Video Transcript:

My topic is about who should operate, which is always politically charged somewhat since none of us like anything to do with regulation, at least some people in Washington do but I do not. On the other hand, this field sort of requires this question to be answered and there is a lot of data that is going to show that; my disclosures really do not have anything to do with what I am going to talk about. Let us start again with a question.

An otherwise healthy 50 year-old asymptomatic man has severe MR and preserved left ventricular function. Which of the following is most relevant in deciding on whether to proceed with mitral valve surgery at this time? So let you look at the echo, and here are the possibilities; a, accessibility to a very high volume mitral valve surgery center; b, age of the patient; c, high probability of repair based on echocardiographic lesions, the patient is a potential candidate for warfarin therapy, pulmonary artery pressure documenting to be 42/28.

So, let us have the answers. Everyone, pick one.

Okay, let us see the answer. So, it is interesting, you see 58%, I think have what is the right answer, which is that you have an asymptomatic patient and the guidelines are clear that if you do not think you are going to repair the valve with a very high probability, you should continue to do followup. We could argue about exercise or we could introduce the concept now of exercise echo to see where the PA pressures go but I do not think that answer A, which is what I predicted would be the second most common, is exactly right and I will show you why in a second. Can we go back to the slides?

So the guidelines again are very, they like to talk about mitral valve repair a lot and I like to show this algorithm a lot, I show it actually to every patient that I operate on. I put this picture up to walk them through why they should be considering surgery or why they have been referred and I put a double circle around the question mitral valve repair likely and I think this is the fundamental question that we have to start taking more seriously. The days of saying, well we have got these boxes and we are going to try and do a repair and when we got there we were surprised. We should not be that surprised anymore. I think the echo lecture this morning and the imaging lectures really show us that. We are doing better. This paper was just published from Jim Gammie, it is looking at 60,000 or 70,000 patients in the STS database, a large number of patients, and you can see I have put some blue up there what I consider still is the gap. Meaning that the gap has narrowed some in terms of, and I drew the line at 90%, well just for the sake of argument use the ACC guideline that 90% repair should be a goal. Again, several centers are in the 95% to 96% repair range depending on the lesion or higher, but you can still see there is a significant practice gap. I do not think in oncology, you would accept a 20% gap in terms of class I care, so I think we do have a significant gap and we need to think about how we are going to address it and of course the thing that always comes up is volume and skill, so let us look at that. Gammie looked at it in this paper in circulation that is sort of an earlier time period in the STS database. Again, they looked at a large number of patients, this time about 14,000 patients and they made a couple of intuitive observations but they are always worth reemphasizing, that is there is a relationship to volume versus repair and that of course makes sense and this was again data sort of in the early part of 2000-2003. You can see that higher volumes centers were sort of getting in the 70% to 75% range at a time where the average repair rate was in the 30's, so it was still double. The other observation that was made, which I do not think gets enough emphasis, is the decreasing mortality rate associated with volume.

