Update on the Dual Epicardial Endocardial Persistent Atrial Fibrillation Study (Staged DEEP)

Interview by Jodie Elrod
Interview by Jodie Elrod

In this article EP Lab Digest® speaks with Jonathan Philpott, MD, director of the Sentara Atrial Fibrillation Surgery Program, and electrophysiologist Ian Woollett, MD of Cardiovascular Associates. Sentara Heart Hospital was the first in the world to begin performing surgeries in the Dual Epicardial Endocardial Persistent Atrial Fibrillation (AF) Study (Staged DEEP) feasibility trial.

Explain the purpose of the Dual Epicardial Endocardial Persistent Atrial Fibrillation Study (Staged DEEP). How is this trial different from the DEEP AF trial?

Philpott: The purpose is to pioneer a safe and effective treatment for patients with symptomatic persistent and longstanding persistent AF or non-paroxysmal AF (NPAF). Right now, there are really no good options for these patients, in terms of high durable success and freedom of AF while off AADs. Endocardial ablation alone has not yielded the results hoped for, and open-heart surgery is frequently an unpalatable option for most patients. NPAF represents the largest subgroup of AF patients and by far the most difficult to treat. Our chief aim was to take the Cox Maze IV, a lesion set that when performed precisely was known to yield high and durable success rates, and transition it to a minimally invasive and totally thorascopic platform. This would begin with epicardial ablation performed on the beating heart, followed immediately in the same room with endocardial mapping and ablation to complete the lesions initiated epicardially. Then we would test both the epicardial and endocardial lesions for block. The concept was independently arrived at at several centers, but was first pioneered in Holland where it showed significantly better success rates than catheter ablation alone. The DEEP AF feasibility trial was initiated to explore the concept. It yielded high success rates in our center, but logistically was very difficult to execute with surgery occurring first, followed immediately by endocardial mapping and ablation. The anesthesia times for these simultaneous cases were excessive, as it required an EP hybrid OR.

While technically successful, it did not appear that the simultaneous platform was going to be something that would be easily adopted across the nation from a resource, work flow, and logistics standpoint. Thus, we sought to simplify the procedure by separating the two procedures by several days during the same hospitalization. This allows the patient to recover between cases. By doing this, a center would not need an EP hybrid OR to carry out the procedure. Instead, their usual work place could suffice, with the operation occurring in a regular cardiac surgical OR, followed by the second procedure in a typical EP lab configured for mapping and ablation of AF. This would open it up widely across the country and improve successful adoption. Secondly, we theorized that by shortening the anesthesia times, the patient would have a much faster recovery. The lesion set, however, remains exactly the same and is very robust. It includes bilateral mapping and ablation for active ganglionic plexi followed by bilateral antral pulmonary vein isolation with the AtriCure Synergy Bipolar RF clamp, connecting lesions across the dome or roof of the LA and the floor created with bipolar epicardial RF arrays, connecting lesions to the LAA and exclusion of the appendage with an AtriCure AtriClip, and finally the initiation of the mitral valve isthmus lesion which is started from the left PV and carried up to the coronary sinus (but not across the annulus, as this will be completed endocardially). On the right atrial side, a transverse SVC ablation is created just above the SA node with the AtriCure Synergy Bipolar clamp for SVC trigger suppression; the connecting lesion then created runs from the inferior vena cava up and across the medial RA free wall and on to the SVC, where it joins with the transverse bipolar line. The lesion set is then finished endocardially with a tricuspid-caval flutter ablation line followed by a transseptal puncture and completion of the MV isthmus line, carrying the line initiated epicardially down and on to the MV annulus.

At this point, we have enrolled the first two patients in the trial. The trial is limited to 30 patients overall, with each site capping out at 10 as a maximum. Although it is still early, we have already noticed a major difference between these first two patients and our experience with DEEP AF. The patients’ time of recovery from the surgical procedure is significantly faster, which was a major surprise. The dual cases are averaging 3–4 hours of initial surgical time instead of 8–10 combined at the single setting, and the patients are being extubated much easier and with significantly less discomfort. Admittedly, we may have just gotten lucky with the first two, but the difference so far has been significant. In DEEP AF, patients were in considerable discomfort following the procedure. At that time, we were also using larger ports, but I also suspect that the duration of the anesthesia, which involved a large double-lumen endotracheal tube, contributed significantly more to the longer recovery time than we initially expected. One would assume that by separating the anesthesia times that the combined anesthesia would be similar or perhaps even slightly longer with the staged approach; however, I think what we are seeing is that by keeping the two much shorter with the recovery in the middle, it has significantly accelerated the overall speed of recovery from each. We will need more time to verify the current findings.

