occlusion (n = 17, nine patients with and eight without residual shunt) were consecutively enrolled to undergo contrast-enhanced MRI (saturation recovery steady-state free precession sequence). The images were analyzed both visually and semiquantitatively for arrival of contrast agent in the left atrium before arrival in the pulmonary veins during a Valsalva maneuver. TEE results were used as the clinical reference.

With an area under the signal intensity-time curve of 0.85, height of the first initial peak in signal intensity in the left atrium proved to be the best discriminator in right-to-left shunt detection. For a cutoff value of 129% (from baseline signal intensity) for this parameter, sensitivity and specificity were 90% (17/19) and 100% (7/7) in patients without PFO devices and 56% (5/9) and 88% (7/8) in patients with PFO devices. The diagnostic accuracy of both visual assessment and semiquantitative analysis was consistently superior before PFO device implantation than after device implantation. The diagnostic accuracy of visual shunt assessment was better than that of semiquantitative shunt assessment in patients with PFO occluders (sensitivity, 67% [6/9] correctly diagnosed; specificity, 88% [7/8]) and those without PFO occluders (sensitivity, 95% [18/19]; specificity, 100% [7/7]).

At present, MRI cannot replace TEE for the exclusion of potential embolic sources, such as thrombus in the left atrial appendage. However, MRI can be an attractive alternative noninvasive technique if TEE is technically unfeasible or is declined by patients.