Share this article on:

Does Cryomaze Injure the Circumflex Artery?: A Preliminary Search for Occult Postprocedure Stenoses

Cheema, Faisal H. MD*†; Pervez, Mohammad B. MD; Mehmood, Mansoor MD; Younus, Muhammad J. MD; Munir, Mohammad B. MD; Bisleri, Gianluigi MD; Barili, Fabio MD, PhD; Ayala, Ivan L. MD; Ad, Niv MD§; Cox, James L. MD; Roberts, Harold G. Jr MD†¶

Innovations: Technology & Techniques in Cardiothoracic & Vascular Surgery: January/February 2013 - Volume 8 - Issue 1 - p 56–66
doi: 10.1097/IMI.0b013e31828e5267
Original Articles

Objective: Ensuring the transmurality of the mitral isthmus lesion, a critical component of the cryomaze, entails mirror-image application of the cryoprobe both on endocardial and epicardial surfaces when carrying out ablation. Concerns of circumflex artery injury have been expressed during the epicardial application of the cryoprobe over the coronary sinus as the artery courses on the posterior surface of the sinus in the atrioventricular (AV) groove. The objective of this study was to analyze the incidence of significant injury to the circumflex artery and its impact on outcomes, if any, in those patients who have undergone cryomaze.

Methods: Between August 2004 and December 2009, a total of 223 patients underwent argon-based cryoablation (120-second application at −140°C). After Western Institutional Review Board approval, 20 consecutive patients with normal results of preoperative coronary angiograms (right dominance, 75%; left dominance, 15%; codominant circulation, 10%) and who were at least 6 months postablation were enrolled in this study. The mean ± SD age was 60.74 ± 14.99 years, 35% were men, and 50% belonged to New York Heart Association class III/IV. The mean ± SD atrial fibrillation duration was 23.83 ± 36.28 months (65% were paroxysmal). Ten percent (n = 2) underwent primary cryomaze, 40% (n = 8) underwent cryomaze plus mitral valve repair, and 50% (n = 10) underwent two or more concomitant valvular procedures. Twelve patients underwent biatrial cryomaze, and eight underwent only left-sided cryomaze. All patients underwent a 24-hour Holter monitoring, electrocardiogram stress test, and a coronary computed tomographic angiogram, as per the protocol of this study.

Results: At discharge, 85% had normal sinus rhythm, whereas 15% of the patients were paced. On a mean ± SD follow-up at 32.57 ± 19.51 months, the Holter and/or pacemaker interrogation revealed AV synchrony in all patients—16 in sinus rhythm and 4 with heart block who converted to AV synchrony after subsequent pacemaker implantation. The stress test was available for 18 patients, and its results were negative in all of them. On the computed tomographic angiogram, 95% of the patients had a completely patent circumflex artery. Stenosis was noticed in only one patient (right dominant circulation), with a 30% to 40% tubular stenosis of the circumflex artery. However, this lesion corresponded to the P1 area of the mitral annulus and was significantly proximal on the circumflex to the P3 area, where the cryoprobe was applied during the cryomaze procedure.

Conclusions: Barring one case of partial circumflex stenosis, likely due to the ongoing normal progression of coronary artery disease, these data derived from a limited prospective trial suggest that epicardial application during the cryomaze procedure does not cause anatomic or physiological compromise of the circumflex artery. Nevertheless, laboratory and anecdotal evidence exist that conflict with this conclusion, and caution should be exercised when applying cryothermy in the vicinity of coronary arteries.

From the *College of Physicians & Surgeons of Columbia University, New York Presbyterian Hospital, New York, NY USA; †Aegis Cardiovascular Research Foundation, Fort Lauderdale, FL USA; ‡Atlantic Coast Radiology, Sunrise, FL USA; §Fairfax Hospital, Falls Church, VA USA; ∥Washington University School of Medicine, St Louis, MO USA; and ¶Jim Moran Heart and Vascular Research Institute at Holy Cross Hospital, Fort Lauderdale, FL USA.

Accepted for publication January 30, 2013.

Presented at the Annual Scientific Meeting of the International Society for Minimally Invasive Cardiothoracic Surgery, May 30 – June 2, 2012, Los Angeles, CA USA.

Supported by a research grant from ATS Medical Inc, which was acquired by Medtronic, Inc, Minneapolis, MN USA, in 2010.

