The Thoracic Veins and Ablation for Atrial Fibrillation
Jothi Kanagalingam, Samuel J. Asirvatham
Cardiovascular Pathology Department, Johns Hopkins Medical Institution – Baltimore, Maryland, Division of
Cardiovascular Diseases, Department of Medicine and Department of Pediatrics and Adolescent Medicine - Mayo
Clinic, Rochester, Minnesota

The seminal observation6,7 that the pulmonary vein can trigger AF put into clinical practice suggestions of earlier investigators and have allowed us to appreciate the role of classic venous structures in the etiology of AF better . While the pulmonary veins are clearly the most important venous figures for AF, other thoracic veins including superior vena cava (SVC), atrial tributaries of the coronary sinus and the coronary sinus itself have all been demonstrated to be potential triggers for AF8,11,12.
In this article, we briefly discuss the collection of the most appropriate patients for venous isolation procedures and provide guidelines for the interventional ablationist to electrically isolate these veins. A cornerstone to understanding the correct technique in isolating the pulmonary veins (PV)
is an appreciation of the regional anatomy and the genesis of the typical pulmonary vein electrograms (PV potentials).
Equally important in appreciating the morphological features of the complex vein is a thorough understanding of what constitutes the PV ostium. Finally, and perhaps most importantly, a discussion on methods to minimize the welldescribed complications associated with these procedures is outlined. A brief discussion of the roles of cardiac imaging in optimizing results with the ablation management of PAF is included.

Correspondence: Samuel J. Asirvatham,Division of Cardiovascular Diseases,Mayo Clinic College of Medicine, 200 First Street SW, Rochester, MN 55905
E-mail:asirvatham.samuel@mayo.edu

Who Is the Ideal Patient for the Invasive Management of Atrial Fibrillation? The pathogenic factors responsible for chronic AF are vastly different from PAF, but care must be taken in directing patients for the isolation procedure. The patients who benefit the most from PV and other venous isolation procedures are the patients with repeated but self-terminating episodes of AF. These patients often have a structurally normal heart, are young and have only mild or moderate enlargement of the left atrium(LV)13. While hypertension and diastolic dysfunction may be present in patients with PAF, the severe comorbidities that accompany chronic AF such as ventricular dysfunction, congestive heart failure and advanced diabetes are usually absent. Patients with chronic forms of AF may benefit from venous isolation; however, very often, additional treatment that would include extensive linear ablation to modify the abnormal atrial substrate, continued antiarrhythmic drug therapy along with possible placement of antitachycardia pacemakers to address the main typical atrial flutters is required. Even when a primary substrate mediates the approach like linear ablation, ablation of fragmented atrial electrograms or targeting the retroatrial ganglionated plexii, knowledge of the regional anatomy and addressing the (PVs) remain critical14,15. It should be emphasized that although patients with PAF tend to be the same patients with structurally normal hearts, normal size atria and of a younger age but the critical feature in assessing which patients are likely to benefit from pulmonary venous isolation which is the paroxysmal nature of the arrhythmia itself16. The fact that the arrhythmia self terminates, is powerful evidence of the lack of severe substrate abnormalities and thus
they enhance the etiological role of the initiating sites for arrhythmia. Accurate selection is particularly important in the Indian subcontinent given the limited number of centers, procedures available and potentially prohibitive costs associated with repeated and failed procedures. The criteria to identify the best patient for these procedures are summarized in Table1 given below17,18.

PULMONARY VEIN ISOLATION
Once the appropriate patient has been recognized with symptomatic PAF despite medical therapy, the next step is knowing how best to electrically isolate the PV. To do this, the operator must be fully conversant with understanding the PV potential and the location of the PV ostium. Once this is done, the actual technique of the ablation must be known, specifically how to validate the PV potentials and to know when the procedure is complete, i.e. entrance block into the vein has occurred.
THE PULMONARY VEIN POTENTIALS
The bipolar electrograms from an electrode tipped catheter placed with the PV produces a characteristic electrogram (Figure 1). In patients with AF, three distinct components for the electrogram are noted. First, a far-field component that represents atrial activation is followed by a period of electrical silence (isoelectric period) and then a sharp distinct nearfield electrogram resulting from PV myocardial activation is identified. The far-field component may be the right or LA, depending on which PV is being mapped and the near-field component is almost always a result of conduction into the PV from the LA (see below). The PV musculature itself is a complex layer of atrial myocardium that extends into the PV for variable lengths. An important part of this complex electrogram is the isoelectric period. While the exact reason for this delay in activation is not completely understood, it importantly correlates with the anatomic ostium of the vein. Thus, the site of electrical delay into the vein or with ectopy out of the vein corresponds to the ostium of the vein. Studies have shown that the PV that are most arrhythmogenic and exhibit the greatest amount of ostial delay (> 30 to 40 ms)8.