Woollett: We have had great success by approaching this in a collaborative rather then a competitive fashion, and as a result have been able to focus on what each specialty (EP and CT surgery) can do to cover the weaknesses of the other. The surgical approach allows for robust lesion sets, but is weak in mapping and is unable to reliably complete a mitral valve isthmus ablation line. Without this line, a very high percentage of the patients wind up with very difficult-to-treat left atrial flutter, and this has been one of the primary weaknesses of the stand-alone MAZE procedure. The EP catheter-based approach allows for meticulous testing of block to the same standards used in the EP lab and also for creation of lesions in areas that cannot be done surgically (tricuspid and mitral valve isthmus lines).

As a result, with the DEEP 1 trial (concurrent surgical and catheter ablations), we were able to achieve a success of greater then 90% in patients with longstanding persistent AF that had been refractory to multiple other therapies (often including previous ablations). We were very pleased with this, but there were concerns about the excessive procedure and anesthesia times, so a decision was made to attempt staging the procedures to hopefully minimize the procedural complications and speed recovery. We felt strongly that both procedures should be done during the same hospitalization, as we were concerned about refractory atrial flutters if the lesion sets (especially the mitral valve isthmus line) were not completed and tested in a rigorous fashion.

Discuss the first cases enrolled.

Philpott: Our two patients are the first to be enrolled, but multiple other centers are preparing to enroll. Stanford Medical Center was set to do their first case, the third in the study. In terms of overall experience, we have done approximately 30 simultaneous hybrid procedures (non staged) in our EP hybrid OR. The last two were the first in a regular OR and staged.

The first case in Staged DEEP was a 70-year-old retired RN with highly symptomatic longstanding persistent AF for four years. She had failed four prior cardioversions and medical management. When cardioverted she would briefly hold sinus rhythm for about a day at a time, and would note that she felt dramatically better in terms of overall energy level, stamina, and exercise ability. She was initially considering a Cox Maze IV, which we were also discussing with her. Ultimately, she elected the hybrid staged approach, as she was worried about the sternotomy scar showing when she would wear a dress or lower neckline apparel. She had competent mitral and tricuspid valves, and moderate dilatation of her left atrium (4.5 cm), a normal ejection fraction, and no coronary artery disease. She had previously undergone an endocardial ablation for an AV nodal reentry tachycardia in 2001. Her post-operative course was uneventful. The case took about three hours to perform; she was extubated in the OR and was observed in the ICU for 24 hours. She had minimal discomfort. At the procedure, bilateral chest drains are placed as the pericardium is opened on both sides. These were removed on the third day. On POD 4, she then completed her endocardial ablation per Dr. Woollett, and was discharged shortly afterward in normal sinus rhythm.

The second Staged DEEP case was a 68-year-old male with longstanding persistent AF with a non-ischemic cardiomyopathy, which had been as low as 20% and was 30% at the time of the procedure. He had a four-year history of AF and was consistently in uninterrupted AF for two years. He had normal coronary arteries, no significant valvular pathology, and an LA size of 4.2 cm. The first stage (thorascopic/epicardial) portion occurred first and took about three hours to perform. He was extubated with minimal discomfort, and was scheduled to undergo the second stage (endocardial) four days later by Dr. Venkat Iyer (partners with Dr. Woollett).

Tell us more about the hybrid procedure being done, including use of the AtriCure Bipolar System.

Philpott: The patient is taken to the OR where general endotracheal anesthesia is induced via a double-lumen endotracheal tube, and short neck lines are introduced (the SVC needs to be clear of catheters as it will be ablated). Cardiac perfusion and a heart-lung machine are prepped and on standby during the case. The chest is prepped and draped, and the right lung is deflated using the double-lumen endotracheal tube. A 5 mm port is introduced in the midaxillary line, and CO2 is insufflated into the chest to help collapse the right lung and also to move the diaphragm inferiorly to improve visualization of the pericardium. A 5 mm high-definition thorascopic camera is then introduced and the pericardium visualized. Two additional 5 mm ports are then introduced on either side of the camera port, and the diaphragm is opened several centimeters above the phrenic nerve. The pericardium is retracted with several stay sutures posteriorly to provide good visualization down to the posterior pericardium, where the right superior and inferior pulmonary veins can be well visualized. A path is opened under the SVC to gain access to the transverse sinus superiorly, and the oblique sinus is opened inferiorly. From these two approaches, the camera is advanced from the right side over to the left above and below to examine the left veins. These paths will be used to create the connecting lesions across the roof and floor of the LA. Testing is now carried out on the pulmonary veins to confirm entrance conduction. If the patient is in sinus rhythm, pacing is also carried out to confirm exit conduction. Mapping for active ganglionic plexi is then carried out with high-frequency stimulation working from the superior dome across both the superior and inferior veins both medially and laterally, and then inferiorly beneath the inferior vein. Active plexi are ablated with an epicardial bipolar device and then are retested. If they remain persistently active, they are ablated again and the process is repeated until no further activity can be replicated.