Disclosures: Faisal H. Cheema, MD, and Harold G. Roberts, Jr., MD, receive research grants from Edwards Lifesciences, Irvine, CA USA, and Medtronic, Inc., Minneapolis, MN USA. Niv Ad, MD, is a consultant for Medtronic, Inc., Minneapolis, MN USA, Atricure, Inc., Cincinnati, OH USA, and Estech, Inc., San Ramon, CA USA. James L. Cox, MD, is a consultant for Atricure, Inc., Cincinnati, OH USA, Estech, Inc., San Ramon, CA USA, SentreHEART, Inc., Redwood City, CA USA, CorMatrix, Roswell, CA USA, and Adagio, Laguna Hills, CA USA. Mohammad B. Pervez, MD, Mansoor Mehmood, MD, Muhammad Jabran Younus, MD, Mohammad B. Munir, MD, Gianluigi Bisleri, MD, Fabio Barili, MD, PhD, and Ivan L. Ayala, MD, declare no conflict of interest.

Address correspondence and reprint requests to Faisal H. Cheema, MD, Division of Cardiothoracic Surgery, College of Physicians and Surgeons of Columbia University - New York Presbyterian Hospital, MHB 7 GN 435, 177 Fort Washington Ave, New York, NY 10032 USA. E-mail:

Atrial fibrillation (AF) is one of the most common cardiac arrhythmias, affecting up to 2.2 million people in the United States. Atrial fibrillation has been associated with a number of cardiac conditions, particularly coronary artery disease (CAD), valvular heart disease, hypertension, cardiomyopathy, and congestive heart failure. In addition, it is responsible for at least 20% of all strokes in the United States. Atrial fibrillation usually results from multiple macroreentrant circuits in the left atrium and, less commonly, in the right atrium. The Heart Rhythm Society classifies AF into three types: paroxysmal, persistent, and long-standing persistent. Paroxysmal AF is defined as an intermittent episode lasting less than 48 hours, with varying frequency that terminates spontaneously. Persistent AF generally lasts longer and requires pharmacologic and/or electrical cardioversion for termination. If an electrical or a chemical cardioversion attempt fails, it is categorized as long-standing persistent AF. In paroxysmal AF, these circuits most commonly arise from the junction of the pulmonary venous endothelium and the left atrial endocardium. With persistent and long-standing persistent AF, macroreentrant circuits arise in the pulmonary veins and other areas of the left and right atria.

The various treatment modalities available for AF include antiarrhythmic drugs, cardioversion, atrioventricular (AV) node ablation, pacemaker insertion, catheter-based ablation, and surgical ablation.1 Pharmacologic therapy is, at best, 50% effective in maintaining sinus rhythm.2 In addition, the long-term deleterious effects of antiarrhythmics can be clinically significant. Cheema et al3 reported only 28% long-term success rate at a mean ± SD of 28 ± 11 months with catheter ablation. Similarly, Weerasooriya et al4 recently reported 29% success rate after a single catheter-based ablation and 63% after multiple catheter ablations in patients with lone AF at 5 years after surgery. Variously reported success rates for persistent AF reveal even more dismal results.5

The Cox-maze III,6 considered to be the standard procedure for the surgical treatment of AF, has been extensively modified during the last decade with the availability of new energy sources to produce the atrial lesions. Although high-intensity focused ultrasound, microwave, unipolar radiofrequency (RF), and laser were once commonly used or evaluated energy sources, bipolar RF and cryoablation are now the primary energy sources being used. Although there has been some debate on the merits of heat versus cold energy sources, the main advantage of cryothermy stems from collagen being impervious to its effects. This results in the apparent lack of permanent collateral damage to surrounding structures such as the valve leaflets, the coronary arteries, the coronary sinus, and the esophagus as well as the absence of tissue vaporization or charring. Cryomaze7,8 consists of a set number of lesions in the left and right atria, made with a flexible argon-based cryothermy ablation probe. To create a cryolesion, the cryoprobe is applied at a temperature of −140°C for 2 minutes. Application of the cryothermic device results in the production of a firm, fibrous scar with smooth endocardium that is electrically silent and has a sharp histological border.9

