Starting with the Cox-Maze procedure, nonconcomitant AF ablation has evolved toward a minimally invasive video-assisted procedure with dedicated ablation tools using cryothermia and/or radiofrequency (RF) energy to isolate the PVs and create linear lesions without opening the heart. These ablation tools that replaced the cut-and-sew technique cannot always guarantee durable transmural lesions. Furthermore, linear lesions such as the isthmus lines cannot be made solely from the epicardium, and proving epicardial bidirectional block can be challenging. Therefore, our Maastricht group introduced a novel hybrid procedure for patients with paroxysmal AF (PAF) and persistent AF based on a combined simultaneous thoracoscopic surgical and transvenous catheter procedure. These two complementary techniques performed in combination bear the potential of treating AF on a long-term basis with a single off-pump procedure. The workflow of the stepwise approach was designed to overcome the limitations of both thoracoscopy and catheter ablation, the absence of mapping in surgery, and the difficulty of consistently achieving transmural lesions with catheter ablation. During a stepwise ablation, structures contributing to the initiation and maintenance of AF (PVs, left atrial roof, left atrial posterior wall, mitral isthmus and cavotricuspid isthmus [CVI]) are sequentially targeted as needed.

Our hybrid procedure was originally performed using bilateral thoracoscopy and consisted of a minimal lesion set of bilateral PVI made with a bipolar RF clamp. The end point for the PVI was entrance and exit block. In the case of sinus rhythm after PVI, reinduction of AF was attempted by pacing, and if AF was not inducible, isoproterenol was infused. If AF had not been terminated or was still inducible, linear lesions were added. First, a roof line was performed followed by a floor line, left atrial isthmus line, right atrial intercaval line, superior caval vein isolation, and/or a CVI line if needed. In 23% of the patients, the epicardial lesions were not transmural, and an endocardial touch-up was necessary, proving the importance of simultaneous endocardial mapping and ablation. A single-procedure success rate (sinus rhythm without antiarrhythmic drugs and/or redo procedure) of 79% at 1 year for PAF (11 of 14 patients) and 90% for persistent AF (9 of 10 patients), and an overall single-procedure success rate of 83% at 1 year was obtained. Our results were published in 2012 along with an editorial by Calkins discussing the potential limitations of this novel approach. In essence, Calkins discussed the need for expertise of both teams in the same center and the challenging logistics of organizing a simultaneous joint procedure. The need for expertise is still a challenge, but the logistics have been solved with the creation of dedicated hybrid rooms used for the percutaneous treatment of patients with structural heart disease.

Based on this need for surgical expertise and the better understanding of the potentials of the hybrid procedure, it seemed important to minimize the surgical technique further by limiting the approach to one side of the chest. This made the procedure more teachable and increased our patient referrals. Therefore, we developed a left-side-only thoracoscopic approach to facilitate surgery, to avoid right-sided complications and reduce postoperative pain. We had previous experience with a right-side-only approach, but many cardiac surgeons do not like to close the left atrial appendage (LAA) through the transverse sinus, and LAA occlusion is a bit more difficult to control with that approach.

When approaching the heart through the left chest, the camera has to be positioned inside the transverse and oblique sinuses to dissect the pericardial folds between the superior vena cava and the right superior PV (RSPV), between the inferior vena cava and the right inferior PV (RIPV), and the roof and posterior wall of the left atrium. Although this results in a reduced working space compared to the right-sided approach, the dissection is more straightforward because important cardiac structures are easily visualized, and the surgeon works toward the pericardium (safe space) instead of toward the heart. After this dissection, the connecting roof and inferior lines are created with the Coolrail device (Atricure, Inc.) as well as a line medial to the right PVs inside the oblique sinus. Performing these lines also melts the fat on the pericardial reflection between the transverse and oblique sinus and therefore facilitates the passage of the light dissector around the left PVs. When creating these linear ablations, it is mandatory to follow three steps to avoid the potential for an atrioesophageal fistula. The first step is to identify the anatomic position of the esophagus by looking for the tip of the transesophageal echocardiograpy (TEE) scope bulging through the pericardium. The second step is to push the ablated tissue away from the posterior pericardium. The third step is to flush the pericardial space with saline for at least 30 seconds after the ablation and before removing the ablation device. Bipolar RF clamp ablation of the left PVs is performed medial to the ligament of Marshall and includes the coumadin ridge to exclude these structures. The remaining ganglionic plexi are then ablated. Next, we carefully position the clamp in the oblique sinus and guide it toward the opened reflection at the base of the RIPV and over the right PVs, and the ablation is performed. Finally, LAA occlusion (LAAO) is performed under echocardiographic guidance.

