Results of surgery for aortic root aneurysm in patients with the Marfan syndrome

Modern cardiac surgery dramatically changed the natural history of patients with the Marfan syndrome [1–3]. In 1968, Bentall and DeBono described an operative procedure whereby the aortic root and ascending aorta were replaced with a tubular Teflon graft containing a mechanical heart valve, and the coronary arteries were reimplanted by suturing the graft to the aortic sinus wall around their orifices [2]. Although certain aspects of that procedure were modified to minimize operative and late complications, composite replacement of the aortic valve and ascending aorta with reimplantation of the coronary arteries remains the standard operation for patients with aortic root aneurysm and the Marfan syndrome [3]. In 1992, David and Feindel described a method of reconstruction of the aortic valve in patients with aortic root aneurysm, whereby the three sinuses of Valsalva and the ascending aorta were excised, and the aortic valve was re-implanted inside a tubular Dacron graft [4]. This and other types of aortic valve-sparing operations are alternative procedures to the composite replacement of the aortic valve and ascending aorta in patients with aortic root aneurysm and macroscopically normal aortic valve cusps [5]. Many surgeons believe that aortic valve-sparing operations should not be used in patients with the Marfan syndrome [3]. We believe that both the aortic and the mitral valve of patients with the Marfan syndrome should be preserved whenever possible, and have adopted this approach in our practices over the past decade. This study is a review of our experience with surgery of the aortic root in patients with the Marfan syndrome. From 1979 to 1998, 78 consecutive patients with the Marfan syndrome according to Gent criteria [6] had surgery for aortic root aneurysm at Toronto General Hospital. There were 61 men and 17 women. The mean age of the patients was 35 years (range, 16–70). The operation was performed electively in 63 patients and as an emergency in 15 (acute type A aortic dissection in 14 patients and cardiogenic shock in one). Aortic valve-sparing operations were performed in 42 patients and composite replacement of the aortic valve and ascending aorta in 36. Aortic valve-sparing operations have been performed only since May 1988 and in patients with normal or mildly diseased aortic valve cusps. Composite replacement of the aortic valve and ascending aorta was performed throughout the period of study, and since 1988, only in patients with grossly abnormal aortic cusps. Table 1 shows the preoperative characteristics of patients who had aortic valve-sparing operations and those who had aortic root replacement. Table 2 shows the operative data. Patients have been followed up at annual intervals and Doppler echocardiographic studies were obtained for those who had aortic valve-sparing operations. An annual CT scan of the chest and abdomen was also obtained in patients who had aortic dissection. The mean follow-up was 41±35 months (range, 5–133) for patients who had aortic valve-sparing operations, and 65±50 months (range, 3–216) for patients who had composite replacement of the aortic valve and ascending aorta. No patient was lost to follow-up. Preoperative characteristics of the patients who had valve repair and replacementa Percentages are shown in parentheses. Operative data Clinical and echocardiographic variables were entered in datasets using dBASEIV, and analyzed with the SAS for PC statistical program. Categorical variables were compared using Chi-square or Fisher's exact test where appropriate. Time-related events were evaluated univariately by the Kaplan–Meier method and log-rank statistic. Multivariable analysis was not performed because of the small sample size and too few events. There was one operative death in a patient who had an emergency operation because of cardiogenic shock due to end-stage aortic insufficiency, severe left ventricular dysfunction, coarctation of the aorta and a massive aneurysm of the aortic root. A valved conduit was used for aortic root replacement. This patient died of myocardial failure. Nine patients required re-exploration of the mediastinum for bleeding; five patients in the valve-sparing group and four in the root replacement group. There was no perioperative stroke, myocardial infarction, or any infection. There were four late deaths, all in the aortic root replacement group. The cause of late death was valve-related in two patients (one due to endocarditis and one sudden death, probably an acute myocardial infarction), and due to vascular complications in two patients (one rupture of the false lumen and one rupture of a true thoracic aneurysm). Fig. 1 shows the survival in both groups of patients. The survival at 5 years was 100% for patients who had aortic valve-sparing operations and 88±6% for patients who had aortic root replacement (P=0.04). Survival of patients who had aortic valve-sparing operations and aortic root