Recently, there have been concerns over the possible acute and chronic effects of cryothermia on coronary artery patency and integrity. Holman et al10 performed an experimental canine study in which, after application of a cryothermic probe in the territory of the left anterior descending artery, they observed the presence of cryothermia-induced coronary artery damage that resulted in decreased blood flow at the cryolesion site, possibly due to coronary artery stenosis. Subsequent quantification of radioactive tracer microspheres demonstrated decreased myocardial blood flow in the left anterior descending artery territory. Similarly, Bakker et al11 demonstrated in their experimental animal study that cryoablation in the area of the circumflex artery resulted in thrombus formation, as demonstrated during histological follow-up. Studies focusing on the coronary anatomy have shown that the risk for coronary artery injury, especially to the circumflex artery, is highest in the dominant left coronary artery system because of the close proximity of the mitral annulus and the circumflex artery during mitral valve surgery.12–15 Raza et al16 described a case in which they encountered circumflex artery injury after minimally invasive mitral valve repair and left atrial cryomaze. The injury was successfully mitigated by revascularizing the circumflex artery using percutaneous intervention. However, the authors were unable to determine whether this circumflex injury was a direct result of cryomaze or an iatrogenic suture-needle injury sustained while carrying out the concomitant mitral annuloplasty. A more recent report demonstrated right coronary artery (RCA) spasm after cryomaze,17 but no long-term stenotic injury was documented. A systematic literature search confirmed a complete dearth of any definitive studies addressing this issue in humans.

The objective of this study was to determine whether the coronary arteries after cryomaze are healthy from an anatomic and a physiologic perspective. Although the coronary computed tomographic angiogram (CTA) will produce a comprehensive image of any possible anatomic damage, a stress electrocardiogram (ECG) will determine whether the functional element of the coronary arteries is intact. Hence, the specific aim of this study was to analyze the incidence of postoperative injury to the circumflex artery in patients with no significant CAD on their preoperative coronary angiograms.

Back to Top | Article Outline


Our cryomaze database, consisting of 223 patients from August 2004 to December 2009, was searched for patients who underwent open chest or minimally invasive cryomaze, with or without a concomitant valve procedure at least 6 months before the date of follow-up coronary CTA. Exclusion criteria included a history of CAD, as identified by preoperative coronary angiography; previous ablation attempt(s) for AF; previous thoracic procedures; pregnancy; and patients at risk for iodinated contrast due to renal dysfunction. After obtaining investigation review board approval from the Western Institutional Review Board (Western Institutional Review Board study number 1113589), a total of 20 patients matching the abovementioned criteria consented for participation in the study. They were then prospectively scheduled for coronary CTA, ECG stress test, and 24-hour Holter monitoring or pacemaker interrogation. Data analysis was accomplished using the Statistical Package for the Social Sciences version 16.0 (SPSS Inc., Chicago, IL USA). Basic demographic data and preoperative risk factors are given in Table 1. The methods of this study are summarized in Figure 1.

Back to Top | Article Outline


Cryolesions were created using an argon-based cryomaze surgical ablation system with a 10-cm malleable probe (Cardioblate CryoFlex system; Medtronic Inc, Minneapolis, MN USA, previously owned by ATS Medical, Minneapolis, MN USA) for all procedures.

Back to Top | Article Outline

Surgical Procedure

All cryomazes, carried out by the same surgeon, were created using the abovementioned device after initiating cardiopulmonary bypass and cardioplegic arrest. For the left atrial lesion set (Fig. 2), the oblique sinus was widely opened to provide epicardial access to the coronary sinus. A vertical left atriotomy was carried out. A box lesion was made around the pulmonary veins using a 10-cm–long cryocatheter (probe temperature, −140°C; mean ± SD total duration, 8.7 ± 2.8 minutes), with each application lasting 2 minutes. A linear cryolesion was made from the box across P3 as the mitral isthmus lesion. Through the oblique sinus, a linear cryolesion was then made directly on the epicardial surface of the left atrium and the coronary sinus that “mirrored” the endocardial mitral isthmus lesion. The os of the left atrial appendage was oversewn endocardially in two layers with either continuous 4-0 Prolene (Ethicon Inc, Somerville, NJ USA) or Goretex (W.L. Gore & Associates, Inc, Newark, DE USA). For the cryomazes performed between April 2005 and December 2006, a cavotricuspid isthmus lesion was the only right-sided lesion. The right-sided lesion set (Fig. 3) before and after this era began with a horizontal atriotomy in the midportion of the free wall of the right atrium. The medial aspect of the atriotomy was continued as a linear cryolesion across the posterior tricuspid annulus. At the posterior aspect of the atriotomy, a cryolesion was made from the superior vena cava down to the inferior vena cava. A stab wound was made in the right atrial appendage that provided access for the probe to create a linear cryolesion medially across the anterior tricuspid annulus. The cryocatheter was then turned laterally, and a linear cryolesion was made toward the right atriotomy, leaving a 3-cm corridor between the lesion and the atriotomy. The operative details are given in Table 2. The following lesions were created: full left side only (40%), full left side and full right side (35%), and full left side and right cavotricuspid isthmus lesions (25%).

