- Email: [email protected]
Options for repair of a bicuspid aortic valve and ascending aortic aneurysm
Options for Repair of a Bicuspid Aortic Valve and Ascending Aortic Aneurysm Thoralf M. Sundt III, MD, Bassem N. Mora, MD, Marc R. Moon, MD, Marci S. Bailey, MSN, Michael K. Pasque, MD, and William A. Gay, Jr, MD Division of Cardiothoracic Surgery, Department of Surgery, Washington University School of Medicine, St. Louis, Missouri
Background. Intrinsic abnormality of the aortic wall may explain the association of bicuspid aortic valves with ascending aortic aneurysms. Separate valve and graft repair of such lesions, rather than composite valve graft replacement, is more straightforward but leaves potentially abnormal sinuses behind. Methods. Between January 1985 and January 1998, 45 patients underwent separate valve and graft (n ⴝ 27) or composite valve graft (n ⴝ 18) for an ascending aortic aneurysm and bicuspid aortic valve. Perioperative events and late results were compared. Results. Patients undergoing separate valve and graft were older (mean age, 60 ⴞ 13 vs 42 ⴞ 12 years, p < 0.001) and were more likely to have purely stenotic (48% vs 6%, p ⴝ 0.003) than purely regurgitant (11% vs 72%, p < 0.001)
disease. They were also more likely to require concomitant coronary artery bypass grafting (56% vs 6%, p ⴝ 0.001). There were no significant differences in operative risk and no known late complications related to recurrent aneurysms. Conclusions. Root replacement with a composite valve graft can be accomplished with low operative risk and is the first choice for repair of this lesion. Separate valve and graft repair, however, yields satisfactory early and late results and remains an acceptable option, especially when the coronary ostea are not displaced or when concomitant procedures must be performed.
A
of hemodynamically significant aortic stenosis. Studying patients with functionally normal bicuspid valves and those with varying degrees of hemodynamic abnormalities, Hahn and associates [6] demonstrated that root enlargement was widespread. Enlargement was greatest among those with regurgitant lesions. Furthermore, recent histochemical studies have revealed an increased rate of smooth muscle cell apoptosis in the aortic media of patients with bicuspid aortic valves, even in the absence of gross dilation [7]. The presence or absence of intrinsic abnormality of the aortic wall has profound implications for the proper surgical management of this condition. A strategy of valve replacement and separate supracoronary graft repair of the ascending aorta, leaving the sinuses intact, is straightforward and obviates risks related to coronary osteal reimplantation, but may also subject the patient to the risk of reoperation for aneurysmal dilation of the unreplaced root elements. Composite root replacement is a well-established technique, with the operative mortality rate reported in several series to be as low as 2% for elective procedures [8, 9]. It is now widely accepted as the standard of care for patients with Marfan’s syndrome, in whom the risk of aneurysmal dilation of the sinuses has been shown to be considerable [10]. Without displacement of the coronary ostea, however, composite root replacement can be difficult, and it may be a more time-consuming procedure than separate valve replacement and supracoronary graft replacement of the ascending aorta. There is, as yet, no consensus regarding
n association between bicuspid aortic valve and diseases of the aorta has long been recognized, although the causal relationship remains undefined. Abbott and Hamilton [1] reported this observation, as well as an apparent association with propensity for rupture of the ascending aorta, in 1928 based on study of coarctation in the adult. They hypothesized a developmental relationship between bicuspid aortic valve disease and intrinsic weakness of the aorta. Subsequently, McKusick and colleagues [2] reported an association between bicuspid aortic valve and dissection of the ascending aorta with cystic medial necrosis. Originally favoring a hemodynamic explanation, further clinical observations led them to shift their position, favoring a common developmental defect with variable expression among individuals leading to a bicuspid aortic valve, coarctation of the aorta, and cystic medial necrosis [3]. These seminal observations have been reconfirmed by others [4], and the hypothesis of intrinsic abnormality has increasingly been embraced. Several recent echocardiographic studies have supported this notion. Pachulkski and colleagues [5] found that the aortic root diameter was significantly larger among patients with bicuspid aortic valves than among normal controls, even in the absence
Presented at the Forty-sixth Annual Meeting of the Southern Thoracic Surgical Association, San Juan, Puerto Rico, Nov 4 – 6, 1999. Address reprint requests to Dr Sundt, Division of Cardiothoracic Surgery, Washington University School of Medicine, Suite 3106 Queeny Tower, One Barnes Hospital Plaza, St. Louis, MO 63110; e-mail: [email protected].
© 2000 by The Society of Thoracic Surgeons Published by Elsevier Science Inc
(Ann Thorac Surg 2000;69:1333–7) © 2000 by The Society of Thoracic Surgeons
0003-4975/00/$20.00 PII S0003-4975(00)01220-0
1334
SUNDT ET AL BICUSPID AORTIC VALVE AND ASCENDING ANEURYSM
the management of ascending aneurysms associated with bicuspid valve disease. We therefore reviewed our institutional experience with this condition over the preceding 13 years to determine comparative operative risk and long-term outcome of separate valve and graft repair versus composite root replacement for ascending aortic aneurysms in association with bicuspid aortic valves.
Material and Methods Retrospective review of the computerized database for the Division of Cardiothoracic Surgery at Washington University identified 126 patients undergoing either composite root replacement or separate valve and graft repair of aortic valve pathology and an ascending aortic aneurysm between January 1985 and January 1998. In 45 patients, the aortic valve was documented to be bicuspid in either the operative note or the pathology report. The decision to address dilation of the ascending aorta was made by the operating surgeon. There is some variation among surgeons; however, it is our general practice to intervene routinely on an ascending aorta greater than 6 centimeters in diameter as determined by preoperative computed tomography, echocardiography, or intraoperative inspection. In the presence of a bicuspid valve, we consider graft replacement when the ascending aorta is more than 4 centimeters in diameter. Aortic “tailoring” for dilation of the ascending aorta is only rarely practiced. Patients undergoing this procedure were therefore excluded from this study. Preoperative demographic characteristics, operative variables, and perioperative events were retrieved from the database. Late follow-up information was obtained by postal questionnaire or telephone interview between July 1, 1998, and December 31, 1998. Follow-up was 100% complete. The mean follow-up was 5.6 (⫾ 4.1) years, and the total number of patient-years of follow-up was 243 years.