The more you do of this operation like most operations, the better you get at recognizing the 2% or 3% of catastrophes that are about to happen, so I do think that it is not just your repair rate but I do think there is going to be a safety issue when you do very few mitral operations versus when you do a lot of them. We should not be surprised by that. The other issue about volume I think was addressed in this excellent paper from Cleveland which had a lot of data in it about degenerative repair and rates and various predictors and all kinds of things, but one thing I thought that was interesting to point out was that it is not just going to the Cleveland Clinic. It mattered a lot which surgeon you got at the Cleveland Clinic and I see lots of patients that tell me we may go to the Cleveland Clinic and I say who is the surgeon and they say we do not know yet. That is true, and I am sure the cardiologists here, some of you have had that conversation, but it does matter quite a bit because some surgeons at Cleveland were predictive for more replacement and one surgeon or some surgeons now I am sure are predictive of more repair, so it is not just volume or center, it is something else. There is something else there. This concept and Craig Miller actually sent me this paper after a meeting about two to three years ago. He e-mailed it to me. I was getting ready to take off on a plane and it was a paper I thought that he had recognized and I have since recognized was very provocative By Ben Bridgewater talking about best practice standards. It was the first time someone had actually put pen on paper and started to define this and one of the things they were saying in the United Kingdom was; surgeons should do 25 repairs a year and centers should do 50 a year, so at least we are starting the concept of volume. They were not getting into lesions yet except a little bit with Barlow's but at least they were starting to put parameters on volume and I will show you, this is data from 2008. It has just come out in the last two or three weeks and we actually were asked to write a commentary in this. It is published actually as a book. A commentary about their mitral data, so I had a look at this beforehand and I just want to point out that their experience in the United Kingdom where they track and actually unlike the STS where they do not differentiate at least yet, we will in the future be able to differentiate the type of etiology versus the operation in the STS database, those valve elements have been worked on by our committee including Steve and several people in this room and that will be coming so we will get more information. They do already, so they are not validated but they do at least ask people to check the box degenerative or non- degenerative and so this is the degenerative surgery arm and you can see that just like in the United States, repair is going up, replacement is coming down. Their repair rate now is a little over 60%, so they are looking a lot like the United States in terms of repair versus replacement for degenerative disease, but this is the figure that shows volume versus repair rate and look at the percentage of repair and the number of operations and what you can see, which you would expect, some low volume centers have very low repair rates and most of the high volume centers have high repair rates but regardless of whether you want to take the highest volume centers, you can still see there is an outlier that is very low and even when you get in this sort of middle category, I tell you there is a big swing between almost the highest repair center and one of the lower repair centers and the difference was only 20 to 30 cases. So, it is not just volume. There is something else to it obviously and of course that is going to be skill and I think what we have to do is say okay what else do we do to narrow this practice gap. One thing of course could be concentrated volume but there is certainly going to be more. This was their summary of their findings and this was an eight year study, so this is all inclusive, all patients that had mitral surgery in the United Kingdom over eight years, many patients. Another corollary by the way which I am sure we see in the United States and we do not talk a lot about, one of the reasons patients are not referred in a timely fashion to surgery and Bob will bring up in the discussion this paper just published by David Bach showing at the University of Michigan patients having indications for surgery not being referred for surgery.

One of the reasons I think globally patients are not referred is because they are worried the surgeons are not going to do a good operation, so you wait for the patient to get sicker and sicker and a lot of these patients were referred when they were class III and class IV and in fact a majority of them in the United Kingdom over that eight years were class III or class IV. Asymptomatic patients were rarely referred there. The mortality rate just like we saw in Gammie's paper was low or just one of the other elements Gammie had shown was lower for repair versus replacement. It is probably safer to do valve repair in most patients. Again repair rates crept up to 67% over the time of the study and there was a wide variation between hospitals performing repair versus replacement. So, Ani coined a phrase about postcode lottery. It should not matter within an 80-mile radius where you live as to which operation you do but today in the United Kingdom that is in fact what happens. Well, the other thing is this complexity of lesions. We talked about skill and Bob introduced that in the last set of guidelines and my partner Ani and I wrote this last year in the JTCBS and we are writing about an article that was talking about recurrence of MR and we were saying that basically complexity of the valve that you are confronted with and the type of repair you do is probably the fundamental thing we have not talked enough about to narrow this practice gap, so we are very interested in the concept of trying to differentiate degenerative disease and for you cardiologists that are here the next time you see a patient try and sort it out, is it single segment FED. This is the type of disease Craig was mentioning and these are older patient that go to Mayo Clinic and they are in their 60s and they have maybe unrelated two vessel coronary disease, but they generally have simple pathology and most of those patients by the way have simple triangular resection. They are not very complicated repairs. These are the patients Craig was saying and in a quaternary level, mitral valve referral practice in Northern California. This is what you get. You are happy the day you see P2. Most of the time you are seeing Barlow's patients, young, asymptomatic people, multisegment disease, giant leaflet segment distention, deep clefts, calcified anterior papillary muscles, you see these fissures, which is the beginning of displacement of the hinge of the valve toward the atrium and off the atrioventricular groove that leads to these micro-fissures that probably evolve into calcification. Totally different, you do not have to be a surgeon to know which one of these two valves you would like to try and repair. There are very different valves and again with imaging, it should not be a surprise anymore, which valve you are going to the operating room with. Keep in mind and I will show you in a second most repair rates in most centers were these two lesions, type I and anuloplasty, AFib the most common etiology are type II P2, a simple posterior leaflet prolapse which is about 60% of mitral valve repair and then all those pictures that I showed to you, 60% of the valves you would actually see just in a 100-case practice in a regular hospital would be this type II P2, not quite sure why that is the leaflet that actually prolapses most commonly but those valves get repaired very frequently in centers and you can read between the lines in papers if you want to understand this. So don't look at papers that are talking about what we repair. Go read the literature and read between the lines. Here is a paper that was talking about 3D echo and how 3D echo can accurately predict what the surgeon saw that they had a very high agreement between the preoperative 3D and what the surgeon saw but what they also provided not intentionally but just in the paper was the repair rates and you can see the overall repair rate was 50%, posterior leaflet was higher, bileaflet lower, anterior leaflet in this series is very low. So again, there is a clue. Posterior leaflet repairs are simpler to do, they are going to be done in maybe not a reference mitral center much more efficiently, anterior leaflet particularly in this center. There is a message there. Same thing, here is another paper, it doesn't matter what they are talking about, this one was talking about minimally invasive surgery doing side surgery incredibly safely but essentially no mortality, very high volume over 150 patients over an 18-month period published just last year, very current series. Overall repair rate is 66% but that is what you cannot tell the patient.