Minor variations occurred in the procedures because of the differences inherent to the robotic and open approaches, but these variations pertained to access rather than to the cryolesions made.

Back to Top | Article Outline


The most recent follow-up details are summarized in Table 3. Prospective data were collected for the evaluation of coronary vessel injury and residual AF. Standard protocol included clinical follow-up at 1, 3, 6, and 12 months. The patients with postoperative AF received amiodarone. If the AF did not resolve within a month, outpatient electrical cardioversion was carried out with transesophageal echocardiography guidance. The computed tomographic angiogram follow-up time ranged from 6 to 69 months. To analyze the incidence of postoperative injury to the circumflex artery, the patients were brought in for follow-up at a minimum of 6 months after the cryomaze (mean, 2.7 years) and were assessed by coronary CTA. During the routine follow-up and/or immediately before CTA, every patient underwent a 24-hour Holter monitoring to assess his/her cardiac rhythm.

Back to Top | Article Outline


A total of 20 consecutive patients, fulfilling the inclusion criteria, were included in this study. The mean ± SD duration of AF was 23.83 ± 36.28 months. Atrial fibrillation type included paroxysmal (65%, n = 13), and long-standing persistent (35%, n = 7), (n = 16), dilated cardiomyopathy (n = 10), and rheumatic (n = 3). The patients’ mean ± SD age was 60.74 ± 14.99 (range, 38–86) years, with 65% belonging to the female sex. Most patients (90%) reported symptoms preoperatively, and 50% belonged to the New York Heart Association class III/IV. The preoperative mean ± SD ejection fraction was 50.26% ± 10.07%; and the postoperative, 51.68% ± 8.49%. Twenty-five percent (n = 5) underwent the procedure robotically, and the remaining 75% of the subjects (n = 15) underwent a conventional transsternotomy cardiac procedure. Ten percent (n = 2) underwent primary cryomaze, 40% (n = 8) underwent cryomaze plus mitral valve repair, and 50% (n = 10) underwent two or more concomitant valvular procedures. The repair rate for the mitral procedures was 83% overall and 100% for nonrheumatic etiologies. There were no perioperative deaths. Postoperative complications included pleuropericarditis (n = 1), pleural effusion (n = 1), respiratory failure (n = 1), transient altered mental status (n = 1), and transient ischemic attack (n = 1). The mean ± SD extubation time was 18.58 ± 14.41 (median, 14.25) hours, the intensive care unit duration was 81.06 ± 57.60 (median, 72.0) hours, and the total length of hospital stay was 12.11 ± 5.35 (median, 12) days. The hospital morbidity and mortality data are summarized in Table 4. At discharge, 85% had normal sinus rhythm, whereas 15% of the patients were paced. On the mean ± SD follow-up at 32.57 ± 19.51 months, the Holter or pacemaker interrogation revealed AV synchrony (16 in sinus rhythm, 4 paced) in all patients, with heart block in four patients who converted to AV synchrony after subsequent pacemaker implantation. At discharge, 15% (n = 3) of the patients were on antiarrhythmics; and at the time of follow-up, 10% (n = 2; Fig. 4). Anticoagulant use was 25% (n = 5) at the last follow-up. Two patients refused the stress test; however, its results were negative in 100% of the remaining 18 patients.

The postcryomaze coronary CTA demonstrated that 95% of the patients had a completely patent circumflex artery. Significant stenosis was seen in only one patient (right dominant circulation), with a 30% to 40% tubular stenosis of the circumflex artery (Fig. 5). However, this lesion corresponded to the P1 area of the mitral annulus and was significantly proximal on the circumflex artery to the P3 area, where the cryoprobe was applied during the cryomaze procedure.

Back to Top | Article Outline


As cryotechnology continues to evolve and becomes more widely accepted in clinical practice, its risk to the coronary arteries that are in anatomic proximity to the areas of targeted ablation remains an issue of debate. Although a rare complication, coronary artery injury has been previously associated with cryoablation in the literature.10–17 Although more recent reports, mostly single patient, have investigated the potential damage of cryothermy to the coronary arteries, the literature is largely devoid of any major studies. Speculation about possible cryothermy-mediated coronary injury, based on random case reports, calls for a proper investigation to search for any occult coronary artery stenosis after the cryomaze procedure.

