Lessons From a 5-Year Follow-Up Study

Yay. Ablation works well.

[circ.ahajournals.org]

Researcher,

Many thanks for that.

Am I correct in concluding that after ONE procedure the success rate (NSR at median follow up) was 46.6% and after somewhere between ONE TO THREE procedures was at median follow up 79.5%?? I'm assuming that there's more analysis in the main article that is missing in the abstract pertaining to average success rate at TWO and THREE procedures.

Regards,

Mike

Who carried out the ablations, the doctors listed in the study?

Don't ablation results reflect the competency of any given electrophysiologists?

Carol

Carol - yes, definitely. Jackie

The results don't indicate the operators are part of the "A" team, when compared to facilities on a world-wide basis. Not the "C" team either. More like the "B" team, IMHO.

The Five-year follow-up

[Thanks, researcher for pointing us to the article, with its reporting of results for any unusually long follow up period (but see below).]

The prospective patient who is trying to get a general picture of his career as a CA patient* would want to note the following:

1) The rate of single-procedure success was 46% (74 of 161 Ss).
So there is around a 50% chance that one CA will not be enough for you. And this % is for paroxysmal patients with normally functioning left ventricles (although the Table of Patient Characteristics on p 4 indicates that the AF would not be labeled as "lone" ).

2) Sixty of those who failed had a second procedure and twelve of these had a third procedure (78 of 161 Ss or 48%). These additional procedures put an additional 33% (53 Ss) into NSR for some number of years from .33 to 5.5 (!). The remaining Ss exhibited improvement, defined as " either 90% reduction of symptomatic ATa (recurrent atrial tachycardia) before the last follow-up and off AAD or 90% reduction of symptomatic ATa while taking previously ineffective AAD."

So there is a very good chance* that you will be free of AF (and AT) if you are willing to have 1-3 CA's, and that however many you need will almost certainly result in clinical improvement. There are also indications that this intervention will prevent progression from paroxysmal to chronic AF.

* I am not sure what to make of the statement that the median follow-up was 4.8 years -- with a range of .33 to 5.5 years (!). Certainly the distribution would be skewed towards the longer durations. Perhaps I missed this detail.

The question for you is, "How can I improve my odds of success over those found in this study?"

One is to make sure that you are effectively experimenting with supplements and life-style changes, as described many time in posts and other materials available on this site.

Another is to seek out top-tier EPs (as suggested by Carol, Jackie and GeargeN).

The other is to consider surgical interventions that *might* reduce the rate of PV-LA reconnection that is found in the great majority of AF recurrences.

I am referring to both hybrid or convergent approaches, in which the PV isolation is done in a way that should result in lesions are reliably transmural and *may* not break down as readily.

I am also thinking of TTM, which is (relatively!!) minimally invasive surgery which accomplishes a more complete lesion set.

These days, I believe that the costs and benefits of CA as described in the "follow-up" study and the potential for better results from convergent and TTM mean that the latter procedures deserve your consideration when first you survey treatment possibilities for AF.

-- Dick

P.S. There was a discussion of dealing with AF or AT (or ATa) that occurs during the 3-month blanking period on p 3 of the article, which should interest the many of you who have been concerned by these occurrences.

Actually, the results are pretty impressive but that would be as expected as Kuck's group in Hamburg is as good as any of the top groups including Natale. If you take the single procedure results as is and assume a log linear decline Kaplan-Meier model, the results imply 12 month efficacy of 85.3% for single procedure without AAD for methods dated 2003-2004.

The equation for estimating one year efficacy is as follows:

E (12 months) = 0.466 ^ (1/4.8)

My other thought is that I don't know of any other study from top groups that tracked patients that long, never mind less than top groups. Also, it is great to see so little progression to chronic AF. If we include the second and third procedure remediations, the results are even more encouraging. I think most folks would be really happy with a near 80% NSR rate at 5 years even if it takes 3 tries to get there but those 3peats are a small fraction compared to 2peats.

I forgot to mention in my last posting that the standard way of expressing success rate is % in NSR off AAD at 12 months. That was why I back estimated to 12 month efficacy for comparison purposes with other quoted success rates. Again I think it would be great for all the EP ablation centers to track patients for 5 years and to use that as a metric of how their EP specialists are doing in comparison to others.

As my mother-in-law used to say, "Everyone sees things from their own seat in the theater."