Surgical Technique Surgical procedures were carried out by the members of the Division of Cardiothoracic Surgery at Washington University both at Barnes Hospital and Jewish Hospital. Procedures were carried out under hypothermic cardiopulmonary bypass with cardioplegic arrest for myocardial protection. Initially antegrade cold crystalloid cardioplegia was employed. More recently cold blood cardioplegia, often given both antegrade and retrograde, has been used. Profound hypothermia and circulatory arrest were employed when concomitant hemiarch or full arch replacement was undertaken. The operating surgeon made the choice between procedures. Composite root replacement was performed most often using the open coronary button technique as previously reported by Kouchoukos and colleagues [8]. Separate graft replacement of the ascending aorta was performed with proximal anastomosis to the supracoronary ridge. The aorta was transected completely before distal anastomosis.
Ann Thorac Surg 2000;69:1333–7
Table 1. Demographic Characteristics of the Study Population Variable
CVG (n ⫽ 18)
Age (y) Mean ⫾ SD 42 ⫾ 12 Range 25–70 Male sex 15 (83%) Marfan’s syndrome 1 (6%) Diabetes mellitus 2 (11%) Hypertension 6 (33%) Tobacco abuse 6 (33%) Renal dysfunction 1 (6%) Hypercholesterolemia 3 (16%) Peripheral vascular 2 (11%) disease Cerobrovascular 1 (6%) disease Previous CAB 0 Functional aortic valve pathology Stenosis 1 (6%) Regurgitation 13 (72%) Mixed 4 (22%) Ascending aortic 6.1 ⫾ 1.8 cm diametera
Separate V ⫹ G (n ⫽ 27) 60 ⫾ 13 32–78 21 (77%) 0 4 (15%) 12 (44%) 8 (30%) 0 7 (26%) 0
p Value ⬍ 0.001 0.65 0.40 0.94 0.54 0.94 0.40 0.72 0.15
3 (11%)
0.64
3 (11%)
0.26
13 (48%) 3 (11%) 11 (41%) 5.4 ⫾ 1.2 cm
0.003 ⬍ 0.001 0.33 0.21
a
Determined by visual inspection at surgery, computed tomography, or echocardiography.
CAB ⫽ coronary artery bypass; V ⫹ G ⫽ valve plus graft.
CVG ⫽ composite valve and graft;
Statistical Analysis All continuous data were expressed as the mean plus or minus the standard deviation. Two-group comparisons for continuous data were made using the Student’s t test. Categorical data were tabulated, and two-group comparisons were made using the 2 test for 2 ⫻ n tables. When 2 ⫻ 2 tables were analyzed, Fisher’s exact test was used. The survival analysis was done using the Kaplan-Meier estimator. Comparisons were made using the Mantel log-rank test. The Cox proportional hazards model was used to examine the effect of age (continuous variable) on the survivor function. A p value of less than 0.05 was considered significant for all statistical calculations. The SYSTAT system for statistics was used for all data analysis (version 6.0 for Windows; SPSS, Chicago, IL).
Results Study Population The demographic characteristics of the study groups are shown in Table 1. Patients undergoing composite root replacement were significantly younger than those having a separate valve and graft, and were more likely to have predominantly regurgitant rather than stenotic lesions. The only patient in the study group who met clinical criteria for the Marfan’s syndrome underwent composite root replacement. Aortic diameters were similar in both groups.
Ann Thorac Surg 2000;69:1333–7
SUNDT ET AL BICUSPID AORTIC VALVE AND ASCENDING ANEURYSM
Table 2. Operative Characteristics Variable
a
Table 3. Perioperative Results
CVG Separate V ⫹ G (n ⫽ 18) (n ⫽ 27) p Value
CPB time (min) Mean ⫾ SD 169 ⫾ 45 Range (109 –272) Aortic occlusion time (min) Mean ⫾ SD 132 ⫾ 43 Range (77–241) Associated procedures CAB 1 (6%) Arch replacement 1 (6%) Aortic annular enlargement 0 VSD closure 0 Repair aortic coarctation 1 (6%) Circulatory arrest used 1 (6%) Valve size (mm) Mean ⫾ SD 26.3 Range 25–31 Mechanical valve 12 (92%)a Biological valve 1 (8%)a
1335
163 ⫾ 53 (92–294)
0.70
115 ⫾ 28 (60 –168)
0.15
15 (56%) 6 (22%) 1 (4%) 1 (4%) 0 6 (22%)
0.001 0.22 0.46 0.46 0.40 0.22
25.0 21–31 16 (59%) 11 (41%)
0.15
Valve prosthesis type known for 13 of 18 CVG patients.
CAB ⫽ coronary artery bypass; CPB ⫽ cardiopulmonary bypass; CVG ⫽ composite valve and graft; V ⫹ G ⫽ valve plus graft; VSD ⫽ ventricular septal defect.
Operative procedures tended to be more complex in the separate valve and graft group, as shown in Table 2. Concomitant coronary artery bypass was performed more often in the separate valve and graft group, as was aortic arch replacement, although the latter did not reach statistical significance. Despite the more frequent use of circulatory arrest in the separate valve and graft group, the mean cardiopulmonary bypass times for both groups were nearly identical. There was a trend toward somewhat longer aortic occlusion time among patients undergoing composite root replacement, reflecting the increased technical complexity of this procedure. Composite replacement was done almost exclusively with mechanical prostheses. Mechanical valves also predominated in the separate valve and graft group, but more narrowly so.
Variable
CVG (n ⫽ 18)
Mortality IABP support Re-exploration for bleeding Prolonged ventillation Stroke Renal dysfunction Intraoperative blood used Postoperative blood used Total units pRBCs Total units FFP Total units platelets
0 0 0 1 (5.5%) 2 (11%) 2 (11%) 10 (56%) 9 (50%) 2.1 ⫾ 2.5 0.8 ⫾ 1.6 1.0 ⫾ 1.8
Separate V ⫹ G (n ⫽ 27) p Value 2 (7.4%) 2 (7.4%) 2 (7.4%) 4 (14.8%) 0 3 (11%) 18 (67%) 12 (44%) 3.44 ⫾ 4.1 2.5 ⫾ 5.0 1.7 ⫾ 3.2
0.51 0.51 0.51 0.23 0.16 1.0 0.76 0.49 0.16 0.10 0.32
CVG ⫽ composite valve and graft; FFP ⫽ fresh frozen plasma; pRBC ⫽ packed red blood cells; V ⫹ G ⫽ valve plus graft.
ing, and separate graft replacement of the ascending aorta. On postoperative day 1 he exhibited hemodynamic instability and underwent operative exploration. He was found to have right heart failure with thrombosis of a bypass graft to the posterior descending coronary artery. He expired despite graft thrombectomy and institution of ventricular assist device support. There were no statistically significant differences in rates of perioperative intraaortic balloon support, stroke, renal dysfunction, or respiratory insufficiency between groups. There were also no statistically significant differences in perioperative transfusion or in the rates of exploration for bleeding.