At a minimum, all cryomaze lesion sets include the Cox-maze III lesions made in the left atrium because of the observations that a left-sided maze alone decreases the incidence of AF18 by 85% to 90%. In the left atrium, a complete set of cryomaze lesions includes a box lesion around all four pulmonary veins, an endocardial mitral isthmus lesion, a “mirror-image” epicardial mitral isthmus lesion that includes the coronary sinus, and closure of the left atrial appendage os. In the right atrium, a cavotricuspid isthmus lesion has been made as the lone right-sided lesion. In fact, this was our preferred lesion during the era from April 2005 to December 2006. However, in view of the fact that macroreentrant circuits can develop in the upper two thirds of an enlarged right atrium despite a well-constructed cavotricuspid isthmus lesion, more recently, our preference is a complete set of Cox-maze III right-sided lesions. The cryolesions across the AV groove that are present in the complete right atrial lesion set can potentially freeze the RCA. However, the limited number of complete right-sided lesions in this study makes drawing conclusions regarding the RCA and cryothermic injury difficult. Nevertheless, none of the CTAs or stress tests in this study support RCA injury.

The circumflex artery courses near the anterolateral commissure, and the coronary sinus is adjacent to the annulus of the posterior mitral valve leaflet; both vessels lie within the left AV groove. This is even more clinically relevant in the left dominant coronary system in which the circumflex artery runs in the AV groove on the posterior surface of the heart and terminates in the posterior descending artery. The RCA runs in the AV groove on the anterior surface of the heart and normally terminates in the posterior descending artery in a right dominant or codominant circulation. On the basis of these observations, it is hypothesized that, during the left-sided cryomaze, when an epicardial lesion is made over the coronary sinus (to ensure transmurality of the counter endocardial mitral isthmus lesion), the circumflex artery is at risk for injury.

To reliably cure long-standing persistent AF, a transmural mitral isthmus lesion that includes the coronary sinus at that level is crucial. Failure to achieve this lesion can produce atypical left atrial flutter that is notoriously difficult to manage, either medically or with catheter intervention. From our current understanding of cardiac anatomy and available power sources, reliably producing a transmural mitral isthmus lesion requires cardiopulmonary bypass. In Cox’s original description, the mitral isthmus lesion was created by extending the inferior limb of the left atriotomy to the P3 portion of the mitral annulus. The coronary sinus was exposed after the fat pad was dissected and then epicardial ablation was carried out with a cryoprobe. This reusable device (Frigitronics; Cooper Surgical, Pleasanton, CA USA) used nitrous oxide and achieved temperatures of −60°C. No reports of coronary injury were ever documented with this approach. In addition, nearly all procedures reproducing the Cox-maze III lesion set, including those using bipolar RF, generally use cryothermy when creating lesions around the coronary sinus as well as the mitral and tricuspid annuli. Again, there are no reports of coronary injury associated with these procedures.

The previously mentioned case reports make apparent that the exact pathophysiology of any potential vessel injury and the final manifestation of the injury remain unclear. As such, a comprehensive approach to identify potential injury is warranted. For example, if the injury manifests itself as physical vessel damage, as in the case report by Doguet et al,19 it would be detected by coronary CTA. On the other hand, if the damage comes in the form of predisposition to coronary vasospasm under stress, as in the case report by Berreklouw et al,20 a stress ECG would likely detect that derangement. Only with documented evidence of the vessels being intact anatomically and physiologically can the true safety of the procedure begin to be assessed.

In light of our findings, it is possible that the underlying cause of circumflex vessel injury in the case reports of Doguet et al19 and Raza et al16 is from suture-related injury secondary to the concomitant mitral valve surgery. Rajbanshi et al17 reported a case of RCA spasm after ablation at the vicinity of the right atrial lesion. The coronary spasm was successfully reversed with the use of intracoronary nitroglycerin, indicating that the cause of spasm may have had other contributing factors. Recently, Wong et al21 reported circumflex artery injury after catheter ablation. In this report, 15 of the 54 patients undergoing ablation had subclinical narrowing that reversed upon administration of intracoronary glycerine trinitrate. We conducted a detailed literature search on postablation vascular injury. The most significant studies that met our criteria are summarized in Table 5.

Although one case of coronary stenosis in our series is debatable, it could have several contributing factors. Because of its location in the circumflex artery, we believe its etiology to be the normal ongoing progression of CAD in an octogenarian. At discharge, the patient had a normal sinus rhythm. Two years after surgery, she had a negative stress test and continued NSR. This circumflex lesion corresponded to the area of the posterior mitral annulus near the P1 area. Therefore, it was unlikely to have been caused by the cryoprobe because the ablation was applied at P3, which was anatomically remote from the site of coronary stenosis.

