- Email: [email protected]
Composite replacement of the aortic valve and ascending aorta
Composite replacement of the aortic valve and ascending aorta This report comprises 16 consecutive patients with ascending aortic aneurysms caused by cystic medial necrosis. We replaced the ascending aorta and aortic valve with a tightly woven Dacron graft containing a Lillehei-Kaster valve prosthesis and implanted the coronary ostia in the sides of the graft. All but two patients had massive aortic insufficiency. Postoperative catheterization was performed in 13 patients, and all surviving patients have been seen within the past 6 months. There was one perioperative death (6 percent) and two late deaths. Eleven surviviors are in Class I and two are in Class II (N.Y.H.A.). Angiographically demonstrated late complications have included pseudoaneurysms of the coronary ostium (two), paravalvular leak (one), and pseudoaneurysm of the distal suture line (one). Two of these four patients were asymptomatic. Two of the four patients have had successful repair of these defects and a third is awaiting operation. Composite replacement carries a low operative risk and minimizes problems of intraoperative bleeding. In view of the incidence of late suture line problems, routine angiography 6 to 12 months postoperatively is recommended. If new symptoms occur or if there is a change in the cardiac silhouette on chest roentgenogram, the patient should be recatheterized.
John E. Mayer, Jr., M.D. (by invitation), W. G. Lindsay, M.D. (by invitation), Y. Wang, M.D. (by invitation), C. R. Jorgensen, M.D. (by invitation), and D. M. Nicoloff, M . D . , Minneapolis, Minn.
V_^ystic medial necrosis primarily affects the ascending aorta and sinuses of Valsalva, the result being aneurysum formation and secondary aortic insufficiency. Surgical treatment for this condition has evolved through several stages since the initial report of Bahnson and Nelson 1 in 1956. In 1964 Wheat and Colleagues 2 reported replacement of the aortic valve and ascending aorta. In that procedure the coronary ostia were left attached to the rim of aortic wall, and a graft to replace the ascending aorta was sewn to the skirt of an aortic valve prosthesis and rim of aortic wall to encompass the coronary ostia. Bentall and DeBono 3 described the technique in which the ascending aorta and aortic valve are replaced with a composite unit of tubular qraft sewn to the aortic anulus with anastomosis of the coronary ostia to the sides of the graft. Since those reports, others have reported generally favorable experience with this technique or slight modifica-
From the Departments of Surgery and Medicine, University of Minnesota Hospitals, Minneapolis, Minn. 55455. Read at the Fifty-eighth Annual Meeting of The American Association for Thoracic Surgery, New Orleans, La., May 8, 9, and 10, 1978. Address for reprints: D. M. Nicoloff, M.D., Department of Surgery, University of Minnesota Hospitals, Minneapolis, Minn. 55455. 816
tions. 4 - 1 0 However, disagreement exists as to whether the composite graft technique represents optimal treatment for aortic root aneurysms with aortic insufficiency,11-14 and there is insufficient long-term follow-up information. The present report reviews a 4Vi year experience with composite grafts with data on postoperative catheterization and long-term follow-up. Patients Sixteen patients ranging in age from 13 to 61 years underwent replacement of the aortic valve and ascending aorta with a composite graft from July, 1973, through April, 1978. All had aneurysms of the ascending aorta and all but two had associated massive aortic insufficiency. One patient had an aneurysm involving the anulus and aortic root without aortic insufficiency, and a second patient had calcific aortic stenosis (94 mm. Hg gradient). Six patients had clinical features of the Marfan syndrome, one of whom had previously had a supracoronary aortic graft inserted for a dissection. Histologic proof of cystic medial necrosis was present in the seven cases in which a biopsy of the aortic wall was submitted. In all cases the aneurysm was typical of that associated with cystic medial necrosis. Acute or semi-acute (less than 2 weeks) dissections were present in five patients. Only one patient was asymptomatic
0022-5223/78/120816+08S00.80/0 © 1978 The C. V. Mosby Co.
Volume 76 Number 6 December, 1978
Replacement of aortic valve and ascending aorta
"\Jfr
^J&
817
+\Jh-
Decompress LV . ,, LC artery A \ y / / - y perfusion
iC artery perfusion
^Mt
Fig. 1. Technique of composite replacement (see text). LC, Left coronary, RC, Right coronary, LV, Left ventricle. prior to operation, with five in New York Heart Association Class II and eight in Class III. The two patients with acute dissections were placed in Class IV. Surgical technique (Fig. 1) In all cases cardiopulmonary bypass with moderate systemic hypothermia (30° C.) and arterial return through a femoral artery cannula was utilized. Coronary perfusion was used in all cases (Fig. 1, c). Since February, 1976, (Six patients), a second infant heat exchanger has been added to the coronary perfusion
lines to lower the temperature of the coronary blood to 18 to 20° C. The technique of composite replacement is similar to that described by others3' 5- 7 but differs in that a Lillehei-Kaster pivoting disc prosthesis has been exclusively utilized in this series (Fig. 2). Valve sizes are listed in Table I. The prosthesis is sewn into the proximal end of a tightly woven Dacron graft,* and the graft-valve composite unit is then sewn to the aortic anulus with 2-0 Mersilene mattress sutures reinforced *USCI, Division of C. R. Bard, Inc., Billerica, Mass.
