Comparison of Cox-Maze III and catheter ablation

ROCHESTER, MINNESOTA. The Cox-Maze III procedure is often described as the “gold standard” for curing atrial fibrillation (AF). The procedure uses a cut-and-sew protocol for creating lesions forming a “maze” that conducts electrical impulses from the sinoatrial (SA) node to the atrioventricular (AV) node while, at the same time, interrupting any “rogue” circuits. Creating the full set of maze lesions involves open-heart surgery and the use of a heart/lung machine (cardiopulmonary bypass). For this reason, the Cox-Maze has been primarily used in afibbers who needed valve surgery or other procedures requiring access to the outside of the heart.

Although the cut-and-sew approach is still used by some surgeons, it has largely been replaced by the use of radiofrequency-powered devices to create the maze pattern. The main advantage of the original cut-and-sew is that it ensures “transmurality”; that is, the creation of lesions through the entire thickness of the heart tissue.

A group of cardiothoracic surgeons (NOTE: No electrophysiologists were involved in this study) from the Mayo Clinic now report on a study comparing the efficacy and safety of the Cox-Maze III cut-and-sew procedure (CSM) with that of radiofrequency catheter ablation (CA). The study involved 97 lone afibbers (no structural heart disease) who underwent CSM during the period 1993 to 2007. Average (median) age at time of surgery was 56 years, 68% of patients were male, and 69% had paroxysmal AF. The outcome for this (CSM) group was compared to that of a group of 194 lone afibbers who underwent CA during the period 1997 to 2007. Median age at time of ablation was 54 years, 71% of patients were male, and 71% had paroxysmal AF. The standard CSM procedure was used throughout the period 1993 to 2007, while 4 different versions of the CA procedure were used.

CA Procedure
# of Patients
Right-sided ablation only
Focal AF ablation
Segmental PVI
Circumferential PVI

NOTE: The use of right-sided ablation only and focal ablation is now considered obsolete. At time of last follow-up (average of 5.6 years for CSM group and 3.1 years for CA group), the outcomes were as follows:

CSM Group
CA Group
Complete success(1)
Partial success(2)
Still on warfarin

(1) Freedom from AF without use of antiarrhythmics
(2) Freedom from AF with use of antiarrhythmics

It should be noted that 24% of the patients in the CA group required a second ablation during the follow-up period, while 4.6% required a third procedure. In the CSM group, 6.5% underwent a catheter ablation to fix a failed maze procedure.

The procedure-related complications observed in this study are substantially more frequent and more serious than I ever recall seeing before.

CSM Group
CA Group
Respiratory failure
Renal failure
Wound infection
Pacemaker installation
AV node ablation
Pulmonary vein stenosis*
Pericardial effusion
Vascular access problems
Total complications

*More than 50% reduction in vein diameter. NOTE: In their attempt to deal with the procedure-related pulmonary vein stenosis, the surgeons performed 18 balloon angioplasties and 11 stenting procedures in 14 patients.

The authors of the study conclude that, “Compared with catheter-based ablation the Cox-Maze procedure results in greater freedom from AF and less medical treatment with antiarrhythmic drugs and warfarin anticoagulation during follow-up.”
Stulak, JM, et al. Ablation of atrial fibrillation: comparison of catheter-based techniques and the Cox-Maze III operation. Annals of Thoracic Surgery, Vol. 91, June 2011, pp. 1882-89

Editor’s comment: This study would appear to be an attempt to justify the use of the highly invasive Cox-Maze III procedure to treat patients with lone AF. The efficacy and safety of the procedure compared with that of radiofrequency catheter ablation is highly favourable of the maze and very much biased against catheter ablation. Thus, it is not surprising that members of the Mayo Clinic’s electrophysiology department refused to participate in the study and to act as co-authors of the article.

The complete success rate (no AF, no antiarrhythmics) of 56% achieved after an average of 1.3 CA procedures per patient is clearly very low when compared to the 70 to 90% final complete success rates now achieved in world class ablation centres. The total complication rate for the CA group at 26% (major complication rate of 24%) is clearly totally out-of-line with that reported in other studies. A study involving 517 patients who underwent RF ablation at Johns Hopkins in Baltimore reported a total complication rate of 5%, of which, pulmonary vein stenosis accounted for 0.2%, and pacemaker installation and AV node ablation accounted for 0%. A study involving 400 patients who underwent RF ablation at the Cleveland Clinic reported a major complication rate of 1.6%, of which, pulmonary veins stenosis accounted for 0.25%, and pacemaker installation and AV node ablation accounted for 0%. A study involving 1642 RF ablation procedures carried out at the University of Michigan reported a total complication rate of 3.5% with pulmonary veins stenosis accounting for 0.1%, and pacemaker installation and AV node ablation accounting for 0%.

Clearly the success rate is unacceptably low and the complication rate unacceptably high in this study of the efficacy and safety of radiofrequency catheter ablation as performed at the Mayo Clinic. The complication rate for the Cox-Maze is also unacceptably high making it a very poor choice for lone afibbers. It is not clear who actually performed the catheter ablations discussed in the report, but it is safe to say that whoever did them were not competent to do so. The Mayo Clinic (Rochester) does have very competent electrophysiologists on staff, notably Dr. Douglas Packer.