Although this subset of patients was selected from a much larger cohort, the results did support our contention that cryomaze was highly effective in relieving AF burden. In addition, despite mitral valve repair and the lack of coronary grafts to improve left ventricular function, the ejection fraction did not deteriorate with surgery (50.26% ± 10.07% preoperatively vs 51.68 ± 8.49% postoperatively), as stated earlier in the Results section. This was probably a reflection of the preponderance of mitral repairs rather than replacements as well as the previously documented salutary effect of a successful maze on left ventricular function.

Cryomaze can reliably produce a full set of transmural lesions and, thus, greatly simplify the creation of an effective operation to cure AF. However, concerns that the mitral isthmus lesions are producing occult injuries to the circumflex arteries have been raised. It is hoped that the results of this study, albeit based on a relatively small number of patients, will begin to allay those fears. It is encouraging that all of these patients had no significant AF burden by 24-hour Holter or pacemaker interrogation despite all patients having a long preoperative history of AF. In addition, none of the CTAs or stress tests are consistent with cryothermy-induced coronary injury. The location of the stenosis in the lone patient with a 30% to 40% circumflex lesion is in the P1 area. This area is never frozen during the procedure, and thus, the lesion is not consistent with a cryomaze injury. Expanding the enrollment of patients into this study protocol, particularly those with left dominant circulation, is crucial in confirming our preliminary conclusions.

A secondary observation of this study is the uniform lack of postoperative AF burden. We believe that the success rate of our study in achieving AV synchrony has been due to the cryomaze technique faithfully replicating most lesions of the Cox-maze III lesion set.

Although the data from this small study do not show any adverse complications from cryoablation, in no way do we advocate the carte blanche freezing of coronary arteries. Some laboratory and anecdotal clinical evidence supports cryothermia-induced coronary injury, so we believe that avoiding the intentional freezing of coronary arteries is prudent. Although larger studies are needed to definitively investigate the possibility of vessel injury in cryomaze, this is the first formally designed study investigating the phenomenon in humans. Nevertheless, this small cohort of patients demonstrates no evidence of cryothemia-related coronary injury and is consistent with our much larger overall experience with cryomaze. We have yet to recognize acute or chronic coronary injury associated with cryomazes performed on our patients. However, the scope of this study is restricted to only midterm follow-up, which precludes us from drawing any conclusions of the effects of cryoenergy on the development of CAD in the long-term. Whether cryoablation gives rise to the development of arteriosclerotic plaques over an extended period and whether cryolesions form a nidus for such plaque formation are questions that cannot currently be answered because of the inadequate length of this study. Thus, the relatively small numbers and the midterm follow-up dictate that we continue to exercise caution when freezing in the vicinity of the coronary arteries. Understanding well the limitations of studying only 20 patients, our long-term goal is to add to this study protocol. Because our current policy for persistent and permanent AF is a complete biatrial lesion set, we will also include the RCA as an additional site of investigation for potential injury.

Back to Top | Article Outline


In conclusion, our preliminary data from this relatively small cohort suggest that cryomaze is safe with regard to coronary artery injury. Barring one case of partial circumflex stenosis, likely due to the ongoing normal progression of CAD, these data derived from a limited prospective trial suggest that epicardial application during the cryomaze procedure does not cause anatomic or physiological compromise of the circumflex artery. The results of this study, albeit based on a relatively limited sample size, will begin to allay the fears that the transmural mitral isthmus lesions are producing occult injuries to the circumflex arteries. Whereas these data confirm the safety of cryoablation, in no way do the authors advocate the carte blanche freezing of coronary arteries. Caution should be exercised when applying cryothermy in the vicinity of coronary arteries while creating the Cox-maze III lesion set.