The key portion of the PV isolation procedures is the circumferential ablation, in the atrium proximal to the ostium of the PV that converts this ostial delay to complete conduction block and thus the abolition of the PV potential. Although the exact pathology of the ostium is unknown, an abrupt change in myocardial fiber orientation, fibrosis from mechanical wear and tear or other cardiac diseases, interdigitation of venular smooth muscles with cardiac syncytial muscle and possible nodal remnants located near the ostia have been thought to be the reason.

THE PULMONARY VEIN OSTIUM

An important conundrum exists with regard to the PV ostium19. On the one hand, it is critical for electrophysiologist to know exactly where the PV ostium is located since ablation within the vein results in PV stenosis without getting rid of the entire arrhythmogenic substrate. On the other hand, even with an autopsied heart, it is not possible to know exactly where the PV ends and the LA begins. Usually, four PVs (two to eight veins) drain into the posterior wall of LA. The actual venoatrial junction may in some cases be an abrupt transition from a cylindrical vein to the more bulbous atrium (often the case with the left sided veins), but other times, the transition is extremely gradual with a funnel-like opening or a cloacal entrance that in turn is the drainage site for more than one vein (Figure 2). Since the exact anatomic location of the ostium is difficult to know, electrophysiologists should be aware of this difficulty and err on the side of ablating more proximally into the LA. Imaging studies, as well as electrical maneuvers, can aid the electrophysiologist in approximating the site of the PV ostium.

CT/MRI/ANGIOGRAPHY

Commonly used methods to appreciate PV anatomy and aid the ablation procedure include CT scanning, MRI, as well as retrograde pulmonary venography performed during the ablation procedure. CT or MRI have the advantage of giving exquisitely detailed anatomic information on the number and morphology of the PVs. The methods of branching aberrant or unusually positioned veins as well as complex drainage pattern can all be viewed by the operator prior to manipulation in the patient’s LA. A major disadvantage of these off-line imaging resources, however, is that changes in left atrial volume (LAV) and the relative anatomy of the atrium occur during ablation procedures as a result of dehydration or

branch point of small venous tributaries as being the entrance of the PV into the LA. Intracardiac Ultrasound The linear phased-array imaging with the probe placed in the right atrium via a femoral vein can be very useful for online imaging during PV ablation procedures20,21. Intracardiac ultrasound using this technique can also aid transseptal puncture, assess for the development of thrombus and quickly identify potential causes of hypotension including cardiac perforation20. One of the most important uses for this technique, however, is to visualize in real-time the PV ostial region and assess where the ablation catheter is in relation to the vein. The intracardiac ultrasound probe is placed close to the fossa ovalis, and the image adjusted for that the lateral wall of the LA and PV can be visualized. With clockwise rotation, the right-sided PV can be seen oblique or short axis, where as the left sided PVs are seen in their long axis with slight counter clockwise rotation. Doppler evaluation can assess pulmonary venous flow to quickly detect spasm or thrombus formation as well as dissection.

ELECTRICAL METHODS TO IDENTIFY
THE PULMONARY VEIN OSTIUM

The use of endocardial pacing techniques can be very helpful in identifying the “electrical ostium” of the PV22. When a circumferential mapping catheter is placed inside the vein, the classical PV potential is seen. If a mapping catheter is then used to pace in the LA, left atrial signal is captured and the PV potential is easily visualized. However, when pacing inside the vein, the PV potential is captured, and because of the saturation from the pacing artifact, the pulmonary vein potential is no longer seen. The concept of pacing near the PV is described below. Thus, prior to delivery of ablation energy if the operator is concerned whether the ablation catheter is inside the vein, pacing can be initiated. If the PV potential is “captured”, then the catheter is in the vein. With continued pacing, the catheter is dragged back, and when the PV potential is “released”, then the ostium has been crossed and the catheter is now in a safe location in the atrium for ablation to proceed.