Late Results Actuarial survival was superior ( p ⫽ 0.04) among patients undergoing composite root replacement, as shown in Figure 1. However, multivariate stepwise logistic regression analysis of risk factors for late death, including the type of aortic procedure performed, diabetes mellitus, preoperative renal dysfunction, hypercholesterolemia, peripheral vascular disease, previous or concomitant coronary bypass or aortic arch replacement, and
Perioperative Complications There was no difference in perioperative mortality between groups (Table 3). Two patients undergoing separate valve and graft reconstruction died within 30 days of surgery. One was a 52-year-old man presenting with aortic stenosis, who had undergone mechanical aortic valve replacement and supracoronary graft replacement of the ascending aorta. Twenty-five days after hospital discharge he was readmitted with a markedly elevated prothrombin time (29 seconds) and near syncope. He experienced a cardiopulmonary arrest shortly after arrival and could not be resuscitated. Postmortem examination demonstrated mediastinal hematoma. The second perioperative mortality occurred in a 68-year-old man with a history of previous coronary artery bypass, who underwent aortic valve replacement, repeat bypass graft-
Fig 1. Kaplan-Meier actuarial survival estimates for composite valve and graft (CVG) and separate valve and graft (V ⫹ G) replacement of the aortic valve and ascending aorta for bicuspid aortic valve and associated ascending aortic aneurysm. Multivariate analysis identified only age, and not procedure, as a predictor of late death.
1336
SUNDT ET AL BICUSPID AORTIC VALVE AND ASCENDING ANEURYSM
functional pathology of the valve, yielded only age as a significant predictor of death. There were no known late reoperations for remaining or recurrent pathology of the aortic root. Similarly, there were no late deaths known to be related to recurrent aneurysms of the aortic root.
Comment The association between intrinsic aortic disease and bicuspid aortic valve disease is of practical importance. Bicuspid aortic valve disease occurs in 1% to 2% of the population. In autopsy studies it has been reported in more than one-half of patients with anatomically isolated aortic stenosis and in more than 25% of those with clinically pure aortic regurgitation [11]. Conversely, among 85 cases of bicuspid aortic valve identified at autopsy, Roberts [11] reported that 61 patients (72%) had stenotic or mixed lesions, 11 (13%) had purely regurgitant lesions, and 13 (15%) had functionally normal valves. The risk of aortic dissection after aortic valve replacement for a bicuspid aortic valve is well documented. The studies by Prenger and colleagues [12] and Pieters and colleagues [13] demonstrated a substantial risk (22% to 27%) of aortic dissection after AVR when the aorta is more than 5 centimeters in diameter. In the study by Prenger and colleagues, all patients with aortic dissection had a diameter of more than 5 cm. Accordingly, it is our practice to be particularly aggressive with the treatment of the dilated aorta during aortic valve replacement in the presence of a bicuspid valve. Debate over the choice between composite root replacement and separate valve and graft replacement of the ascending aorta is not new; however, little has been said on the subject in the context of bicuspid aortic valve disease. The latter technique has proved inadequate in the presence of Marfan’s syndrome [10], the archtypical connective tissue abnormality, and composite root replacement is now the standard of care in this circumstance. Composite root replacement can be accomplished in this condition with very low operative risk under elective circumstances [9], particularly in experienced centers [14]. Gott and colleagues [9] recently reported an operative mortality of 1.5% among 455 patients with Marfan’s syndrome undergoing elective root replacement at 10 experienced centers. The series of 138 such consecutive elective procedures without a single operative death reported from Johns Hopkins sets the standard for the field [14]. There is, however, no such consensus regarding the surgical management of associated bicuspid aortic valve disease and ascending aortic aneurysmal disease. Although arguments for composite root replacement have been made on theoretical grounds, there is little information regarding the actual long-term risk of dilation of the sinuses. The results of this study are consistent with previous reports demonstrating similar operative risk for both procedures. In 1983, Grey and colleagues [16] compared their early and late results of a separate valve and supracoronary graft repair versus insertion of a valved conduit for aortic regurgitation and ascending aortic
Ann Thorac Surg 2000;69:1333–7
aneurysms. Finding a higher operative risk among those with annuloaortic ectasia undergoing separate valve and graft and, conversely, a higher operative risk for valved conduits when used among those with atherosclerotic aneurysms, they concluded that both approaches have a place in the surgical armamentarium. None of their patients undergoing the separate valve and graft procedure required reoperation, although 5 of their patients had undergone previous separate valve and graft repairs elsewhere. All of these patients were classified as having annuloaortic ectasia. None were thought to have Marfan’s syndrome. No mention was made of bicuspid aortic valve disease. These authors advocated a selective approach based on anatomical considerations. The same year, Antunes and colleagues [17] compared composite root replacement, separate valve and graft repair, and aortoplasty or “tailoring.” Again, equivalent perioperative results were obtained. Although there were no late aneurysms among the separate valve and graft group, the authors expressed a preference for the more radical root replacement on theoretical grounds. Carrel and associates [18] compared their results with the same procedures, finding that the highest operative risk was for composite root replacement and the lowest for tailoring, leading them to the opposite conclusion—that the least radical procedure (tailoring) should be employed whenever possible. In their study the only predictors of late death were age and concomitant coronary artery disease. Two patients in their series required reoperation for dilation of the sinuses after separate valve and graft repair; however, there is no mention regarding Marfan’s syndrome or bicuspid aortic valve disease. Adams and associates [19] reviewed 53 consecutive patients undergoing aortic root procedures, including 14 undergoing separate valve and graft repair. They observed no evidence of late dilation of the sinuses among the supracoronary graft patients and advocated a selective approach. The largest comparative series in the literature has been reported from Stanford University. Yun and colleagues [15] compared the early and late results among 390 patients undergoing separate valve and graft repair (255 patients) or composite root replacement (135). There was no statistically significant difference in operative risk between groups. Multivariate analysis yielded age and coronary artery disease as predictors of late death. The type of operation was not predictive of either early or late death. Seven patients in supracoronary graft group required reoperation. All were described as having had evidence of “gross” annuloaortic ectasia at operation. Of these patients, 5 had acute or chronic dissections and 4 had known Marfan’s syndrome. The authors concluded that a separate valve and graft continued to have a place, but that this approach was contraindicated in the presence of a connective tissue disorder. These findings were remarkably similar to those of Lawrie and colleagues [10] in their follow-up of 277 patients undergoing supracoronary and graft repair of ascending aortic aneurysms. Six patients required reoperation on their ascending aorta, half of whom had Marfan’s syndrome and half of whom had “medial degeneration/necrosis”. Should patients