replacement. There were no thromboembolic events in patients who had valve-sparing and there were two strokes in patients who had aortic root replacement with a mechanical aortic valve. Neither one of these patients had undergone mitral valve replacement. Both patients recovered completely. There were four episodes of prosthetic valve endocarditis in three patients. One patient was deemed inoperable and died. He was an elderly man who had undergone two previous aortic root and mitral valve replacements and multiple co-morbid conditions. The other two patients underwent aortic root re-replacement and survived. One of them had endocarditis twice, 4 years apart. There were no major bleeding complications in patients on oral anticoagulants for mechanical aortic valves. There were seven aortic root re-replacements in five patients. Three reoperations were because of prosthetic valve endocarditis and four due to biological valve failure (two bioprostheses and two homografts). No patient died at reoperation. There was no reoperation in the aortic valve-sparing group. Fig. 2 shows the freedom from reoperation; at 5 years it was 100% for patients who had aortic valve-sparing and 89±6% for patients who had aortic root replacement (P=0.04). Freedom from reoperation in patients who had aortic valve-sparing operations and aortic root replacement. Four patients, two from each group, required replacement of the entire thoracic and abdominal aorta because of expansion or rupture of the false lumen. One patient with a ruptured false lumen died. He had undergone aortic root replacement with a mechanical valve. Another patient became paraplegic. He had undergone an aortic valve-sparing operation. This last patient also needed replacement of the transverse aortic arch. The other two patients, one from each group, did well. The most recent Doppler echocardiographic study in patients who had aortic valve-sparing operations showed no aortic insufficiency in six patients, trivial in 13, mild in 20, and moderate in three. These three patients with moderate aortic insufficiency had mildly elongated aortic cusps and required shortening of the their free margins with a 6-0 ePTFE suture (expanded polytetrafluoroethylene), in addition to an aortic annuloplasty and remodeling of the aortic root. The standard treatment for aortic root aneurysm in patients with the Marfan syndrome is composite replacement of the aortic valve and ascending aorta [3]. This operation improved the poor prognosis of patients with aortic root aneurysm and the Marfan syndrome [1]. In a recent report by Gott et al. on the results of aortic root surgery in patients with the Marfan syndrome from multiple institutions, the operative mortality was 1.5% for elective operations and 11.7% for emergency operations [3]. The actuarial survival rates in that collective series were 84% at 5 years, and 75% at 10 years. Coincidentally, the long-term survival in our much smaller series of patients who had aortic root replacement is remarkably similar to that reported by Gott et al. [3]. Actually, our patients were included in that series. Since 1988, we have been preserving the aortic valve in patients with aortic root aneurysms as long as the aortic cusps are normal or only minimally diseased. We have not excluded patients with the Marfan syndrome. The appropriateness of this operation in this group of patients has been controversial because many surgeons believe that the aortic cusps of patients with the Marfan syndrome are histologically abnormal. Indeed, histological studies by Fleischer et al. [7] revealed a high degree of structural deterioration in aortic cusps removed at the time of aortic root replacement in patients with the Marfan syndrome. Although we do not dispute those findings, the experience with mitral valve repair in patients with myxomatous disease of the mitral valve, including those with the Marfan syndrome, indicates that histochemically-abnormal leaflets and chordae tendineae can last a long time if properly anchored by the surrounding structures [8,9]. By analogy, the aortic valve cusps should not fail if the aortic annulus and the sinotubular junction are stabilized by corrective surgery [5]. It has been our experience that the aortic valve cusps of patients with the Marfan syndrome are remarkably normal when the aortic root aneurysm is not excessively large. However, as the aortic root dilates, the mechanical stress on the cusp increases and the degenerative process is accelerated. The cusps become thinner and elongated, and stress fenestrations appear in the commissural areas. These cusps should not be repaired and aortic root replacement is certainly the appropriate treatment. Most of these patients have aortic root diameters in excess of 55 mm. However, if surgery is performed before the aortic root reaches that size, the aortic cusps are frequently normal or only minimally stretched, and a valve-sparing operation is feasible and probably the best treatment for these patients. In our study, the group of patients