The groove between the left and right atrium is now fully developed. Using a lighted dissector introduced inferiorly and posteriorly, a tape is passed around the right-sided veins. This tape is attached to a small red catheter, which attaches to the bottom jaw of the minimally invasive AtriCure Synergy Bipolar RF clamp. Using the red catheter to guide the bottom jaw, the Synergy clamp is guided into position onto the antrum of the left atrium, just medial to the junction of the right pulmonary veins where it is clamped and fired. It is then repositioned and fired again for a total of five applications. The clamp is then withdrawn and the pulmonary veins are tested for entrance block (and rarely, for exit block if the patient is in sinus rhythm). Connecting lesions are then created from this PVI line across the floor and the roof of the LA with a linear bipolary RF array.

On the right atrium, the AtriCure Synergy Bipolar clamp is then used to create a transverse SVC line to isolate SVC triggers above the SA node. A connecting lesion is then created on the lateral aspect of the right atrium traveling from the transverse SVC ablation down and across the medial RA free wall and onto the IVC with the linear RF bipolary array devices (one short/one long as needed). This completes the right side. The pericardium is loosely closed, the right lung is reinflated, and a small chest tube is inserted. The 5 mm port sites are now closed.

The left side is then initiated using the same 5 mm port technology and isolation of the left lung to visualize the left pericardium. The left side is opened under the phrenic nerve, and the roof and floor connecting lesions on the dome and floor of the LA that had been started from the right side are now visualized. Baseline testing for conduction is now made on the pulmonary veins as was done on the right. Mapping for active GPs is now carried out just as on the right with immediate ablation of any active GP sites. The ligament of Marshall is divided and the left veins encircled again using a lighted dissector with a tape attached that is used to a second minimally invasive Synergy clamp (there is one designed for the right side and one designed for the left). This is guided into position around the antrum of the left atrium just medial to the junction of the left pulmonary veins, where it is clamped into position and fired. It is then repositioned, clamped and fired, and the process is repeated for a total of five applications. Testing is then carried out to confirm entrance block (and exit block if the patient is in SR). The connecting lesions from the roof and floor are then extended to join with the left antral pulmonary vein isolation line, thus completing a posterior box. A connecting lesion is then created to the base of the LAA and a second one toward the MV annulus using epicardial linear bipolar devices. The end of the MV line is marked with a small hemoclip as a radiopaque marker to guide the forthcoming endocardial ablation. Then the LAA is excluded using an appropriately sized FDA-approved AtriClip closure device.

For the endocardial aspect, catheters are introduced into the right atrium and a right tricuspid-caval line is created. The right epicardial lines are tested. Typically a double transseptal puncture is then performed, and the right and left PVI and posterior box are tested. While rare, breaks in the epicardial lines do occur, particularly on the connecting lesions created with the non-clamp epicardial tools, and these are now completed endocardially. Finally, the MV isthmus lesion is extended down and on to the MV annulus, and testing for block of this line is completed. Isoproterenol challenge is then performed up to 20 mcg to try and induce AF.

How many patients are expected to be enrolled in the trial? What is the patient inclusion criteria?


  • 30 total
  • 10 per center
  • NPAF or persistent AF, and especially longstanding persistent AF
  • Atrial size less than 6 cm
  • EF greater than or equal to 30%
  • No significant untreated critical CAD or valvular heart disease
  • BMI less than 40

Which other centers will be participating in the Staged DEEP trial? When is the estimated trial completion date?

Philpott: The centers identified so far are Sentara Heart Hospital, Vanderbilt Heart Institute, Stanford Medical Center, the University of Brussels, and Academic Medical Center in Amsterdam. Penrose Medical Center in Colorado Springs is also hopefully coming on board.

Amsterdam and Brussels have high-volume centers and have just formalized their IRBs. When they come on line, the study enrollment is expected to accelerate dramatically. This makes predictions on the estimated trial completion date difficult, but our hope is that we will finish enrollment in the next nine months to a year.

If the safety and efficacy endpoints are met, it is expected that a pivotal trial would follow, which would allow treatment-specific, head-to-head comparisons with an accepted treatment alternative. The elements of such a trial have not yet been defined.

What are the primary endpoints? When will first results from the trial be available?

Philpott: The primary endpoints are safety and efficiency at one year. This includes no atrial fibrillation or atrial tachycardia episodes greater than 30 seconds off antiarrhythmic medications and following the new HRS consensus statement.

Patients will be followed for one year to determine endpoints, and then results will follow shortly afterward — thus, approximately two years. However, there may be potential for interim reporting.

