Intraprocedure visualization of the esophagus using interventional C-arm CT as guidance for left atrial radiofrequency ablation

Abstract

Rationale and objectives: During radiofrequency catheter ablation for atrial fibrillation, the esophagus is at risk for thermal injury. In this study, C-arm computed tomography (CT) was compared to clinical CT, without the administration of oral contrast, to visualize the esophagus and its relationship to the left atrium and the ostia of the pulmonary veins (PVs) during the radiofrequency ablation procedure.

Materials and methods: Sixteen subjects underwent both cardiac clinical CT and C-arm CT. Computed tomographic scans were performed on a multidetector scanner using a standard electrocardiographically gated protocol. C-arm computed tomographic scans were obtained using either a multisweep protocol with retrospective electrocardiographic gating or a non-gated single-sweep protocol. C-arm and clinical computed tomographic scans were analyzed in a random order and then compared for the following criteria: (1) visualization of the esophagus (yes or no), (2) relationship of esophageal position to the four PVs, and (3) direct contact or absence of a fat pad between the esophagus and the PV antrum.

Results: The esophagus was identified in all C-arm and clinical computed tomographic scans. In four cases, orthogonal planes were needed on C-arm CT (inferior PV level). In six patients, the esophageal location on C-arm CT was different from that on CT. Direct contact was reported in 19 of 64 of the segments (30%) examined on CT and in 26 of 64 (41%) on C-arm CT. In five of 64 segments (8%), C-arm CT overestimated a direct contact of the esophagus to the left atrium.

Conclusions: C-arm computed tomographic image quality without the administration of oral contrast agents was shown to be sufficient for visualization of the esophagus location during a radiofrequency catheter ablation procedure for atrial fibrillation.

Fig. 1. Visualization of the esophagus

On this C-arm CT reconstructed image of the heart, the cross-lines are centered on the esophagus on the 3 spatial planes (axial, sagittal and coronal). The availability of 3D planes helped confirm the findings on the axial plane.

Fig. 2. a–c: Example of assignment of the esophagus location to a predefined left atrium area

These 3 axial CT images of the left atrium show the ostia of the LSPV (fig. 2a), RSPV (fig. 2b), LIPV and RIPV (fig.2c). In this particular case, the esophagus location corresponded to area 5 at the axial level of the LSPV, to area 3 at the level of the RSPV, and to area 4 at the inferior PV level. The structures delineated by a triangle represent the presence of a fat pad.

Fig. 3. a, b: ECG-gated and non-gated C-arm CT scans

For the purpose of esophagus visualization, C-arm CT ECG-gated (fig. 3a) and non-gated (fig. 3b) images did not show significant differences, as shown in these two different patients.

Fig. 4. a, b: Temporal changes

Axial CT (fig. 4a) and C-arm CT (fig.4b) images of the same patient show a change in the esophagus position from the clinical CT to the time of the procedure.

Fig. 5. a–d: Examples of visualization of the esophagus

Axial CT (fig. 5a) and C-arm CT (fig.5b) images of the same patient show excellent visualization of the esophagus, comparable to the corresponding CT image. In a different patient, axial CT (fig. 5c) and C-arm CT (fig.5d) images show how even in images affected by severe artifacts (worst case), the visualization of the esophagus was still possible.