STS Guidelines Wrong to Call for More Surgical AF Ablation

John Mandrola, MD


January 13, 2017

The Society of Thoracic Surgeons (STS) recently published an extensive literature review and guideline statement on surgical treatment of atrial fibrillation.[1] The document was clearly written and well referenced. Its problem was the conclusions.

The authors, largely respected and experienced surgeons, call for more surgical ablation of AF. Their take of the literature is that surgical ablation is safe and provides long-term rhythm control; thus, we should do the beneficial procedure on more patients.

I strongly disagree—for three reasons.

Weak Evidence Base

The first problem with their conclusions is the literature itself. It's unconvincing. As an example, consider the evidence base for surgical ablation at the time of mitral-valve surgery. I chose this category because the authors give it a class I, level A recommendation.

Several of the cited trials of surgical ablation vs none are single-center studies with small numbers of patients, often less than 100 patients total. Most of these studies were published a decade ago and included only 1-year follow-up in which AF was assessed by a single ECG or Holter. Interestingly, as the electrophysiology community moves to more aggressive AF monitoring with implantable loop recorders (ILR), the surgical authors write, "[These ILR] systems have their own set of interpretive challenges and infrastructure requirements that may not significantly [affect] practice." To support less aggressive AF monitoring, the surgeons cite a study of 47 patients that found no difference between Holter and ILR recordings.[2] AF monitoring aside, these small studies yielded no useful data on long-term clinical outcomes such as rates of stroke or mortality.

The Cardiothoracic Surgical Trials Network (CTSN) investigators led the largest and most contemporary trial of concomitant surgical AF ablation during mitral-valve surgery.[3] This non–industry-sponsored multicenter randomized controlled trial compared surgical ablation or no ablation in 260 patients with persistent or longstanding-persistent AF. They found no statistical differences in 1-year mortality (6.8% vs 8.7% for the control group, P=NS). At 1 year, more patients in the ablation group were free from AF on a 3-day Holter monitor (63.2% vs 29.4%, P<0.001), but this did not lead to improvement in functional class or other quality-of-life measures. Patients in the ablation group were nearly three times more likely to require a permanent pacemaker (21.5% vs. 8.1% per 100 patient-years, P=0.01).

Surgeons from the Czech Republic reported nearly identical findings in a smaller trial of concomitant surgical ablation vs no ablation.[4] In this study of mixed valvular and coronary operations, ablation resulted in fewer AF episodes at 1 year but no difference in the rates of mortality, stroke, or heart failure. Pacemakers were required in 6% of ablation group vs 1% of the control group (P=0.07).

These studies confirm the results of many smaller studies and propensity-matched observational studies showing that, yes, surgical ablation can reduce the number of AF episodes noted at 1 year, but that does not translate into clinical benefits. Plus, surgical ablation comes at a cost. One cost is that the procedure adds almost $900 in surgeons' billing. (I know this because makers of the ablation tools conveniently post the reimbursement schedule online.) Another cost is a higher rate of pacemaker implantations.

The disconnect between suppression of AF episodes and clinical outcomes isn't surprising. The medical community has not yet decided on the strength of AF episodes as a surrogate marker of clinical outcomes. Witness the lack of temporal relationship between AF episodes and stroke. Many paragraphs in the STS guidelines make the case that having AF is bad and therefore fixing it is good. The problem is that there is no evidence that reducing AF episodes by drugs or ablation changes the disease course or reduces clinical events. Surgeons and electrophysiologists alike have trouble defining "success."

Lack of a Standard Procedure

My take of the surgical literature is that they haven't worked out a standard approach. In a 2010 editorial[5], aptly named "The longstanding, persistent confusion surrounding surgery for atrial fibrillation,"surgical pioneer Dr James Cox (Washington University, St Louis, MO) called the simultaneous introduction of new lesion patterns and new energy sources a violation of a cardinal rule of science. Namely, that all but one variable in an experiment be held constant. Cox wrote that the common practice of incriminating the energy source in surgical ablation failures rather than the lesion pattern was a "result of the influence exerted by industry on the mind-set of surgeons."

Here is my understanding of the complex history of surgical AF ablation: The cut-and-sew Cox-Maze III operation delivers the greatest freedom from AF. The problem is that it takes time (on pump) and expertise. The development of new tools, such as radiofrequency clamps, cryoablation probes, and even microwave energy sources led to the ability to approximate the cut­-and-sew procedure and reduce pump time. Putting new tools in the hands of innovative surgeons, however, has led to heterogeneity in lesion sets and energy sources. Some do right atrial lesions, some don't. Some do pulmonary-vein isolation alone, some do the full Cox-Maze. This hodgepodge of techniques complicates our understanding of the results.

In the CTSN[3] study, for example, the surgical ablation group was split into pulmonary-vein isolation only vs full biatrial Maze. In a group of patients with advanced mitral-valve disease, both techniques performed equally—an observation that surprised the surgical community. In a letter to the New England Journal of Medicine, Dr Cox called this finding "not valid" because the surgical technique was incomplete.[6] Most electrophysiologists would find this equivalence less surprising given the results of the STAR AF II trial. [7]

If surgical leaders can't agree on the best technique(s), what does that entail for translation of this procedure to the broader community where most heart surgery is done?

Surgical Ablation Requires Expertise

Here's a sentence from the STS guideline statement that included five references: "Based on available data, current ablation techniques are safe and should be applied at the time of open atrial procedures, even for high-risk patients." Four of the five referenced studies came from a single expert group of surgeons in the US, while the other came from one expert group in Europe.[8–12] That's the problem. It's hardly controversial to ask whether the safety and efficacy of surgical ablation reported in the literature can be replicated in the real world. The surgical authors did not cite a single contemporary database study supporting the translation of these procedures outside of expert centers.

At the risk of sounding like a neurosurgeon at an anticoagulation symposium, my experience with contemporary surgical ablation is not reassuring. I've done many redo procedures in patients who have had surgical ablation (from multiple centers). I find that the pulmonary veins aren't isolated and the mitral and cavotriscupid isthmus are only partially ablated. Often, there is significant delay of left atrial appendage activation—which suggests misplacement of lines. A basic rule of physiology is that partial ablation in the atria sets the stage for reentry. It's much easier to manage atrial fibrillation than an injured atrial flutter.

In a 2014 editorial, surgical ablation expert Dr Ralph Damiano (Washington University) echoed these concerns.[13] He noted the importance of surgical expertise and institutions' investment and commitment to obtain adequate follow-up. Damiano, a colleague of Dr Cox, also emphasized the importance of doing a full Cox-Maze lesion set. Despite the CTSN study,{3} he believes limited ablation is "less efficacious."


Considering the weak evidence base for surgical ablation (including no evidence of long-term clinical benefit), the lack of a standard approach, and zero published evidence it can be performed successfully outside of expert centers, only one conclusion is possible. We should not encourage more surgical ablation.

What we should encourage is more study of this technique. If I were in charge of healthcare, I'd allow this procedure to be done, but only by surgical groups who publicly report their clinical outcomes. I'd say no reimbursement without data. That would encourage surgeons to 1) record their results and add to our knowledge base, 2) learn to make complete lines, and 3) collaborate with electrophysiologists in the wise selection of patients.



Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.