Published:January 03, 2008DOI:
        Dr Brian F. Buxton(Melbourne, Australia). This paper has some fascinating aspects to it. One is the sequential rate of angiography and the second is the excellent results obtained by off-pump surgery. Let me just run through some of the aspects of this study. This is a single-center, in fact almost a single-surgeon, retrospective observational study of 240 out of 402 patients who had off-pump surgery, 60% of whom had repeat angiography unrelated to symptoms. The sequential angiograms were obtained the day after surgery and at 1 year and 5 years. I will confine most of my comments to the main groups rather than the subset analyses presented toward the end.
        First, the 5-year angiographic patency studies of the ITA grafts were absolutely excellent, with 92% and 93% 5-year patency for the OPCAB series. In particular, I noted that most or many of the right ITA grafts were in situ. In contrast, the RGEA had 83% and saphenous vein 74% 5-year patency results. There were insufficient radial arteries in this small series to comment on the patency rates. However, these results overall do confirm the excellent results in the hands of the Korean surgeons in terms of their OPCAB series.
        The second point is that the overall patency in relation to target arteries provides no real new information. We know that the right ITA patency is superior to the LCX and RCA, and that was confirmed by these studies.
        I think of special interest in this paper are the data provided by the progression of disease in the FitzGibbon 50% stenosis or FitzGibbon B group. This showed several interesting facts. In some of the patients the lesions recovered and in other patients new type B lesions appeared.
        Let me run through the progression of disease. The ITA grafts showed very little progression of disease. There were two time frames, zero time to 1 year and 1 year through 5 years, and these failure rates were 4% and 6%, respectively. If we then look at the RGEA grafts, similar figures were 10% for the first 1-year increment and then 6% for the 4 years after that. In contrast, the saphenous vein grafts had double the failure rate of the arterial grafts, even in that relatively short follow-up period of 5 years.
        We noted the same observations in our own series published by Shah and associates, that is, that there is very little progression of disease in arterial grafts but quite significant progress in arterial grafts in a slightly longer follow-up period.
        It does appear that saphenous vein graft patency may have improved more recently with careful preparation, use of vasodilators, adequate storage techniques, and secondary prevention of the patients after surgery. It may be that this improvement in graft patency, if it is real, will require longer follow-up of these studies to obtain sufficient late data.
        Let me just run over some of the deficiencies of this excellent paper. First, it is a single-surgeon, single-institution study, and one raises the question about generalizability of these results. Are we always as good as the Koreans?
        Second, it is a retrospective study and therefore data quality comes into question.
        I think the most disappointing thing from my perspective is that there was no comparison between off-pump and on-pump surgery. There were no control data, which would be absolutely fascinating. I suspect this could be done in the future prospectively in a way that would give us some really meaningful answers to the question, is off-pump surgery patency as good as on-pump? I think that is the real question that most of us have in mind.
        Fourth, there is not really good documentation of the native vessel disease. In other words, how many vessels had serious competitive flow at low-grade stenoses?
        I have four questions: What technique did you use to implant the right ITA? Did you go across the anterior midline or through the transverse sinus as a Y graft from the left?
        Dr Kim. In this series, the right ITA was used to revascularize the LAD territory by crossing the midline, the ramus, or high obtuse marginal branch through the transverse sinus, and sometimes the RCA or posterior descending artery as an in situ graft. If the right ITA was too short to reach the left coronary territory or if the left coronary territory could not be completely revascularized with bilateral in situ ITA grafts, a Y graft was constructed.
        Dr Buxton. Second question: Most surgeons prefer or perhaps are more familiar with the Kaplan–Meier estimation of graft patency, and this can be done at zero time, 1, and 5 years. Why did you choose to use elective time estimates of 1 and 5 years?
        Dr Kim. We studied 240 patients who received all the early, 1-year, and 5-year follow-up angiograms. For the analysis of the serial changes over the 5-year period, we used nonparametric χ2 test with McNemar examination. We did not use the Kaplan–Meier estimation because we could not recognize the exact time of graft occlusions.
        Dr Buxton. Perhaps there are not so many assumptions made in your results as ours that might derive from a Kaplan–Meier analysis.
        One of the difficult things about looking at graft angiograms, and particularly anastomotic lesions and stenoses, is the validity and reproducibility of the observations. Can you tell us how many people looked and how independent the observers were in assessing the level of graft stenoses?
        Dr Kim. Our cardiologists performed the angiograms and put their initial interpretations. For this study, one physician initially reviewed all the angiograms and consensus was reached after review.
        Dr Buxton. One last question: Forty percent of the patients had missing data. Is there any reason for that? Did their grafts all fail? What happened? Why were they missing?
        Dr Kim. As I mentioned before, patients who missed any of the three postoperative serial angiograms were excluded from this study. In this study, 240 patients who received all three follow-up angiograms—early, 1 year, and 5 years—evaluation of the anastomotic sites and patency of the grafts, were studied.
        Dr Volkmar Falk(Leipzig, Germany). I enjoyed your paper, and I have one question. Would it be of value to go back and assess the extent of coronary artery disease by applying the SYNTAX score, which was designed to measure the extent of coronary artery disease? You are in a unique position in that you have angiograms at the time of surgery and 5 years after. You could finally tell us whether the extent of coronary artery disease at the time of surgery is really a predictor of late graft failure. That is one question that is still kind of open and no one is really addressing this. Would you consider doing that? I think this would be a great adjunct to your current study.
        Dr Damiano. Can you clarify what you mean by SYNTAX score?
        Dr Falk. I'm sorry; I thought that was common knowledge already. The SYNTAX trial was designed to compare treatment of triple vessel disease either by stenting or CABG. It was just finished, so the enrollment is just finished. Specifically for this trial, a new scoring system was developed to determine the extent of coronary artery disease, because most of the trials do not really look at this. Let's say they enroll patients with triple vessel disease, but we all know that there are various types of triple vessel disease that may alter the outcome more so than any of the other risk factors that are commonly applied, such as diabetes or hypertension. The extent of coronary artery disease can be graded. There is a good score for this that was actually designed for the SYNTAX trial but is available. It should be great to look at your angiograms again, apply the score, and then after a period of time assess whether the extent of coronary artery disease plays a role for late graft occlusion.
        Dr Damiano. It is a good idea, but I think he will probably have to go back and get the score and then report the data at a later meeting.
        Dr Kim. I would like to make some comments. As Dr Buxton indicated, there are limitations to the present study. It was a retrospective study of a single surgeon in a single institution. It was not performed in a randomized manner with regard to the type of conduit and the target vessels. The majority of ITA grafts were used to revascularize the left coronary territory and the majority of RGEA grafts were used to revascularize the RCA territory. This might serve as confounding variables. We have two staff surgeons in the adult cardiac division of our institution. I do most of the isolated CABGs with off-pump technique and the other surgeon does most of the valvular and aortic operations. That is why this study was a single-surgeon experience from a single institution. As we performed OPCAB in most of patients during the study period, it was difficult to obtain an on-pump control group to see the difference between off-pump and on-pump CABG.