Ann Thorac Surg 2000;69:1333–7
with bicuspid aortic valve and aneurysmal dilation of the aorta be considered among this group? It is our current practice preferentially to perform root replacement with a composite valve and graft in this population, particularly among younger patients. The low operative risk observed for composite root replacement is reassuring, and confirms other studies [8, 9, 14, 15] indicating that this procedure can be undertaken in the current era with acceptable operative results despite greater technical complexity. We are not reluctant, however, to carry out the technically easier and more expeditious separate valve and supracoronary graft procedure if the coronary ostea are not significantly displaced or when complex concomitant procedures must be performed. The apparent absence of late aneurysmal dilation of the sinuses supports this as a safe option, at least in selected cases. The similar operative risk observed in both groups, despite a higher mean age and more frequent comorbidities in the separate valve and graft group, supports the notion that this simpler procedure may be more appropriate when these additional factors are operative. The principal weakness of this study is its retrospective nature and the limitations so imposed. Beyond the subjective assessment of the surgeon, as reflected in the operative note, the degree of sinus dilatation and coronary displacement in each case is unknown. Clinical experience suggests that bicuspid aortic valve and ascending aortic aneurysm may be a heterogeneous group, with some having annuloaortic ectasia and true root aneurysms and others having relatively normal sinuses and dilation above the sinotubular ridge. The mechanisms of such aneurysms, as well as their therapeutic implications, may be dissimilar. Future studies should include prospective clinical investigation with clear documentation of the pathological anatomy, as well as histologic, biochemical, and molecular biological analysis of resected aortic tissue from such patients. Furthermore, consideration should be given to serial follow-up imaging studies for patients with bicuspid aortic valves undergoing the separate valve and graft procedure. The authors thank Thomas B. Ferguson, Sr, MD, Professor Emeritus, and the following former members of the Division of Cardiothoracic Surgery at Washington University who performed the operations on patients described in this series: R. Morton Bolman III, MD, James L. Cox, MD, Bill B. Daily, MD, T. Bruce Ferguson, Jr, MD, Nicholas T. Kouchoukos, MD, Michael Rosenbloom, MD, Thomas L. Spray, MD, and Clarence S. Weldon, MD. We also gratefully acknowledge Richard B. Schuessler, PhD, for assistance with the statistical analyses.
SUNDT ET AL BICUSPID AORTIC VALVE AND ASCENDING ANEURYSM
1337
References 1. Abbott ME, Hamilton WF. Coarctation of the aorta of the adult type. Am Heart J 1928;3:381– 421, 574 – 618. 2. McKusick VA, Logue RB, Bahnson HT. Association of aortic valvular disease and cystic medial necrosis of the ascending aorta. Report of four cases. Circulation 1957;16:188–94. 3. McKusick VA. Association of congenital bicuspid aortic valve and Erdheim’s cystic medial necrosis. Lancet 1972;1: 1026–7. 4. Edwards WD, Leaf DS, Edwards JE. Dissecting aortic aneurysm associated with congenital bicuspid aortic valve. Circulation 1978;57:1022–5. 5. Pachulski RT, Weinberg AL, Chan KL. Aortic aneurysm in patients with functionally normal or minimally stenotic bicuspid aortic valve. Am J Cardiol 1991;67:781–2. 6. Hahn RT, Roman MJ, Mogtader AH, Devereux RB. Association of aortic dilation with regurgitant, stenotic and functionally normal bicuspid aortic valve. J Am Coll Cardiol 1992; 19:283– 8. 7. Bonderman D, Gharehbaghi-Schnell E, Wolleneck G, Maurer G, Baumgartner H, Lang IM. Mechanisms underlying aortic dilatation in congenital aortic valve malformation. Circulation 1999;99:2138– 43. 8. Kouchoukos NT, Wareing TH, Murphy SF, Perrillo JB. Sixteen-year experience with aortic root replacement. Ann Surg 1991;214:308–20. 9. Gott VL, Greene PS, Alejo DE, et al. Replacement of the aortic root in patients with Marfan’s syndrome. N Engl J Med 1999;340:1307–13. 10. Lawrie GM, Earle N, DeBakey ME. Long-term fate of the aortic root and aortic valve after ascending aneurysm surgery. Ann Surg 1993;217:711–20. 11. Roberts WC. The congenitally bicuspid aortic valve. Am J Cardiol 1970;26:72– 83. 12. Prenger K, Pieters F, Cheriex E. Aortic dissection after aortic valve replacement: incidence and consequences for strategy. J Card Surg 1994;9:495–9. 13. Pieters FA, Widdershoven JW, Gerardy AC, Geskes G, Cheriex EC, Wellens HJ. Risk of aortic dissection after aortic valve replacement. Am J Cardiol 1993;72:1043–7. 14. Gott VL, Cameron DE, Pyeritz RE, et al. Composite graft repair of Marfan aneurysm of the ascending aorta: results in 150 patients. J Card Surg 1994;9:482–9. 15. Yun KL, Miller DC, Fann JI, et al. Composite valve graft versus separate aortic valve and ascending aortic replacement. Is there still a role for the separate procedure? Circulation 1997;96(Suppl II):II-368 –II-375. 16. Grey DP, Ott DA, Cooley DA. Surgical treatment of aneurysm of the ascending aorta with aortic insufficiency. J Thorac Cardiovasc Surg 1983;86:864–77. 17. Antunes MJ, Baptista AL, Colsen PR, Kinsley RH. Surgical treatment of aneurysms of the ascending aorta associated with severe aortic regurgitation. Thorax 1984;39:305–10. 18. Carrel T, von Segesser L, Jenni R, et al. Dealing with dilated ascending aorta during aortic valve replacement: advantages of conservative surgical approach. Eur J Cardiothorac Surg 1991;5:137– 43. 19. Adams RD, Goldin MD, Najafi H. Selectivity in aortic root reconstruction. J Card Surg 1994;9:500–7.
DISCUSSION DR MICHAEL J. REARDON (Houston, TX): This is a very nice paper. Of course it is fairly easy when you have a great big aorta to decide what to do. The interesting question is, how small do you go in replacing these aortas, what is your lower limit cutoff? DR SUNDT: Doctor Kouchoukos taught me to replace an aorta that is greater than 4 cm in diameter in the presence of a
bicuspid aortic valve. There are sufficient data in the literature that the risk of dissection following aortic valve replacement for bicuspid aortic valve is significantly greater, particularly when the aorta is greater than 5 cm in diameter. I replace aortas down to about 4 cm in diameter as a general rule, and have felt this is a reasonable practice.
