Introduction

We present a new, easily applicable approach for the guidance of cryo balloon pulmonary vein isolation (PVI) procedures that uses the combination of a 3D mapping system image integration module and CT-derived anatomy. The aim of this study was to investigate (1) an alternative use for an established radiofrequency image integration module for cryo procedures; (2) a guidance technology for cryo PVI based on integrated CT anatomy; and (3) its clinical impact.

Methods

Computer tomographic LA-angiography was performed in n=50 consecutive patients prior to a cryo PVI procedure, and a 3D-reconstruction of the cardiac anatomy was segmented. N=25 patients were treated using conventional fluoroscopy; n=25 patients were treated using the image integration technique. In the image integration group, the CARTO3 UNIVU (Biosense Webster, USA) module was used for image integration of 3D anatomy and fluoroscopic imaging; the 3D-reconstructed cardiac anatomy was congruently aligned to its fluoroscopic counterparts in an initial registration process. Transseptal puncture and cryo PVI were guided by 3D-overlay imaging. During the cryo PVI procedure, the integrated image of the LA and the PVs served as a guidance map for catheter navigation and balloon catheter placement. In order to increase clarity, left and right PVs were visualized separately and superior and inferior PVs were displayed in distinct, separate colors. For each targeted PV, the integrated LA/PV anatomy was used to guide the balloon catheter and verify correct balloon positioning. By selective positioning of the Achieve mapping catheter in different segmentations of each PV’s branches, the contact between the balloon sphere and PV antrum was optimized if necessary (Figure 1). Beyond LA and PV anatomy, further important anatomic landmarks were automatically co-registered and could be visualized during procedures: the aortic root, which would not be clearly identifiable during transseptal puncture with conventional fluoroscopic guidance only and the right phrenic nerve (Figure 2). Imaging and procedural data were stored for later analysis.









Results

In both patient groups, all procedures were feasible and no procedure related complications were observed. In the 3D group, image integration was successfully performed in all patients and used to guide cryo PVI procedures. Relevant baseline characteristics of both patient groups were not significantly different. Procedure time was 116.3 ± 29.0 min in the conventional group vs. 101.2 ± 20.9 min in the 3D group (p=0.04), and fluoroscopy time was 31.7 ± 11.7 min in the conventional group vs. 20.1 ± 7.9 min in the 3D group (p<0.001) (Fig. 6). The dose area product was 18530.5 ± 11025.1 μGym² in the conventional group vs. 5084.3 ± 3306.8 μGym² in the 3D group (p<0.001). In the technical image integration accuracy analysis, the offsets between fluoroscopic and integrated anatomy contours were determined to be lower than 2mm in all imaging planes. The average time needed to register the CT image integration on fluoroscopic imaging was 37.4±10.3 sec.

Conclusions
3D image integration guidance for cryo balloon PVI procedures is feasible, accurate, and safe. It may significantly reduce fluoroscopy and procedure time. This approach combines 3D anatomic representations (as previously demonstrated in RF ablation using mapping systems) with well-known fluoroscopic images to guide cryo PVI.