The most difficult part of this left-side-only approach is completely isolating the RIPV. Although we perform this surgically with a bipolar RF clamp, we do not recommend that this step be performed unless the surgeon is very experienced. Rather, we advise that the lateral part of the RIPV be isolated by the electrophysiologist (EP) with the endocardial catheter. This is a safe and straightforward alternative because the surgeon has already created a linear lesion on the inferior border of the RIPV, the superior border of the RSPV, and the medial side of the PVs in the oblique sinus. Thus only a lesion around the lateral side of the right PVs remains to be created by the EP for completion of the PV isolation. The left lung is insufflated, and the EP continues with endocardial confirmation of transmurality of the epicardial lesion set and, if needed, an endocardial touch-up. Then areas involved in the maintenance of AF, left-sided atrial tachycardias, and flutter are mapped and ablated.

The challenge of any epicardial ablation tool on the beating heart is the creation of reliable, safe, and long-lasting transmural lesions. The quality of each lesion can be impacted by the presence of epicardial fat, thickness of the tissues, histologic variations, and cardiac output. During surgery, if no intraoperative testing is performed to confirm lesion integrity and the efficacy of the lesion set, it is likely that the procedure will be less successful for the long term (see Chapter 20). Performing endocardial mapping after an epicardial ablation immediately provides valuable information regarding the efficacy of the ablation tool used and whether the number of applications (amount of energy delivered) was sufficient. Apart from the bipolar RF clamping devices and properly used cryoprobes, none of the other tools have proven to provide consistent, reliable transmural lesions. This problem was emphasized in the C atheter A blation vs. Thoracoscopic S urgical A blation in Long-­standing Persistent A trial F ibrillation (CASA-AF) randomized controlled trial (RCT). The single-center CASA-AF pilot study in patients with long-standing persistent AF (LSpAF; 26 surgical ablations, 25 catheter ablations) showed a clear superiority of the thoracoscopic approach compared to the percutaneous approach. To ensure robust assessment of conduction block for surgical ablation lesions, block was confirmed by EP-guided testing. Of the 22 patients who had both surgical and EP catheter testing, 3 patients were found to have remaining PV connections, and 2 patients had conduction into the box lesion, all of which were deemed isolated after surgical pen testing. Therefore, EP catheter testing prompted further ablation to achieve conduction block in 23% (5 of 22) of patients. Haldar et al. concluded, “In LSPAF, meticulous EP guided thoracoscopic SA [surgical ablation] as a first line strategy can provide excellent single procedure success rates as compared to catheter ablation (CA), but there is an increased upfront risk of non-fatal complications.” The follow-up randomized controlled trial (RCT) did not achieve similar outcomes for the surgical group, and this time the authors concluded that “Single procedure thoracoscopic SA [surgical ablation] is not superior to CA in treating LSPAF.” So, what changed? Analyzing the article, apart from the multicenter RCT, the only major difference between the two studies was the surgical device used for the creation of the connecting lesions (roof and floor line), which was the Atricure Coolrail device for the first study and the Atricure Multi-Polar Pen (MLP) for the second. In addition, the results of the EP-guided testing of surgical ablation conduction block and the need for further ablation to achieve conduction block were not published. Our group has studied both devices in our hybrid procedures. Whereas the Coolrail device resulted in more than 90% acute isolation, the MLP device always needed an endocardial touch-up to complete the lesion. The nontransmurality of the surgical lesion set easily explains the difference in outcome of the two studies. A hybrid procedure would have overcome this problem and proven the superiority of the combined procedure as demonstrated in the ( H ybrid Thoracoscopic Surgical and Transvenous Catheter A blation Versus T ransvenous C atheter A blation in P ersistent and Longstanding Persistent A trial F ibrillation (HART-CAP AF) randomized controlled trial.