who had aortic valve-sparing operations was different from the group who had aortic root replacement, as indicated in Table 1. Patients who had aortic valve-sparing operations were less symptomatic, and had better left ventricular function, lower grade aortic insufficiency, fewer co-morbid conditions and smaller aortic root aneurysms than patients who had aortic root replacement. Although these differences may have played a role in the clinical outcomes, it is important to notice that patients who had aortic valve-sparing operations had no cardiac or valve-related events postoperatively. This is in contrast with patients who had aortic root replacement with biological or mechanical valves. Anticoagulation-related hemorrhage, thromboembolism and endocarditis remain as serious problems with mechanical heart valves, whereas primary tissue failure and endocarditis remain as the main problems with biological valves. None of these problems appear to be an issue in patients who have aortic valve-sparing operations. As with any type of heart valve repair, the most important adverse effect with aortic valve-sparing operations is valve failure. There is certainly a learning curve with any procedure, and aortic valve-sparing operation is no exception. Initially, we used the reimplantation technique in patients with annuloaortic ectasia, and the remodeling technique in those with normal aortic annulus [5]. The advantage of the reimplantation technique is that it addresses all problems, the dilated annulus, sinuses of Valsalva and sinotubular junction. Its disadvantage is that the aortic valve cusps are placed inside a cylindrical conduit without the sinuses of Valsalva, which may alter their motion and increase mechanical stress. Cochran et al. proposed a modification to the reimplantation technique to create pseudo-sinuses [10]. Yacoub et al. have exclusively used the remodeling procedure, even in patients with the Marfan syndrome [11]. We believe that these patients often have a dilated aortic annulus, and if the remodeling procedure is used, an aortic annuloplasty should be added [5]. After a decade of experience with various types of aortic valve-sparing operations in over 150 patients with ascending aorta and/or aortic root aneurysms, we have learned that the reimplantation technique provides for a more stable valve function than the remodeling procedure in patients with annuloaortic ectasia. Although the number of patients who had reimplantation of the aortic valve is much smaller than those who had remodeling of the aortic root, no patient in the first group has more than mild aortic insufficiency after 5–10 years of follow-up. Because of the excellent functional results obtained with aortic valve-sparing operations in the first 5 years of our experience, we began to expand the indications of these procedures, and included patients with prolapsed aortic cusps due to elongation of their free margin [12]. If one, or even two cusps were slightly elongated, we shortened their length with a double layer 6-0 ePTFE suture [12]. Six of our patients with the Marfan syndrome had aortic cusp-shortening with ePTFE sutures, and after 1–5 years, three patients have none or mild aortic insufficiency and three have moderate aortic insufficiency. If this type of repair proves durable, many more patients with aortic root aneurysms could have aortic valve-sparing operations instead of aortic root replacement. This is a retrospective study with a small sample size. In addition, patients who had aortic root replacement had a higher prevalence of risk factors and longer follow-up than those who had aortic valve-sparing operations. These variables may have biased our results. It is conceivable that with a larger sample size, ample event rate and a randomized trial, multivariable adjustment might dilute the benefit seen with aortic valve-sparing operations. This remains to be seen and should provide the hypothesis for a future multi-center collaborative clinical trial. Aortic valve-sparing operations in patients with the Marfan syndrome are safe and provide excellent clinical results during the first decade of follow-up. Because the feasibility of aortic valve-sparing operations is highly dependent on the quality of the aortic cusps, surgery in patients with the Marfan syndrome should be performed before the aneurysm reaches 55 mm and severe aortic insufficiency develops. We believe that surgery should be considered when the diameter of the aortic root reaches 50 mm and the aortic valve cusps are echocardiographically normal. Dr R. Dion (Brussels, Belgium): Thanks to the leading experience of Dr David, we started in 1995 our program of aortic root remodeling, and we did that in 62 patients, 14 of whom had Marfan syndrome, with excellent results. My question is a technical one: intraoperatively the aortic annulus diameter plays a key role. if that diameter exceeds 28 mm, and especially in the Marfan syndrome, we now add an external annuloplasty to prevent further dilation of the annulus. What