Editorial commentTotal arch repair versus hemiarch repair in the management of acute DeBakey type I aortic dissection

Because of the improvements in the management of acute type A aortic dissection obtained during the last two decades, a controversial debate came out some years ago: Should the replacement of the aorta remain limited as it was traditionally or should we resort to extended aortic replacements? This issue was perfectly illustrated by two major articles by Westaby et al. on the one hand [1] and Kazui et al. on the other hand [2] in the early 2000s. Using a somewhat conservative and stereotyped method, the former obtained a hospital mortality rate of 6% and a rate of reoperation of 5% at 4 years. Conversely, the latter proposed to perform a complete replacement of the aortic arch in most patients. He obtained a hospital mortality rate of 4% and a freedom from reoperation rate of 75% at 10 years. Therefore, both strategies seemed to be efficient and safe and, thus, legitimate. In their article published in the present issue of the journal, Kim et al. address again this yet unresolved issue. They conclude that ‘total arch repair was associated with greater risks of mortality and permanent neurological injury compared with hem-arch repair’ and that ‘rates of reoperation or distal aortic dilatation were not significantly different between the two strategies.’ Therefore, ‘these findings support a conservative surgical approach to circumvent this life-threatening situation’ [3]. The results of the reported experience and the conclusion drawn by the authors seem to go somewhat against the current trend. Indeed, more and more published studies favor extended replacements [4–6]. So, are Kim et al. advocating a somewhat obsolete or ‘old-fashioned’ policy? Some points, indeed, in this study might be questioned. The nature and the extension of the aortic replacement do not depend only on the surgeon’s preference. When the intimal tear is located on the ascending aorta or in the concavity of the arch, it is legitimate to perform a limited replacement. But when it extends to or is completely located in the transverse arch or when this segment is severely damaged by the dissecting process, or when the dissection is retrograde, it is mandatory to replace this segment of aorta. Furthermore, as pointed out by the authors, those patients requiring complete replacement of the arch are more prone to have severe anatomical and physiological damages. It is therefore little surprising that the results in this cohort of patients are less favorable than in patients requiring a limited replacement. In the authors’ experience, the rate of late reoperations was very low and there was no difference between both groups. This is not surprising for several reasons: The mean follow-up was exactly 4 years. This might be too short a period to observe the need for distal aortic replacement. Also, if we look at longer intervals we may observe that the number of patients at risk was extremely low, making the statistical significance nil or irrelevant. There were only eight (4%) Marfan patients. It is well known that this group pays the highest toll to late reoperations and that the risk of such procedures in those patients is up to sixfold higher than for non-Marfan patients. A larger cohort might have made some difference [7]. Nevertheless, it is ascertained that the type and quality of the initial surgical procedure have a great influence on the immediate and long-term results. In this regard, there might be other reasons than the extension of the aortic arch replacement for explaining the good results obtained, in particular in the limited replacement group: the right axillary artery was cannulated in 70% of the patients and all had the aortic repair performed with an ‘open distal anastomosis’ during a period of circulatory arrest. Until recently, this was seldom carried out during limited replacements and most patients had the ascending aortic replacement performed with femoral cannulation and aortic cross-clamping. This has led in many instances to malperfusion syndromes, false channel ruptures, and/or severe residual distal aneurismal dilatation and late reoperations or death. It is therefore logical that, in the present report, the up-to-date techniques have benefited mainly to this group of patients. Finally, we know that the technique of total arch replacement requires a certain expertise and experience. But even though we are informed that the procedures were distributed among six surgeons, we do not know whether they all had the same experience or whether the most experienced ones have performed the total arch replacements, considering that surgery of acute type A dissection is generally performed in emergency and not always by the most experienced surgeons. However, given the daily circumstances in which most acute type A dissections are operated on, the present report may shed some reasonable light on the issue. There is no doubt that the initial emergency surgery should take into account all the damages induced by the dissecting process on which the immediate and long-term results will depend. This means that the modern techniques such as right axillary artery cannulation (or any mode of cannulation allowing permanent antegrade aortic perfusion), open distal anastomosis, Selective Antegrade Cerebral Perfusion, and hypothermia (preferably moderate) must be used in all cases as they are critical for the immediate and, possibly, late results. However, as John Elefteriades stated one decade ago ‘…acute type A dissection is an inherently lethal condition. Our first job is to produce a live patient’ [8]. This may imply that there should not be one single surgical strategy and technique carved in marble, but that it should be adapted to the circumstances, to the patient’s clinical condition, to the pathological features and the anatomical damages. About one decade ago, we asked ‘Acute type A dissection: Can we dramatically reduce the surgical mortality?’ [9]. It seems that the answer is ‘yes’. Would this happen through limited or extended aortic replacements? Apparently the debate is still open.