Back to Top | Article Outline


1. Rahmanian PB, Filsoufi F, Salzberg S, Coppolino A, Castillo JG, Adams DH. Surgical treatment of atrial fibrillation using cryothermy in patients undergoing mitral valve surgery. Interact Cardiovasc Thorac Surg. 2008; 7: 990–995.
2. Geha AS, Abdelhady K. Current status of the surgical treatment of atrial fibrillation. World J Surg. 2008; 32: 346–349.
3. Cheema A, Vasamreddy CR, Dalal D, et al.. Long-term single procedure efficacy of catheter ablation of atrial fibrillation. J Interv Card Electrophysiol. 2006; 15: 145–155.
4. Weerasooriya R, Khairy P, Litalien J, et al.. Catheter ablation for atrial fibrillation: are results maintained at 5 years of follow-up? J Am Coll Cardiol. 2011; 57: 160–166.
5. Tilz RR, Rillig A, Thum AM, et al.. Catheter ablation of long-standing persistent atrial fibrillation: 5-year outcomes of the Hamburg Sequential Ablation Strategy. J Am Coll Cardiol. 2012; 60: 1921–1929.
6. Prasad SM, Maniar HS, Camillo CJ, et al.. The Cox maze III procedure for atrial fibrillation: long-term efficacy in patients undergoing lone versus concomitant procedures. J Thorac Cardiovasc Surg. 2003; 126: 1822–1828.
7. Neuwirth R, Fiala M, Branny P, et al.. Long term effectiveness of surgical cryoablation for chronic atrial fibrillation in patients undergoing surgery for severe mitral valve regurgitation [in Slovak]. Vnitr Lek. 2007; 53: 151–156.
8. Gammie JS, Laschinger JC, Brown JM, et al.. A multi-institutional experience with the CryoMaze procedure. Ann Thorac Surg. 2005; 80: 876–880.
9. Mikat EM, Hackel DB, Harrison L, Gallagher JJ, Wallace AG. Reaction of the myocardium and coronary arteries to cryosurgery. Lab Invest. 1977; 37: 632–641.
10. Holman WL, Ikeshita M, Lease JG, Smith PK, Ungerleider RM, Cox JL. Cardiac cryosurgery: regional myocardial blood flow of ventricular cryolesions. J Surg Res. 1986; 41: 524–528.
11. Bakker PF, Elbers HR, Vermeulen FE, Robles de Medina EO. Effects of cryothermia during cold cardioplegia on epicardial and intramural coronary arteries. Ann Thorac Surg. 1993; 55: 127–130.
12. Virmani R, Chun PK, Parker J, McAllister HA Jr. Suture obliteration of the circumflex coronary artery in three patients undergoing mitral valve operation. Role of left dominant or codominant coronary artery. J Thorac Cardiovasc Surg. 1982; 84: 773–778.
13. Cornu E, Lacroix PH, Christides C, Laskar M. Coronary artery damage during mitral valve replacement. J Cardiovasc Surg (Torino). 1995; 36: 261–264.
14. Kaklikkaya I, Yeginoglu G. Damage to coronary arteries during mitral valve surgery. Heart Surg Forum. 2003; 6: E138–E142.
15. Pessa CJN, Gomes WJ, Catani R, Prates JC, Buffolo E. Anatomical relationship between the posterior mistral valve annulus and the coronary arteries. Implications to operative treatment. Braz J Cardiovasc Surg. 2004; 19: 372–377.
16. Raza JA, Rodriguez E, Miller MJ. Successful percutaneous revascularization of circumflex artery injury after minimally invasive mitral valve repair and left atrial cryo-MAZE. J Invasive Cardiol. 2006; 18: E285–E287.
17. Rajbanshi BG, Rodrigues E, Lynch JJ, Gulati R, Sundt TM 3rd. Coronary artery spasm after Cryo Maze III procedure. Ann Thorac Surg. 2011; 92: 1884–1887.
18. Haïssaguerre M, Jaïs P, Shah DC, et al.. Spontaneous initiation of atrial fibrillation by ectopic beats originating in the pulmonary veins. N Engl J Med. 1998; 339: 659–666.
19. Doguet F, Le Guillou V, Litzler PY, et al.. Coronary artery dissection after surgical cryoablation procedure. Ann Thorac Surg. 2009; 87: 1946–1948.
20. Berreklouw E, Bracke F, Meijer A, Peels KH, Relik D. Cardiogenic shock due to coronary narrowings one day after a MAZE III procedure. Ann Thorac Surg. 1999; 68: 1065–1066.
21. Wong KC, Lim C, Sadarmin PP, et al.. High incidence of acute sub-clinical circumflex artery ‘injury’ following mitral isthmus ablation. Eur Heart J. 2011; 32: 1881–1890.
22. Calkins H, Langberg J, Sousa J, et al.. Radiofrequency catheter ablation of accessory atrioventricular connections in 250 patients. Abbreviated therapeutic approach to Wolff-Parkinson-White syndrome. Circulation. 1992; 85: 1337–1346.
23. Solomon AJ, Tracy CM, Swartz JF, Reagan KM, Karasik PE, Fletcher RD. Effect on coronary artery anatomy of radiofrequency catheter ablation of atrial insertion sites of accessory pathways. J Am Coll Cardiol. 1993; 21: 1440–1444.
24. Sueda T, Shikata H, Mitsui N, Nagata H, Matsuura Y. Myocardial infarction after a maze procedure for idiopathic atrial fibrillation. J Thorac Cardiovasc Surg. 1996; 112: 549–550.
25. Ouali S, Anselme F, Savouré A, Cribier A. Acute coronary occlusion during radiofrequency catheter ablation of typical atrial flutter. J Cardiovasc Electrophysiol. 2002; 13: 1047–1049.
26. Manasse E, Medici D, Ghiselli S, Ornaghi D, Gallotti R. Left main coronary arterial lesion after microwave epicardial ablation. Ann Thorac Surg. 2003; 76: 276–277.
27. Fayad G, Modine T, Le Tourneau T, et al.. Circumflex artery stenosis induced by intraoperative radiofrequency ablation. Ann Thorac Surg. 2003; 76: 1291–1293.
28. Roberts-Thomson KC, Steven D, Seiler J, et al.. Coronary artery injury due to catheter ablation in adults: presentations and outcomes. Circulation. 2009; 120: 1465–1473.
29. Wong KC, Sadarmin PP, Prendergast BD, Betts TR. Acute occlusion of left circumflex artery following radiofrequency catheter ablation at the mitral isthmus. Europace. 2010; 12: 743–745.
30. Spar DS, Silver ES, Hordof AJ, Torres A, Liberman L. Coronary artery spasm during radiofrequency ablation of a left lateral accessory pathway. Pediatr Cardiol. 2010; 31: 724–727.
31. Paul T, Bökenkamp R, Mahnert B, Trappe HJ. Coronary artery involvement early and late after radiofrequency current application in young pigs. Am Heart J. 1997; 133: 436–440.
32. D’Avila A, Gutierrez P, Scanavacca M, et al.. Effects of radiofrequency pulses delivered in the vicinity of the coronary arteries: implications for nonsurgical transthoracic epicardial catheter ablation to treat ventricular tachycardia. Pacing Clin Electrophysiol. 2002; 25: 1488–1495.
33. Aoyama H, Nakagawa H, Pitha JV, et al.. Comparison of cryothermia and radiofrequency current in safety and efficacy of catheter ablation within the canine coronary sinus close to the left circumflex coronary artery. J Cardiovasc Electrophysiol. 2005; 16: 1218–1226.
Back to Top | Article Outline