is your attitude in this situation? Dr Tambeur: As far as Marfan patients are concerned, they all have an annuloplasty, meaning they either get a David 1 or a David 2, and in addition, they get an annuloplasty or annular stabilization in that procedure. Dr Dion: Even if the annulus is, let us say, 25 mm? Dr Tambeur: Even if it is smaller. Dr T. Orszulak (Rochester, MN, USA): I also share your enthusiasm for the use of the valve-sparing procedure for Marfan's. I have three questions. You outline some discrepancy, in that, with valve-sparing you had roughly 17% of patients who had grade 3, and 17 who had grade 4 aortic regurgitation. How did you decide to go ahead and repair those? Tirone David has mentioned that patients who have severe AI may not be able to have valve repair? What was different about your group of patients? Secondly, did any of the patients that had repair immediately convert at the same operation to valve replacement? Thirdly, did you find any difference between the Tirone 1 and Tirone 2 in your results? Dr Tambeur: The decision is clearly an intraoperative decision. If the leaflets, as they are seen clinically, are good enough to allow a repair procedure, they get repaired. There was no conversion at any time during this study in all these patients. As far as the third question is concerned, we don't know which procedure will be the better one. The jury is still out on that one. Dr W. Harringer (Hannover, Germany): I would just like to add a little bit on our information of 102 classical David type 1 procedures, 24 of which were Marfan syndromes. What we have experienced is that three patients required reoperation, two basically to severe AI and one because of extensive expansion of the aneurysm. What we have seen in the failure rate is that both patients in the Marfan group had cusp prolapse which they experienced over the years. Both patients did not have a perfect coaptation early on, immediately after that repair, which leads me immediately to my question. Now, have you experienced any progressive deterioration over time in your aortic insufficiency over the years? The second question is: could you give me the rationale with regard to flexibility, whatsoever, between David type 1 and David type 2, because in my understanding, the annuloplasty and stabilizing the annulus take away quite a bit of flexibility of the left ventricular outflow tract, so that is something I haven't understood yet? Dr Tambeur: As far as the flexibility goes, we think that the stabilization of the annulus is the more important part of the procedure regarding the long-term outcome of what we have tried to achieve here. We have not seen any significant deterioration in aortic valve function, which means we have not seen any aortic regurgitation increasing by more than one grade over the follow-up of these patients, which reaches up to 11 years now. Dr M. Sarsam (Belfast, UK): Well, as you know, we were also interested in this operation, and I think really David 2 should be called the Yacoub operation because he described it much earlier than David. The question I have for you is, it is confusing to say the aortic annulus diameter is 50 or 60 mm. I think you must be talking about either the diameter of the sinotubular junction or the diameter of the aneurysm, because I have never seen the annulus itself, the measurement you use for sizing a valve, as big as 50 or 60; in fact the biggest one we had was 34 mm. The reason we always do the remodeling procedure is to preserve the sinus function, I think, to answer the question about the difference between the Tirone David 1 and what you call Tirone David 2. We never had to use an annuloplasty ring or anything like that, and we haven't had any deterioration in valve function in the Marfan group, and now we have up to six years of follow-up, at least in Belfast. I think one important point in this operation is that there is a secondary dilatation in the leaflet, and there are some cases where the aorta is enlarged but the valve is minimally incompetent, and those are actually the most difficult to remodel, because if you reduce the size of the aorta you make the leaflet prolapse, and in those cases it is better to choose a slightly bigger size of graft and also suspend the commissural point at a higher level. Dr Tambeur: To answer, the annulus and root diameters were all measured echocardiographically and they are actual diameters at the respective level. We think that annuloplasty is an essential part of this surgical procedure, certainly as far as Marfan patients are concerned, and the diameter of the graft should be commensurate with the length of the free margin of the leaflets, which means that you take for a reimplantation procedure a graft which has about the same diameter as the length of the free margin of the leaflets. For the remodeling procedure, you take a graft with a diameter which is about 10% less than the length of the free margin of the leaflets.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.