Background. Intrinsic abnormality of the aortic wall may explain the association of bicuspid aortic valves with ascending aortic aneurysms. Separate valve and graft repair of such lesions, rather than composite valve graft replacement, is more straightforward but leaves potentially abnormal sinuses behind. Methods. Between January 1985 and January 1998, 45 patients underwent separate valve and graft (n ⴝ 27) or composite valve graft (n ⴝ 18) for an ascending aortic aneurysm and bicuspid aortic valve. Perioperative events and late results were compared. Results. Patients undergoing separate valve and graft were older (mean age, 60 ⴞ 13 vs 42 ⴞ 12 years, p < 0.001) and were more likely to have purely stenotic (48% vs 6%, p ⴝ 0.003) than purely regurgitant (11% vs 72%, p < 0.001)
disease. They were also more likely to require concomitant coronary artery bypass grafting (56% vs 6%, p ⴝ 0.001). There were no significant differences in operative risk and no known late complications related to recurrent aneurysms. Conclusions. Root replacement with a composite valve graft can be accomplished with low operative risk and is the first choice for repair of this lesion. Separate valve and graft repair, however, yields satisfactory early and late results and remains an acceptable option, especially when the coronary ostea are not displaced or when concomitant procedures must be performed.
A
of hemodynamically significant aortic stenosis. Studying patients with functionally normal bicuspid valves and those with varying degrees of hemodynamic abnormalities, Hahn and associates [6] demonstrated that root enlargement was widespread. Enlargement was greatest among those with regurgitant lesions. Furthermore, recent histochemical studies have revealed an increased rate of smooth muscle cell apoptosis in the aortic media of patients with bicuspid aortic valves, even in the absence of gross dilation [7]. The presence or absence of intrinsic abnormality of the aortic wall has profound implications for the proper surgical management of this condition. A strategy of valve replacement and separate supracoronary graft repair of the ascending aorta, leaving the sinuses intact, is straightforward and obviates risks related to coronary osteal reimplantation, but may also subject the patient to the risk of reoperation for aneurysmal dilation of the unreplaced root elements. Composite root replacement is a well-established technique, with the operative mortality rate reported in several series to be as low as 2% for elective procedures [8, 9]. It is now widely accepted as the standard of care for patients with Marfan’s syndrome, in whom the risk of aneurysmal dilation of the sinuses has been shown to be considerable [10]. Without displacement of the coronary ostea, however, composite root replacement can be difficult, and it may be a more time-consuming procedure than separate valve replacement and supracoronary graft replacement of the ascending aorta. There is, as yet, no consensus regarding
n association between bicuspid aortic valve and diseases of the aorta has long been recognized, although the causal relationship remains undefined. Abbott and Hamilton [1] reported this observation, as well as an apparent association with propensity for rupture of the ascending aorta, in 1928 based on study of coarctation in the adult. They hypothesized a developmental relationship between bicuspid aortic valve disease and intrinsic weakness of the aorta. Subsequently, McKusick and colleagues [2] reported an association between bicuspid aortic valve and dissection of the ascending aorta with cystic medial necrosis. Originally favoring a hemodynamic explanation, further clinical observations led them to shift their position, favoring a common developmental defect with variable expression among individuals leading to a bicuspid aortic valve, coarctation of the aorta, and cystic medial necrosis [3]. These seminal observations have been reconfirmed by others [4], and the hypothesis of intrinsic abnormality has increasingly been embraced. Several recent echocardiographic studies have supported this notion. Pachulkski and colleagues [5] found that the aortic root diameter was significantly larger among patients with bicuspid aortic valves than among normal controls, even in the absence
Presented at the Forty-sixth Annual Meeting of the Southern Thoracic Surgical Association, San Juan, Puerto Rico, Nov 4 – 6, 1999. Address reprint requests to Dr Sundt, Division of Cardiothoracic Surgery, Washington University School of Medicine, Suite 3106 Queeny Tower, One Barnes Hospital Plaza, St. Louis, MO 63110; e-mail: [email protected].
© 2000 by The Society of Thoracic Surgeons Published by Elsevier Science Inc
(Ann Thorac Surg 2000;69:1333–7) © 2000 by The Society of Thoracic Surgeons
0003-4975/00/$20.00 PII S0003-4975(00)01220-0
1334
SUNDT ET AL BICUSPID AORTIC VALVE AND ASCENDING ANEURYSM
the management of ascending aneurysms associated with bicuspid valve disease. We therefore reviewed our institutional experience with this condition over the preceding 13 years to determine comparative operative risk and long-term outcome of separate valve and graft repair versus composite root replacement for ascending aortic aneurysms in association with bicuspid aortic valves.
Material and Methods Retrospective review of the computerized database for the Division of Cardiothoracic Surgery at Washington University identified 126 patients undergoing either composite root replacement or separate valve and graft repair of aortic valve pathology and an ascending aortic aneurysm between January 1985 and January 1998. In 45 patients, the aortic valve was documented to be bicuspid in either the operative note or the pathology report. The decision to address dilation of the ascending aorta was made by the operating surgeon. There is some variation among surgeons; however, it is our general practice to intervene routinely on an ascending aorta greater than 6 centimeters in diameter as determined by preoperative computed tomography, echocardiography, or intraoperative inspection. In the presence of a bicuspid valve, we consider graft replacement when the ascending aorta is more than 4 centimeters in diameter. Aortic “tailoring” for dilation of the ascending aorta is only rarely practiced. Patients undergoing this procedure were therefore excluded from this study. Preoperative demographic characteristics, operative variables, and perioperative events were retrieved from the database. Late follow-up information was obtained by postal questionnaire or telephone interview between July 1, 1998, and December 31, 1998. Follow-up was 100% complete. The mean follow-up was 5.6 (⫾ 4.1) years, and the total number of patient-years of follow-up was 243 years.
Surgical Technique Surgical procedures were carried out by the members of the Division of Cardiothoracic Surgery at Washington University both at Barnes Hospital and Jewish Hospital. Procedures were carried out under hypothermic cardiopulmonary bypass with cardioplegic arrest for myocardial protection. Initially antegrade cold crystalloid cardioplegia was employed. More recently cold blood cardioplegia, often given both antegrade and retrograde, has been used. Profound hypothermia and circulatory arrest were employed when concomitant hemiarch or full arch replacement was undertaken. The operating surgeon made the choice between procedures. Composite root replacement was performed most often using the open coronary button technique as previously reported by Kouchoukos and colleagues [8]. Separate graft replacement of the ascending aorta was performed with proximal anastomosis to the supracoronary ridge. The aorta was transected completely before distal anastomosis.