ISOLATION OF THE PULMONARY VEINS

Radiofrequency ablation to isolate the PV involves the placement of ablative lesions in the LA proximal to the ostia of the PV. This has been variably referred to as periostial ablation or wide area circumferential ablation. Regardless of the name tag, ablation is in the LA and not at the ostium or in the PV. Variations in the technique based on the use of mapping systems or whether individual veins are isolated exist, but the essentials of the

axis of the vein). This is because earlier activation will always be seen at electrodes closer to the ostium and if the circumferential catheter is tilted inside the vein, then a false impression of a connecting or early site of activation will be given by looking at the electrodes that are closer to the ostium compared to those that are tilted far within. 3. One must have a qualitatively different signal on the ablation catheter than on the mapping catheter. That is, on the ablation catheter the near-field electrogram will be earlier and representing the LA myocardium that is to be ablated. In contrast, on the mapping catheter the near-field signal is the PV potential and occurs after the far-field atrial signal and ostial delay explained above.

VALIDATION OF THE PULMONARY VEIN
POTENTIAL

As described above, most cases of the PV potential is easy to recognize. However, following ablation and with certain anatomic variants (anterior vs. superior vein very close to the appendage), it can be exceedingly difficult to know whether the PV is still being recorded following ablation on the circumferential mapping catheter. Misdiagnosis can lead to excessive ablation that is unnecessary and potentially associated with complications, and if the PV potential that is actually present is not recognized, then inadequate ablation may occur, potentially giving rise to recurrent arrhythmia. When a catheter is placed in the PV in addition to the atrial farfield signal and the PV potential near-field signal, other complex electrograms generated from neighboring structures can also be seen. When the catheter is placed in the left-sided vein, contaminant signals from the LAA, vein of Marshall, adjacent PV, etc. may all be recorded. Similarly, when a catheter is placed in the right upper vein in addition to the signals from the LA and PV musculature, electrograms from the right atrium (RA), SVC and adjacent PVs may be found. Sometimes, despite experience and caution, inadvertent ablation within the PV occurs and these signals become fragmented, further complicating the issue. Some relatively straightforward pacing maneuvers, once understood and mastered, can easily identify what is and what is not the PV potential. Peri-vein pacing The concept underlying this technique is that when complex electrograms are found and pacing is performed at an anatomic site responsible for one of those electrograms, that component will be advanced or pulled towards the pacing stimulus artifact. After a mapping catheter is placed in the left upper PV and three component signals are found, one may be of the LA, another the LAA and yet another is the PV potential. If LA pacing is performed close

to the PV but not within the vein, then the left atrial component of that signal will be advanced towards the pacing stimulus with little or no effect on the PV potential and contaminating LAA signals. On the other hand, if the pacing site is now in the LAA, then the component of the complex electrogram that represents far-field activation of the LAA will now be advanced towards the pacing stimulus. By understanding the anatomy and the principle of peri-vein pacing, the exact origin of each signal can be easily identified. It should be noted that the coronary sinus pacing or right atrial pacing are not a substitute for pacing the LA near the PV to advance the LA component. This is because of the anatomic distance of these pacing sites from the PV. Intra-vein pacing Here, the concept is that when the pacing catheter is now placed within the PV, the only electrogram that will be advanced is PV potential itself. While this maneuver directly allows recognition of the potential we are most interested in, namely the PV potential, the problem is that often high output pacing is required to capture the myocardium, and this may result in far-field capture of anatomic neighbors such as the LAA. A variation of intra-vein pacing is described above in electrical methods to identify the PV ostium.

IDENTIFYING ENTRANCE BLOCK

The primary endpoint for PV ablation procedures is the electrical isolation of the PV, which is identified by entrance block (failure of propagation of LA activation into the PV). It this is mandatory that the ablationist understand that loss of PV potential should not result from ablation if the PV myocardium and fragmentation is a signal, but rather from entrance block. Thus, as ablation is performed in a periosteal manner as described above, one should see a gradual diminution of the far-field atrial signal, and progressive delay in activating the PV potentials, but importantly no change in the amplitude of the pulmonary vein potential itself. These desired changes cannot be over emphasized as to their significance. Sometimes, gradual diminution of the PV potential may be observed – so-called "melting away of potentials". This is a highly undesirable effect as it reflects ablation within the PV itself. Ideally, there should be progressive delays and eventual
loss of conduction into the vein signified by the loss of the PV potentials. The far-field LA signals change very little in this sequence. If a marked changed in these signals is noted, the catheter may have moved deep into the PV. Typically, circumferential or near circumferential ablation is required to get entrance block. If an operator notes that he/she gets entrance block repeatedly with just one or two burns, it is certain that ablation is being performed into the PV and the operator has mistaken a branch of the pulmonary vein for the main PV itself.