What do you find promising and beneficial about the dual epicardial/endocardial procedure?

Philpott: The team approach leverages the skills and strengths of both specialties while minimizing our weaknesses. The OR is a great place to create long lines of scar, but we are highly limited when it comes to really robust and sophisticated mapping and testing. Thus, we are weak at ablation focused to specific foci or singular targets of ectopy. On the other hand, making long lines of solid ablation is time consuming and difficult in the EP lab, but the cath lab is excellent in highly sophisticated EP testing and mapping, and particularly excellent at focused target ablation. It was a perfect match. One specialty’s strengths covers the other specialty’s weakness. The two strengths work together and the weak areas disappear. After working in the environment now for several years, I find that it is a perfect synergy.

The reality of that teamwork is a very robust lesion set. We are hopeful that we have produced a potentially major advancement in terms of a highly successful and durable treatment option for patients with symptomatic persistent and longstanding persistent AF. The number of patients who could potentially benefit is huge. In the end, that is the most satisfying part of this trial. These patients are frequently suffering and have lost a great deal of their quality of life. They are highly limited by fatigue and exhaustion with even minimal activity, or suffer from great anxiety from the palpitations. Secondly, there is the group who suffer from loss of their EF and begin to have heart failure. Offering these highly symptomatic patients a new option that has the promise to render them free from their AF has been the most promising aspect of the procedure.

The second important factor is the reduction of cost and repeat procedures. This has been observed in Europe and is one of the reasons the procedure has seen significant adoption there, as it typically works the first time. The number of patients returning for repeat ablation plummeted as a result of the hybrid procedure, which meant not only happier patients, but much better utilization of limited resources and budgets.

Tell us more about the work being done at the Sentara Atrial Fibrillation Surgery Program and the Sentara Heart Arrhythmia Center.

Philpott: Yes, we have almost completed a quality initiative of Combined EP and Surgical outcome reporting of success rates and complications for endocardial ablation, hybrid ablation, and open surgical ablation (concomitant and stand-alone) using the same metrics across the board. Our go live date for the program will be at the end of this year. This will allow head-to-head comparisons of all of our techniques in the treatment of all forms of AF. The combined outcomes reporting will yield desperately needed insights to guide future patient selection.

This idea is one that patients deserve and have been asking for so they can be better informed of their options. We will be able to accurately discuss all options to treat their AF and discuss the predicted success rates for each option along with the potential complication rates, backed by actual results from our own center.

To accomplish this goal, we adopted the STS AF module for our surgical volume. We then partnered with the Virginia Surgical Quality Initiative program (VSQI) to create and house a mirror image database with the same endpoints for endocardial ablation as well as surgical ablations. We can then upload the data from the STS AF module up to the VSQI-combined database. Once the data pool are combined, we will set goals for success rates for each type of AF: paroxysmal, persistent, and longstanding persistent, and begin the head-to-head comparisons.

The days of surgery competing with endocardial ablation for the treatment of AF are gone in our center, and we hope to help lead the nation forward in realizing that AF and especially NPAF take a team of experts with different strengths who join together in a multidisciplinary effort to pool their strengths and focus them together to the benefit of the patient.

Putting both endocardial and surgical ablation together with the same metrics is an idea of enormous merit. We absolutely support and want to foster the eventual final goal of national outcomes reporting for AF ablations in the same light as CABG mortality, etc. It is time that patients worldwide have access to our actual success rates in every center that engages in advanced therapies for AF.

Is there anything else you’d like to add?

Woollett: I personally am extremely excited about the potential opportunity to find an effective and safe treatment for a group of AF patients that has historically been very difficult to treat. Many of these patients are highly symptomatic and often have been told by other physicians that they have no further options to get back to normal sinus rhythm. My interest and enthusiasm for ablating chronic and longstanding persistent AF had waned due to the poor long-term results from catheter ablation only; our hybrid approach has now made me very hopeful about having something effective to offer these patients.

I should also note that despite clipping the left atrial appendage, we are continuing our patients on anticoagulation based on their CHADS score even if we have not seen any further atrial fibrillation. We are hopeful that there will eventually be good data to support stopping anticoagulation after clipping the appendage, but for now we are taking a very conservative approach as per the AF guidelines.

Disclosures: Dr. Woollett has disclosed that his institution has received a grant from AtriCure. He also reports he is a consultant for St. Jude Medical and his institution has grants/grants pending with St. Jude Medical. Dr. Philpott discloses that his institution, Sentara, has received grants from AtriCure for quality, outcome reporting, and performance improvement.

To link to another article in the November 2012 issue of EP Lab Digest, please visit: http://eplabdigest.com/articles/Community-Memorial-Hospital