Ann Thorac Surg 2000;69:1333–7
Table 1. Demographic Characteristics of the Study Population Variable
CVG (n ⫽ 18)
Age (y) Mean ⫾ SD 42 ⫾ 12 Range 25–70 Male sex 15 (83%) Marfan’s syndrome 1 (6%) Diabetes mellitus 2 (11%) Hypertension 6 (33%) Tobacco abuse 6 (33%) Renal dysfunction 1 (6%) Hypercholesterolemia 3 (16%) Peripheral vascular 2 (11%) disease Cerobrovascular 1 (6%) disease Previous CAB 0 Functional aortic valve pathology Stenosis 1 (6%) Regurgitation 13 (72%) Mixed 4 (22%) Ascending aortic 6.1 ⫾ 1.8 cm diametera
Separate V ⫹ G (n ⫽ 27) 60 ⫾ 13 32–78 21 (77%) 0 4 (15%) 12 (44%) 8 (30%) 0 7 (26%) 0
p Value ⬍ 0.001 0.65 0.40 0.94 0.54 0.94 0.40 0.72 0.15
3 (11%)
0.64
3 (11%)
0.26
13 (48%) 3 (11%) 11 (41%) 5.4 ⫾ 1.2 cm
0.003 ⬍ 0.001 0.33 0.21
a
Determined by visual inspection at surgery, computed tomography, or echocardiography.
CAB ⫽ coronary artery bypass; V ⫹ G ⫽ valve plus graft.
CVG ⫽ composite valve and graft;
Statistical Analysis All continuous data were expressed as the mean plus or minus the standard deviation. Two-group comparisons for continuous data were made using the Student’s t test. Categorical data were tabulated, and two-group comparisons were made using the 2 test for 2 ⫻ n tables. When 2 ⫻ 2 tables were analyzed, Fisher’s exact test was used. The survival analysis was done using the Kaplan-Meier estimator. Comparisons were made using the Mantel log-rank test. The Cox proportional hazards model was used to examine the effect of age (continuous variable) on the survivor function. A p value of less than 0.05 was considered significant for all statistical calculations. The SYSTAT system for statistics was used for all data analysis (version 6.0 for Windows; SPSS, Chicago, IL).
Results Study Population The demographic characteristics of the study groups are shown in Table 1. Patients undergoing composite root replacement were significantly younger than those having a separate valve and graft, and were more likely to have predominantly regurgitant rather than stenotic lesions. The only patient in the study group who met clinical criteria for the Marfan’s syndrome underwent composite root replacement. Aortic diameters were similar in both groups.
Ann Thorac Surg 2000;69:1333–7
SUNDT ET AL BICUSPID AORTIC VALVE AND ASCENDING ANEURYSM
Table 2. Operative Characteristics Variable
a
Table 3. Perioperative Results
CVG Separate V ⫹ G (n ⫽ 18) (n ⫽ 27) p Value
CPB time (min) Mean ⫾ SD 169 ⫾ 45 Range (109 –272) Aortic occlusion time (min) Mean ⫾ SD 132 ⫾ 43 Range (77–241) Associated procedures CAB 1 (6%) Arch replacement 1 (6%) Aortic annular enlargement 0 VSD closure 0 Repair aortic coarctation 1 (6%) Circulatory arrest used 1 (6%) Valve size (mm) Mean ⫾ SD 26.3 Range 25–31 Mechanical valve 12 (92%)a Biological valve 1 (8%)a
1335
163 ⫾ 53 (92–294)
0.70
115 ⫾ 28 (60 –168)
0.15
15 (56%) 6 (22%) 1 (4%) 1 (4%) 0 6 (22%)
0.001 0.22 0.46 0.46 0.40 0.22
25.0 21–31 16 (59%) 11 (41%)
0.15
Valve prosthesis type known for 13 of 18 CVG patients.
CAB ⫽ coronary artery bypass; CPB ⫽ cardiopulmonary bypass; CVG ⫽ composite valve and graft; V ⫹ G ⫽ valve plus graft; VSD ⫽ ventricular septal defect.
Operative procedures tended to be more complex in the separate valve and graft group, as shown in Table 2. Concomitant coronary artery bypass was performed more often in the separate valve and graft group, as was aortic arch replacement, although the latter did not reach statistical significance. Despite the more frequent use of circulatory arrest in the separate valve and graft group, the mean cardiopulmonary bypass times for both groups were nearly identical. There was a trend toward somewhat longer aortic occlusion time among patients undergoing composite root replacement, reflecting the increased technical complexity of this procedure. Composite replacement was done almost exclusively with mechanical prostheses. Mechanical valves also predominated in the separate valve and graft group, but more narrowly so.
Variable
CVG (n ⫽ 18)
Mortality IABP support Re-exploration for bleeding Prolonged ventillation Stroke Renal dysfunction Intraoperative blood used Postoperative blood used Total units pRBCs Total units FFP Total units platelets
0 0 0 1 (5.5%) 2 (11%) 2 (11%) 10 (56%) 9 (50%) 2.1 ⫾ 2.5 0.8 ⫾ 1.6 1.0 ⫾ 1.8
Separate V ⫹ G (n ⫽ 27) p Value 2 (7.4%) 2 (7.4%) 2 (7.4%) 4 (14.8%) 0 3 (11%) 18 (67%) 12 (44%) 3.44 ⫾ 4.1 2.5 ⫾ 5.0 1.7 ⫾ 3.2
0.51 0.51 0.51 0.23 0.16 1.0 0.76 0.49 0.16 0.10 0.32
CVG ⫽ composite valve and graft; FFP ⫽ fresh frozen plasma; pRBC ⫽ packed red blood cells; V ⫹ G ⫽ valve plus graft.
ing, and separate graft replacement of the ascending aorta. On postoperative day 1 he exhibited hemodynamic instability and underwent operative exploration. He was found to have right heart failure with thrombosis of a bypass graft to the posterior descending coronary artery. He expired despite graft thrombectomy and institution of ventricular assist device support. There were no statistically significant differences in rates of perioperative intraaortic balloon support, stroke, renal dysfunction, or respiratory insufficiency between groups. There were also no statistically significant differences in perioperative transfusion or in the rates of exploration for bleeding.