EXIT BLOCK
Exit block is defined as pacing and capture of the PV musculature

but without conduction from the vein to the rest of the atrium. While exit block is closer physiologically to the desired results of preventing arrhythmogenic activity from the PV to exit to the atrium and produce AF, this is difficult to demonstrate in many cases. Often, pacing within the PV, particularly if there is a relatively short sleeve of musculature in the vein, will result in noncapture of the local myocardium, and exit block can not be demonstrated. Sometimes, simple mechanical catheter trauma or spontaneous ectopy with resulting exit block can be easier to demonstrate. With both entrance and exit block, it is important to recheck for these parameters about a half an hour following ablation around the vein with the use of isoproterenol. A recurrence of conduction into the outside of the vein should be addressed by repeating a circumferential ablation procedure often with identifying a gap in the circular series of ablation points.

NON-PULMONARY VEIN FOCI

Thoracic vein triggers for PAF are not confined to the PVs. Well-established alternative trigger sites include the atrial myocardium within the SVC and the vein of Marshall. In addition, the electrical sleeves around the coronary sinus and electrically active connections between the PVs or between the right atrium and right upper PV may require targeted ablation. Superior vena cava isolation Atrial myocardium extends into the SVC for various lengths12,27 this may be continuous with the musculature with the Azygos vein. If ectopy or other arrhythmogenic potentials similar to that seen on the PV are noted with a circumferential mapping catheter placed within the SVC, then circumferential isolation of this vein can be undertaken. The technique is identical to that described for PV isolation except for some important differences (Figure 5). First, the phrenic nerve often lies in the posterior (posterolateral or posteromedial course) aspect of the SVC – RA junction. To avoid damage to the phrenic nerve during ablation, pacing at an output of 5 to 10 milliamp should be performed at all sites prior to delivering radiofrequency energy. The ablation circle can be adjusted to avoid damage to the phrenic nerve in most cases and yet achieve entrance block with loss of the superior vena caval potentials. In some situations however, complete isolation is not possible with this technique because of phrenic nerve stimulation when attempting the preablation pacing. Here, the technique of pacing within the SVC while cryoablating at the ostium can be performed28.
Vein of Marshall Ablation Similar to the approach for the SVC, arrhythmogenic potentials with evidence of ostial delay are found when mapping in the embryological remnant of the left superior vena cava; i.e. the vein of Marshall (Figure 6). In some cases,

narrowing is noted with a clear correlation of narrowing with complaints such as shortness of breath, hemoptasis and cough, then PV angioplasty or stenting should be performed. Unfortunately, there is a greater than 50% recurrence rate of stenosis following PV dilation and stenting procedures. At
times, there may be mild narrowing of the PV with unclear symptomatology. In these situations, quantitative ventilation profusion scanning, along with exercise testing to look for oxygen desaturation, can clarify the issue.
Atrial-esophageal Fistula Formation
This is a well-recognized and unfortunate result of some AF ablation procedures. The anterior wall of the esophagus is immediately related to the posterior, part of LA at variable locations and is at times posterior to the PV ostia. The course of the esophagus is typically oblique but variable from patient to patient and in a given patient varies with peristalsis and patient position. The major symptoms and signs of this dreaded complication are unexplained bacteremia, air found in the LA during echocardiography, dysphasia, hematemesis and rapidly progressive endocarditis. Terminal events may include air embolization to the brain or stroke from embolized vegetation. Immediate recognition of this condition is the only way to avoid certain death. As soon as the possibility is considered surgical exploration or stenting of the esophagus must be immediately offered as options. Avoiding this complication is best done by limiting power delivery in the posterior part of atrium. Intracardiac echocardiography can visualize the oblique sinus and the anterior wall of the esophagus (Figure 8). Changes in echo texture at these locations should immediately result in cessation of ablation energy delivery. Remembering to stop energy delivery whenever the atrial electrogram diminishes (rather than waiting for some preset empiric period of time) will likely help avoid this issue. Some operators will place a temperature probe in the esophagus, and clearly if the temperature rises, energy should be turned off. However, significant damage to the wall of the esophagus may have already occurred before the endoluminal temperature
measured by such probes increases.
THROMBOSIS AND STROKE

A devastating complication of LA ablation, which presently does not have a definite method of prevention, is thrombus formation and stroke. While anticoagulation with heparin is the main method of preventing thrombosis, it is not always sufficient. Heparin needs to be initiated early and experienced operators will fully heparinize with a target aCT of > 300 seconds even prior to transseptal puncture. At the very least, as soon as transseptal puncture is done, complete heparinization is mandatory. Other techniques include careful sheath management