In considering research results, I often attempt to put myself in the position of prospective patient ("PP") who is searching for the best way to deal with his AF. I would suggest that the PP wants to know, “How can the results of this research help me to deal with my AF?” Or, “What is the likelihood that if I take a particular course of action a certain result will occur?”

I would suggest that the goals of the researcher is not to design a study whose results will apply to the individual “you”. He may be trying to advance knowledge in some area without any concern about practical application; or, he might be interested in how to allocate health resources in a way that best serves large groups or populations. So the PP may have to do a bit of work...

Here are some things the PP can do:

He might start with determining how closely he resembles the research subjects. The Ss in this study are described in Table 1atient Baseline Characteristics. The nature of these characteristics suggest that the results may apply to those whose AF in not “lone”, but that they may not if his LV is not functioning well (these Ss were excluded from the study).

Another question is “What were the treatment factors that were essential in causing its effects?” and “Do I have access to such treatment for myself?” I won't go through the details of treatment except to note the use of three-dimensional mapping and an irrigated catheter, both of which techniques have undoubtedly improved since the study was done; and, the circumferential strategy was used as opposed to the segmental approach – the authors cite a study in which this approach was superior to the segmental approach, although I believe more recent studies suggest no difference.

A factor that may have made a difference (as mentioned by others here) was experience level or level of expertise; there was no attempt to assess these. Experience level can of course be measured; the only way I can think of to measure expertise directly is to have another expert looking over the shoulder of an EP being evaluated and making judgments on the moves he makes in response to various ECG signals or other factors.

So, it may be that results would be better, if techniques have improved and you can find an experienced/expert EP.

The last set of questions has to do with the way the independent and dependent factors are measured and the way the results are presented, including the statistics used.

The reliability of measuring the presence or absence of AF is always an issue, what with the possibility of silent AF.

I do not have an idea as to what the numbers/%'s of Ss would with differing outcomes if the same study were done x times. Wasn't this study rather small compared to other outcome studies? And mightn't there be some clinically significant differences found in the %'s reported?

In other words, what level of confidence are we to place in the reported percentages? (It is encouraging that other somewhat similar studies have reported similar outcomes.)

I don't see how Kaplan Meier helps with this type of error.

So the PP will come away from some questions for his doctor about the reliability and validity of the findings and how these might apply to him. Most important, the answers given by the cardiologist or EP will be based on finding from several studies, which would help estimate the range of outcomes for each time period and number of procedures.

Some other comments:

The reaction of the PP to the stated results I think will depend on a number of factors: his hopes for change in QOL, his expectations, how he defines failure, the extent to which he focuses on and how he labels physical symptoms and heart issues (i.e. the cost of remaining in AF and/or continuing to rely on medications), whether he is interested in the best possible treatment or whether he has an absolute probability level that would satisfy him. I would expect PPs to vary in how they would react emotionally to the statistics reported in this study (how people respond to various degrees of risk is a mystery !!!), but there is sure to be variation. These are questions that the PP can try to answer for himself.

This study helps realign PP's expectations by towards thinking of CA treatment as ordinarily two phases so that if more than one procedure needed, this does not mean that the first was a "failure". Labeling CA treatment as a "CA series" would help to dispel this expectation.

The other favor it does for the nervous AFer is to give some guidelines for what to think of and how to deal with AF (or AT or ATa) during the 3-month blanking period. results suggest that a successful CA series may reduce the probability of progression to chronic AF .

--Dick

The Kaplan-Meier model description is simply a plot showing % remaining in NSR on the Y-axis vs time to first AF recurrence on the X-axis. It is the standard for measuring effectiveness for either AAD or ablation. In the vast majority of data presented in papers, the data can be fitted well with exponential decay curves (same as log-linear). I think the important question to ask an EP or hospital when getting a success rate quote is how long of a period they use. The HRS guideline is 12 months in NSR off AAD.

With regard to
================================================
This study helps realign PP's expectations by towards thinking of CA treatment as ordinarily two phases so that if more than one procedure needed
================================================

I absolutely agree. If you speak with Natale's group or other experienced groups, you will in fact be told to expect that. Kuck et al results suggest that there will be a 50/50 chance of requiring a touch up procedure within 5 years even in the best hands as of 2003-04 circumferential PVI approach. I think that the drastic reduction in progression to chronic AF is by itself a terrific result.