Late Results Actuarial survival was superior ( p ⫽ 0.04) among patients undergoing composite root replacement, as shown in Figure 1. However, multivariate stepwise logistic regression analysis of risk factors for late death, including the type of aortic procedure performed, diabetes mellitus, preoperative renal dysfunction, hypercholesterolemia, peripheral vascular disease, previous or concomitant coronary bypass or aortic arch replacement, and
Perioperative Complications There was no difference in perioperative mortality between groups (Table 3). Two patients undergoing separate valve and graft reconstruction died within 30 days of surgery. One was a 52-year-old man presenting with aortic stenosis, who had undergone mechanical aortic valve replacement and supracoronary graft replacement of the ascending aorta. Twenty-five days after hospital discharge he was readmitted with a markedly elevated prothrombin time (29 seconds) and near syncope. He experienced a cardiopulmonary arrest shortly after arrival and could not be resuscitated. Postmortem examination demonstrated mediastinal hematoma. The second perioperative mortality occurred in a 68-year-old man with a history of previous coronary artery bypass, who underwent aortic valve replacement, repeat bypass graft-
Fig 1. Kaplan-Meier actuarial survival estimates for composite valve and graft (CVG) and separate valve and graft (V ⫹ G) replacement of the aortic valve and ascending aorta for bicuspid aortic valve and associated ascending aortic aneurysm. Multivariate analysis identified only age, and not procedure, as a predictor of late death.
1336
SUNDT ET AL BICUSPID AORTIC VALVE AND ASCENDING ANEURYSM
functional pathology of the valve, yielded only age as a significant predictor of death. There were no known late reoperations for remaining or recurrent pathology of the aortic root. Similarly, there were no late deaths known to be related to recurrent aneurysms of the aortic root.
Comment The association between intrinsic aortic disease and bicuspid aortic valve disease is of practical importance. Bicuspid aortic valve disease occurs in 1% to 2% of the population. In autopsy studies it has been reported in more than one-half of patients with anatomically isolated aortic stenosis and in more than 25% of those with clinically pure aortic regurgitation [11]. Conversely, among 85 cases of bicuspid aortic valve identified at autopsy, Roberts [11] reported that 61 patients (72%) had stenotic or mixed lesions, 11 (13%) had purely regurgitant lesions, and 13 (15%) had functionally normal valves. The risk of aortic dissection after aortic valve replacement for a bicuspid aortic valve is well documented. The studies by Prenger and colleagues [12] and Pieters and colleagues [13] demonstrated a substantial risk (22% to 27%) of aortic dissection after AVR when the aorta is more than 5 centimeters in diameter. In the study by Prenger and colleagues, all patients with aortic dissection had a diameter of more than 5 cm. Accordingly, it is our practice to be particularly aggressive with the treatment of the dilated aorta during aortic valve replacement in the presence of a bicuspid valve. Debate over the choice between composite root replacement and separate valve and graft replacement of the ascending aorta is not new; however, little has been said on the subject in the context of bicuspid aortic valve disease. The latter technique has proved inadequate in the presence of Marfan’s syndrome [10], the archtypical connective tissue abnormality, and composite root replacement is now the standard of care in this circumstance. Composite root replacement can be accomplished in this condition with very low operative risk under elective circumstances [9], particularly in experienced centers [14]. Gott and colleagues [9] recently reported an operative mortality of 1.5% among 455 patients with Marfan’s syndrome undergoing elective root replacement at 10 experienced centers. The series of 138 such consecutive elective procedures without a single operative death reported from Johns Hopkins sets the standard for the field [14]. There is, however, no such consensus regarding the surgical management of associated bicuspid aortic valve disease and ascending aortic aneurysmal disease. Although arguments for composite root replacement have been made on theoretical grounds, there is little information regarding the actual long-term risk of dilation of the sinuses. The results of this study are consistent with previous reports demonstrating similar operative risk for both procedures. In 1983, Grey and colleagues [16] compared their early and late results of a separate valve and supracoronary graft repair versus insertion of a valved conduit for aortic regurgitation and ascending aortic
Ann Thorac Surg 2000;69:1333–7
aneurysms. Finding a higher operative risk among those with annuloaortic ectasia undergoing separate valve and graft and, conversely, a higher operative risk for valved conduits when used among those with atherosclerotic aneurysms, they concluded that both approaches have a place in the surgical armamentarium. None of their patients undergoing the separate valve and graft procedure required reoperation, although 5 of their patients had undergone previous separate valve and graft repairs elsewhere. All of these patients were classified as having annuloaortic ectasia. None were thought to have Marfan’s syndrome. No mention was made of bicuspid aortic valve disease. These authors advocated a selective approach based on anatomical considerations. The same year, Antunes and colleagues [17] compared composite root replacement, separate valve and graft repair, and aortoplasty or “tailoring.” Again, equivalent perioperative results were obtained. Although there were no late aneurysms among the separate valve and graft group, the authors expressed a preference for the more radical root replacement on theoretical grounds. Carrel and associates [18] compared their results with the same procedures, finding that the highest operative risk was for composite root replacement and the lowest for tailoring, leading them to the opposite conclusion—that the least radical procedure (tailoring) should be employed whenever possible. In their study the only predictors of late death were age and concomitant coronary artery disease. Two patients in their series required reoperation for dilation of the sinuses after separate valve and graft repair; however, there is no mention regarding Marfan’s syndrome or bicuspid aortic valve disease. Adams and associates [19] reviewed 53 consecutive patients undergoing aortic root procedures, including 14 undergoing separate valve and graft repair. They observed no evidence of late dilation of the sinuses among the supracoronary graft patients and advocated a selective approach. The largest comparative series in the literature has been reported from Stanford University. Yun and colleagues [15] compared the early and late results among 390 patients undergoing separate valve and graft repair (255 patients) or composite root replacement (135). There was no statistically significant difference in operative risk between groups. Multivariate analysis yielded age and coronary artery disease as predictors of late death. The type of operation was not predictive of either early or late death. Seven patients in supracoronary graft group required reoperation. All were described as having had evidence of “gross” annuloaortic ectasia at operation. Of these patients, 5 had acute or chronic dissections and 4 had known Marfan’s syndrome. The authors concluded that a separate valve and graft continued to have a place, but that this approach was contraindicated in the presence of a connective tissue disorder. These findings were remarkably similar to those of Lawrie and colleagues [10] in their follow-up of 277 patients undergoing supracoronary and graft repair of ascending aortic aneurysms. Six patients required reoperation on their ascending aorta, half of whom had Marfan’s syndrome and half of whom had “medial degeneration/necrosis”. Should patients
