Discussion

Presenter: Dr Dror Ben Leviner Dr George Tolis (Boston, Mass). Congratulations on your presentation, and thank you very much for making the paper available about 2 months ago so I had plenty of time to review it. I have a disclosure. I started doing off-pump in 2002, and I was doing about 95%, and now I do about 99% on-pump. So, that's my disclosure before I start your discussion. And the main reason I did that was basically 3 reasons. When I would see a LAD and I could see the apical LAD very well but the more proximal LAD was hidden under fat and so on and so forth, I just didn't want to deal with cutting the fat and digging inside the muscle, so I would put the tip of the LIMA to the apical LAD. When someone had a very tight [inaudible], I would not want to isolate the RCA because that's when all the V-tach starts and you've got to start pacing, and then you chase the vessel down toward the cava, so I would just lift the apex and put the vessel on the PDA. And then when there was a nice juicy ramus right next to the left atrial appendage, it was really difficult to do that and push the appendage out of the way, so I would always graft the OM after 1 or 2 bifurcations closer to the apex. The point I'm trying to make is that even if you're an excellent technical surgeon, on off-pump, you have to graft further distally than you do when you do an operation on-pump. You can do a great job with a more distal vessel and the op note reads exactly the same, but I'm not sure the operation is exactly the same. My question to you is—there's 2 questions that I have for you, and the first one is do you really think that what you've shown here with the statistics is that on-pump surgeons graft their vessels more proximally than off-pump surgeons? And two, in terms of the propensity analysis that you did, are you sure that you're comparing apples to apples? Because like you said, this is not a randomized sample, and I'm not sure that you can take into account patients that someone wanted to do an off-pump and when they saw that the vessels were smaller, they can't isolate them well, they didn't convert, but just decided to change their intent to treat into an on-pump because if that's the case, then that would tilt more toward the favor of off-pump because you're grafting smaller vessels by definition when you're doing on-pump. So, these are my 2 questions but again, congratulations on your talk. Dr Dror Ben Leviner (Haifa, Israel). Thank you, and thank you for the thoughtful questions. So, regarding your first question, I don't know. And as you said, the op notes and for this matter, the Excel tables with all the data don't say where the graft was done. So, distal—unless of course, we have data on grafts or to the RCA itself. I can't remember off the top of my head what the differences were for RCA on- and off-pump. But I can answer what I think the main question is, and there are people here who can answer that much better than I do, and this goes back to Dr Taggart's comment. Should we be grafting very distal or should we be grafting proximal even if the disease isn't proximal? I definitely don't know the answer and there's much smarter people than me in the room because what we say, CABG is better because we graft distal, and no matter how much progression of disease there is, we still get that disease. So, I don't know what the answer is to that. Dr David P. Taggart (Oxford, United Kingdom). If I could just add before we move on to another question, so the specific question you asked regarding the difference in flows and on-pump and off-pump, at my practice, we rarely are in a situation where we're grafting very proximal. We graft in the same place whether it's on-pump or off-pump. We don't make a choice. But what we did show and publish in JTCVS—I think it was in 2005—we look at this very question, and we introduced what we called a flow pressure ratio. So, the most important determination of [inaudible] graft is watch your blood pressure, and what we showed was when a patient had been done on-pump, they came off bypass with a relatively lower blood pressure than the off-pump patients but with far higher flows in the arterial and the venous grafts, and we believe this was an ischemia-reperfusion thing. By having made the territory ischemic, when you reperfused it, you got a higher natural flow. So, that is still my interpretation of what these data show. Dr Tolis. The numbers are great for both. I mean, if I had these PIs finishing a case, it really wouldn't matter, so I'm not entirely sure I see the clinical significance of your findings. Dr Leviner. So, just to answer your question regarding—so, I think if I remember correctly, this was done in 7 centers, and most centers were either heavily toward off-pump or heavily toward on-pump, either/or. There was not that much variation in between the centers themselves—in the centers themselves. I have to admit—this is a confession—I am not sure that propensity score matching this way is the correct way to correct for the differences, but reviewer 2 thought that we should do propensity score matching, so we went ahead and did propensity score matching. Was there anything else? I missed something. Unidentified Speaker 2. [inaudible] from Texas. I have a simple question. Did you see, regardless on-pump or off-pump, see on the degree of coronary stenosis? Because I think in the general practice, that this depends what you measure with the TTFM because if you have low degrees, especially in the left main stenosis, 50% to 60%, you have a very different measurement. Did you see on the degree of the stenosis? That's a very important thing. Dr Leviner. I did not have those data. I'm not sure if the request registry had it? Dr Taggart. We didn't, no. Dr Leviner. We didn't. And just there was I think another comment—and this, again, came up in the reviews many times—what's the clinical significance? So, I think we have to think of this in the context of the other data. Some of our grafts, there was more than 10 mL/min. Not all of the grafts, not on average, but there was more than 10 mL/min difference. Some other studies like the one I showed from JTCVS, there was, for some grafts, almost 20 mL/min difference. And when the cutoff value is set to 15 or 20 and there's a 10 mL difference on average, you have to think if this is relevant. I know you can, this is just statistics, but I think it has some relevance. Unidentified Speaker 2. Thank you. Unidentified Speaker 3. This is [inaudible] from Tokyo. In Japan, we perform average percent of the [inaudible] is nearly 60% and the [inaudible] from measurement is routine procedure in Japan because the government insurance, complete reimbursement for the [inaudible] measurement, and so we have lots of experience of [inaudible] measurement, especially in [inaudible] cases. And my question is, the flow after [inaudible] procedure, it changes because right after the anastomosis, I always observe a better flow in case of the [inaudible] because [inaudible], as Dr [inaudible] said, hyperemia eruption. But so, I measure 3 times during the operation, immediately after the anastomosis, and then finally, before I close the chest, then the flow always decreases. So, I think that this kind of a study, the important point is what time, in which occasion you measure the [inaudible] flow for the patient? So, the timing is a very important factor to compare the flow in this kind of study, so that's my question. Dr Leviner. I agree. For this study, we used only postprotamine measurements. I'm not sure. We didn't have in the database exactly when during the procedure this was done, but assuming postprotamine when you're very close to closing the chest. And by the way, we had another publication, another subanalysis from the REQUEST trial that looked at that question; is there a difference between pre- and postprotamine measurements, and we found no such difference. But I agree, and blood pressure variations affect this hugely, so it really does matter when you measure. Dr Taggart. Final comment. Thank you. I don't think these flows are relevant between on-pump and off-pump because they're so variable. If you've got a flow of 0, you've got a problem whether you are on-pump or off-pump, but the amount of difference flow can make by just changing the blood pressure by 20 mm Hg, you can double the phonograph. So, if you take in all of these variables—I mean, the importance of this technique is really making sure you don't have a graft with no flow or variable flow because then, you have a problem. So, on that point, thank you. We're now going to go on.