You have to look at the echo, turn the color off because this patient has posterior leaflet prolapse and in this center it's a 96% repair rate. Anterior leaflet while the numbers are really tiny but bileaflet repair, I think we can hang our head on that 20%, huge difference whether there was bileaflet or posterior leaflet prolapse, so the message there is not just volume, this is a very high volume center but we have to start thinking about lesions and there is a spectrum of disease and you go from simple prolapse, thin cord, non-distended tissue to the prolapsing segment developing some myxomatous change, relatively simple, then you have the sort of grey zone between fibroelastic deficiency and the true Barlow' disease which is the giant multi-segment valve and another way to think about that is that, that is the truth and you have an increasing repair difficulty as you move along this spectrum. You have got to respect rules and each one of these valves but there is no question that the more segments that are involved and the larger the valve is and the more calcium you may have and the more cordal elongation as opposed to broken cords you have, the more difficult the valve gets to fix. We published this last year talking about how cardiologists could increase the role of mitral valve repair, you are not doing the surgery but you could have a big impact on how we see that practice gap keep getting narrowed and the way you could do it is simply by saying FED or Barlow's. Single-segment, multi-segment, posterior leaflet versus bileaflet or anterior leaflet, because there is no question the likelihood of repair is going to go up substantially when you start matching the lesions and the etiology to the skill of the surgeon, and if you have Barlow's or bileaflet prolapse or complex anterior leaflet prolapse, unless you are with a reference mitral surgeon, these patients will have a high replacement rate and the current literature continues to support that. Again back to this paper from Bridgewater, there were a lot of other concepts in here that are worth mentioning and I like this one. Surgeons undertaking valve repair should subject their results to regular audit. I live in a state where things are audited and actually it is not very bad when you do not live there you sort of think, Gosh, what would that be like. It is not very bad at all. Frankly, it is healthy and I think that in this particular field we need to have some more careful monitoring of the results of repair rates, mortality rates, and quality of repair and audit should include residual and recurrent regurgitation after repair. That is another little buzz where cardiologists say now we are seeing more repairs but we do not like what we are seeing. We have to audit that, so we tried to define that in this article a little bit more. I just highlighted it. If your repair rate is less than 90% that probably means you need to differentiate more complex cases and get them out of your system and the other thing is early recurrence of MR and I will talk about that this afternoon, those are usually technical failure. If that is over 5% that is a problem. Sure, you are going to have some patients where you have a technical failure, we all do but it should be very small, a technical failure rate of 10% to 15%. Again, that suggests; refer more patients out. Another Bridgewater concept that he started was that some surgeons should have sort of super specialty practices and probably the subset that is for they mentioned to were the pan-prolapse they called, or Barlow's valve disease. So there is some logic to that and Bob mentioned this paper in his last talk, Kang, there are three of them now but the bottom line is we move towards asymptomatic surgery, the guess work about repair has to be off the table. I do not think that will change in the guidelines. If you are asymptomatic, these discussions with the patients are all predicated on an extremely high repair rate and that may be provocative. Craig may not agree with me but we said it, in the asymptomatic patient you need to approach 100%, not your whole series but in asymptomatic patients, you need to approach 100%. We know the mortality is going to be less than 1%, it has to be. It was in Gammie's analysis in the STS database. Move away from the argument of mortality, very few patients thankfully do not die any more from mitral valve repair. That is not a good quality indicator anymore. We have got to talk about repair rates and residual MR.