Ann Thorac Surg 2000;69:1333–7
with bicuspid aortic valve and aneurysmal dilation of the aorta be considered among this group? It is our current practice preferentially to perform root replacement with a composite valve and graft in this population, particularly among younger patients. The low operative risk observed for composite root replacement is reassuring, and confirms other studies [8, 9, 14, 15] indicating that this procedure can be undertaken in the current era with acceptable operative results despite greater technical complexity. We are not reluctant, however, to carry out the technically easier and more expeditious separate valve and supracoronary graft procedure if the coronary ostea are not significantly displaced or when complex concomitant procedures must be performed. The apparent absence of late aneurysmal dilation of the sinuses supports this as a safe option, at least in selected cases. The similar operative risk observed in both groups, despite a higher mean age and more frequent comorbidities in the separate valve and graft group, supports the notion that this simpler procedure may be more appropriate when these additional factors are operative. The principal weakness of this study is its retrospective nature and the limitations so imposed. Beyond the subjective assessment of the surgeon, as reflected in the operative note, the degree of sinus dilatation and coronary displacement in each case is unknown. Clinical experience suggests that bicuspid aortic valve and ascending aortic aneurysm may be a heterogeneous group, with some having annuloaortic ectasia and true root aneurysms and others having relatively normal sinuses and dilation above the sinotubular ridge. The mechanisms of such aneurysms, as well as their therapeutic implications, may be dissimilar. Future studies should include prospective clinical investigation with clear documentation of the pathological anatomy, as well as histologic, biochemical, and molecular biological analysis of resected aortic tissue from such patients. Furthermore, consideration should be given to serial follow-up imaging studies for patients with bicuspid aortic valves undergoing the separate valve and graft procedure. The authors thank Thomas B. Ferguson, Sr, MD, Professor Emeritus, and the following former members of the Division of Cardiothoracic Surgery at Washington University who performed the operations on patients described in this series: R. Morton Bolman III, MD, James L. Cox, MD, Bill B. Daily, MD, T. Bruce Ferguson, Jr, MD, Nicholas T. Kouchoukos, MD, Michael Rosenbloom, MD, Thomas L. Spray, MD, and Clarence S. Weldon, MD. We also gratefully acknowledge Richard B. Schuessler, PhD, for assistance with the statistical analyses.
SUNDT ET AL BICUSPID AORTIC VALVE AND ASCENDING ANEURYSM
1337
References 1. Abbott ME, Hamilton WF. Coarctation of the aorta of the adult type. Am Heart J 1928;3:381– 421, 574 – 618. 2. McKusick VA, Logue RB, Bahnson HT. Association of aortic valvular disease and cystic medial necrosis of the ascending aorta. Report of four cases. Circulation 1957;16:188–94. 3. McKusick VA. Association of congenital bicuspid aortic valve and Erdheim’s cystic medial necrosis. Lancet 1972;1: 1026–7. 4. Edwards WD, Leaf DS, Edwards JE. Dissecting aortic aneurysm associated with congenital bicuspid aortic valve. Circulation 1978;57:1022–5. 5. Pachulski RT, Weinberg AL, Chan KL. Aortic aneurysm in patients with functionally normal or minimally stenotic bicuspid aortic valve. Am J Cardiol 1991;67:781–2. 6. Hahn RT, Roman MJ, Mogtader AH, Devereux RB. Association of aortic dilation with regurgitant, stenotic and functionally normal bicuspid aortic valve. J Am Coll Cardiol 1992; 19:283– 8. 7. Bonderman D, Gharehbaghi-Schnell E, Wolleneck G, Maurer G, Baumgartner H, Lang IM. Mechanisms underlying aortic dilatation in congenital aortic valve malformation. Circulation 1999;99:2138– 43. 8. Kouchoukos NT, Wareing TH, Murphy SF, Perrillo JB. Sixteen-year experience with aortic root replacement. Ann Surg 1991;214:308–20. 9. Gott VL, Greene PS, Alejo DE, et al. Replacement of the aortic root in patients with Marfan’s syndrome. N Engl J Med 1999;340:1307–13. 10. Lawrie GM, Earle N, DeBakey ME. Long-term fate of the aortic root and aortic valve after ascending aneurysm surgery. Ann Surg 1993;217:711–20. 11. Roberts WC. The congenitally bicuspid aortic valve. Am J Cardiol 1970;26:72– 83. 12. Prenger K, Pieters F, Cheriex E. Aortic dissection after aortic valve replacement: incidence and consequences for strategy. J Card Surg 1994;9:495–9. 13. Pieters FA, Widdershoven JW, Gerardy AC, Geskes G, Cheriex EC, Wellens HJ. Risk of aortic dissection after aortic valve replacement. Am J Cardiol 1993;72:1043–7. 14. Gott VL, Cameron DE, Pyeritz RE, et al. Composite graft repair of Marfan aneurysm of the ascending aorta: results in 150 patients. J Card Surg 1994;9:482–9. 15. Yun KL, Miller DC, Fann JI, et al. Composite valve graft versus separate aortic valve and ascending aortic replacement. Is there still a role for the separate procedure? Circulation 1997;96(Suppl II):II-368 –II-375. 16. Grey DP, Ott DA, Cooley DA. Surgical treatment of aneurysm of the ascending aorta with aortic insufficiency. J Thorac Cardiovasc Surg 1983;86:864–77. 17. Antunes MJ, Baptista AL, Colsen PR, Kinsley RH. Surgical treatment of aneurysms of the ascending aorta associated with severe aortic regurgitation. Thorax 1984;39:305–10. 18. Carrel T, von Segesser L, Jenni R, et al. Dealing with dilated ascending aorta during aortic valve replacement: advantages of conservative surgical approach. Eur J Cardiothorac Surg 1991;5:137– 43. 19. Adams RD, Goldin MD, Najafi H. Selectivity in aortic root reconstruction. J Card Surg 1994;9:500–7.
DISCUSSION DR MICHAEL J. REARDON (Houston, TX): This is a very nice paper. Of course it is fairly easy when you have a great big aorta to decide what to do. The interesting question is, how small do you go in replacing these aortas, what is your lower limit cutoff? DR SUNDT: Doctor Kouchoukos taught me to replace an aorta that is greater than 4 cm in diameter in the presence of a
bicuspid aortic valve. There are sufficient data in the literature that the risk of dissection following aortic valve replacement for bicuspid aortic valve is significantly greater, particularly when the aorta is greater than 5 cm in diameter. I replace aortas down to about 4 cm in diameter as a general rule, and have felt this is a reasonable practice.