Radiosurgery to Create Lesions in the Pulmonary Vein Antrum: Preliminary Results in a Porcine Model and Possible Implications for the Treatment of Atrial Fibrillation

Objectives:Radiosurgery has found its way into clinical routine practice for a variety of selected benign indications.Radiosurgery of the heart for potential treatment of cardiac arrhythmias has recently been under investigation, exploringthe feasibility of atrial flutter ablation in animals. The most common arrhythmia in humans is atrial fibrillation, whereelectrical isolation of the pulmonary veins has been established in clinical routine as a curative treatment. While invasivecatheter ablation carries significant risks, radiosurgery might be able to non-invasively induce lesions at the pulmonaryvein ostia comparable to catheter ablation to block the electrical veno-atrial connections. We present first results of aporcine feasibility and dose escalation study to create lesions in the left atrial - pulmonary vein junction usingradiosurgery.

Methods:The study was conducted in 10 adult Göttingen-Mini-Pigs (weight 40-60 kg). Two animals were used forbaseline evaluation without radiation exposure and 8 were treated with single fraction radiosurgery. The animalsunderwent general anesthesia and ventilation and were fixated in a tight vacuum bag during treatment. Cardiac MRI wasperformed before radiosurgery to assess heart function, pulmonary vein anatomy and myocardial scarring.Electrophysiological voltage mapping was performed in the left atrium and the target vein using the NavX Ensite System.A homogeneous dose ranging from 17.5 Gy to 35 Gy was applied circumferentially to the upper right pulmonary veinantrum. For compensation of cardiac and respiratory motion during treatment we used margins derived from 4D cardiacCT during inspiration and expiration. Radiation was delivered with a conventional linear accelerator (Varian DHX 2100)using a stereotactic frame and ConeBeam CT for setup. Six months after radiosurgery, MRI and electrophysiologicalmapping were repeated and results were compared to pretreatment findings. Animals were sacrificed and a histologicalexamination was performed to assess radiation related lesions at the target vein antrum and potential collateral damageto adjacent structures such as the bronchus or the aorta.

Results:At baseline all animals had normal heart function assessed by MRI. Voltage mapping consistently showedelectrical potentials in the upper right pulmonary vein. Pacing the target vein prompted atrial excitation, thus provingveno-atrial electrical conduction in this model. 6 months after treatment MRI assessment demonstrated no adverseeffects on general heart function. A dose-dependent decrease of potential amplitude in the target area was noted atdoses above 30Gy though without complete electrical block. Histology revealed a dose-response-relationship of thetargeted area with fat tissue necrosis at doses above 20Gy and partial transmural scarring in the muscle tissue at dosesabove 30Gy. With the highest doses, small damages to the adjacent bronchial tree were noted as well as stenosis of asmaller sub branch of the targeted pulmonary vein.

Conclusions: This proof-of-principle-study demonstrates for the first time that radiosurgery with a conventional linearaccelerator is able to induce focal lesions in anatomically defined areas of the pulmonary vein antrum with a subsequentreduction in electrophysiological signals and the creation of scar. The preliminary results indicate that doses above 30 Gymay be necessary to achieve an electrical conduction block at the pulmonary vein antrum in normal pig hearts at 6months after treatment. Further studies are needed to prove complete pulmonary vein isolation with optimized radiationparameters.