This is an interesting clinical series examining whether epicardial application of a cryoprobe over the coronary sinus and the atrioventricular groove could injure the circumflex coronary artery. In this prospective study, 20 consecutive patients with normal results of preoperative angiograms were followed after a cryomaze procedure. Postoperatively, all patients underwent a computed tomographic angiogram (CTA), and, in 95% of the patients, there was a patent circumflex artery. There was a stenosis in one patient. The CTAs were performed at a minimum of 6 months after the ablation, with a mean of 2.7 years.

This study suggests that, during a reasonably short follow-up period, the cryomaze procedure, as described in this article, did not injure the circumflex artery. However, this study has a number of important limitations. First of all, it is unclear whether, in any of the patients, the cryoprobe was positioned directly over the circumflex artery. In fact, in a typical right dominant system, the epicardial cryoablation would not be near the circumflex coronary artery. Only five patients in this series had a left or codominant system in which there was even a possibility of an injury. Moreover, there are convincing data that cryoablation of coronary arteries causes intimal hyperplasia in experimental models (Holman WL, Ikeshita M, Ungerleider RM, Smith PK, Ideker RE, Cox JL. Cryosurgery for cardiac arrhythmias: acute and chronic effects on coronary arteries. Am J Cardiol. 1983;51:149–155. Mikat EM, Hackel DB, Harrison L, Gallagher JJ, Wallace AG. Reaction of the myocardium and coronary arteries to cryosurgery. Lab Invest. 1977;37:632–641). This type of injury probably would not be identified on CTA, particularly at this short of a follow-up period, and would likely result in stenoses years after ablation. It is the strong opinion of the Editor that cryoablation should never be performed over coronary arteries. These arteries can almost always be avoided with careful examination of preoperative coronary angiograms. This study does not provide enough data to suggest that cryoablation is safe around coronary arteries, and great caution should be exercised in their vicinity.


Atrial fibrillation; Cox-cryomaze; Cryoablation; Circumflex artery injury; CTA

Copyright © 2013 by the International Society for Minimally Invasive Cardiothoracic Surgery. Unauthorized reproduction of this article is prohibited.