The last thing I will show you is this paper, just published this year from Fred Moore's Group in Leipzig because this is real world. These were all surgeons in the group. It is a very high volume center and Fred, I think he runs a meticulous research department and I think his honesty is well known among his peers. This is really what can happen in a real reference center. Overall, repair rate is 94%, posterior leaflet 97%, and anterior leaflet repair rate at 91% and bileaflet 90%, so this is a large series over 1000 patients. All the surgeons in that group by the way, or at least most of the surgeons, they were participating. So, it is possible in our own group. My research fellow, Javier Castillo just presented this in Berlin. Again, we have really been able to get over 95% repair in all of these 550 consecutive patients now, so it is coming. These degenerative patients can have mitral valve repair and this number of 90% is achievable and it is achievable in lots of places and it is achievable in reference centers now. Pat McCarthy wrote this paper earlier this year talking about when is your surgeon good enough, when do you need a referent surgeon, and I will recommend you for those that are interested to read it because it is an excellent summary, not only of the issue of repair versus replacement but also the issue of incisions which has come up more than once this morning already. I think Pat does an excellent job of discussing that. Craig mentioned this and so did Bob about durability of valve repair, should we repair all valves because some of them fail. Well, that is fair enough. Maybe there is 5% or 10% of these complex Barlow's that I happen to like to try and repair them and I think we have an obligation to really follow them carefully to make sure that those repairs are durable as well. There is an issue with the repair but please do not say well but this was Tyrone, so it had to be an excellent repair. Not necessarily true. I am sure it was an excellent repair for the time, and I bet Tyrone is a much better mitral surgeon today than he was in 1995 when he was doing those operations. I know I am. These were the repairs I was doing in 2004, they did not leak, cardiologists were happy. I frankly would not show that in the meeting anymore, I mean that does not look like a very pretty mitral valve repair. It looks like the posterior leaflet still has a fat lip. It has a very asymmetric closure line, not nearly aggressive enough taking out tissue to let the anterior leaflet completely coapt versus today where we are always looking for symmetry and perfect position along the posterior part of the valve and you can get there every time and one of the things we like to do is work backwards, show people bunch of pictures and then show them no matter what the pathology, you can get to that closure line, so Craig repairs are getting better and they are going to be more durable. This book is a good one. It is talking about outliers, what makes people special, and Gladwell talked about 10,000 hour rule, you know, doing the same thing over and over and over is how you really get to be perfect and it is true and this is the reality. This is not that common a disease. Mitral valve repair is not performed that widely among every single member of the society of thoracic surgery, so keep that in mind when you are looking at mitral valve disease. Some are simple, some are very complex. It is not just how much regurgitation there is but what type of regurgitation it is and what skill level do we need. This is not about Ivory Towers and send every patient there; I really do not want you guys to take that message home. The message is identify the etiologies and the lesions and fifth point is; develop local expertise through targeted referral. It is not who is on call that weekend. Every person in every private group or in every big hospital has to have someone that is interested in mitral valve repair and those are the surgeons who should be doing those operations. The question of who should operate right now I think the simplest way to think about it is; you go across the complexity of the disease, identify your city or regional or national reference surgeon.

Thank you very much.