The Journal of Thoracic and Cardiovascular Surgery
8 1 8 Mayer et al.
Table I. Valve sizes (Lillehei-Kaster) Size
No.
18A 20A 22A 25A
2 2 4 8
Table II. Associated procedures Procedure
No.
Coronary artery bypass Mitral valve replacement Arch aneurysm graft
1 1 1
was utilized to keep systolic blood pressure below 120 mm. Hg. All patients were initially treated with Coumadin with recommendations to maintain prothrombin times at two times control. Fig. 2. Lillehei-Kaster aortic prosthesis sewn into tightly woven Dacron graft. with Vi6 inch Teflon felt pledgets (Fig. 1, c). The coronary ostia are sewn to side holes in the graft with continuous 5-0 polypropylene (Prolene) suture (Fig. 1, d). After the completion of the coronary anastomoses, a slit is made in the anterior wall of the graft through which coronary perfusion cannulas are inserted to provide coronary perfusion during the completion of the distal aortic anastomoses. In all but one case the distal anastomosis was to the aorta proximal to the innominate artery. When the distal anastomosis is completed, the coronary cannulas are removed and the slit in the graft is repaired (Fig. 1, e). Air is evacuated from the graft and left ventricle by inserting an air bleed button into the graft, and the aortic cross-clamp is removed. The wall of the aneurysm is trimmed and sewn over the graft to aid hemostasis (Fig. \,f). Three patients underwent additional procedures at the time when the composite graft was inserted (Table II). In one patient the mitral valve was replaced because of severe mitral insufficiency owing to prolapsing leaflets. A second patient had saphenous vein grafts to the anterior descending and right coronary arteries for atherosclerotic lesions. In the third patient the graft was extended to a point distal to the left subclavian take-off because the aneurysm involved the aortic arch. The three arch vessels were anastomosed to the graft as a unit by preserving the aortic wall between their ostia and then suturing the island of aortic wall to the graft. In the immediate postoperative period, nitroprusside
Postoperative follow-up Thirteen of the 16 patients returned for elective postoperative cardiac catheterization. In seven patients a catheter was passed across the atrial septum and into the left ventricle and a second catheter was placed in the ascending aorta so that peak systolic transvalvular gradients could be measured. Supra valvular aortography and coronary arteriography were performed in all patients. Further follow-up has been obtained in our clinics or by telephone contact with the patient or the referring physician. Average length of follow-up is 26 months (range 2 to 51 months). Results Early (Table III). Fifteen of the 16 patients survived the operative procedure and the early postoperative period. The single early death occurred in a patient operated upon for an acute dissection who had renal failure postoperatively. During hemodialysis, massive bleeding occurred around the right coronary orifice, and the patient died. A second patient had transient renal failure and an empyema after a transfusion reaction but survived. A severe ventricular arrhythmia controlled only by Bretylium occurred intraoperatively in one patient who survived, and one patient with an acute dissection had an intraoperative myocardial infarction. The remainder of the patients had uncomplicated courses. No patient was re-explored for postoperative bleeding. The average cardiopulmonary bypass time for elective cases without additional procedures was 115 minutes. The average bypass time including pa-
Volume 76 Number 6 December, 1978
819
Replacement of aortic valve and ascending aorta
Fig. 3. Contrast outlines left coronary artery and small pseudoaneurysm arising from the inferior margin of the suture line. Table m . Early complications (four patients) Complication Operative death Hemorrhage Renal failure Empyeme Myocardial infarction Severe arrhythmia
No. 1(%) 1 2 1 1 1
tients with acute dissections and additional procedures was 161 minutes. Late (Table IV). Of the 15 survivors, two have died in the late postoperative period. Eleven of the survivors are in N.Y.H.A. Class I, and two are in Class II. One late death occurred in a 13-year-boy whose preoperative left ventricular end-diastolic pressure was 38 mm. Hg. He died 5 months postoperatively of end-stage cardiac failure. Repeat catheterization 4 weeks prior to death had shown a 4 mm. Hg peak gradient across the prosthesis, patent coronary anastomoses, and no aortic regurgitation. A second patient died 12 months postoperatively. Routine angiography had shown a small pseudoaneurysm at the distal suture line and patency of the two saphenous vein grafts. Coumadin was discontinued because of the pseudoaneurysm, but the patient remained asymptomatic until she died suddenly at home. No autopsy was performed. Two patients have had pseudoaneurysms form at the anastomoses between the graft and the coronary arteries. One patient with the Marfan syndrome has aneurysms at both ostia which
Table IV. Late complications (seven patients) Complication Death Coronary ostia pseudoaneurysm Aortic paravalvular leak Distal suture-line pseudoaneurysm Transient neurologic changes (? emboli) Spinal cord hemorrhage
No.
2 2 1 1 2 1
were discovered 14 months postoperatively on a routine catheterization (Fig. 3). She is scheduled for reoperation in the near future. The second patient began having chest pain 3 ^ years postoperatively, and a left coronary pseudoaneurysm was found by angiography. A previous angiogram 7 months postoperatively was normal. He underwent reoperation and successful repair of the small suture-line dehiscence. One additional patient with the full Marfan syndrome was found to have a mitral paravalvular leak 10 months postoperatively; the leak was repaired by resuture of the dehisced area of the suture line. An aortogram was normal at this time. This same patient had an aortic paraprosthetic dehiscence 20 months after the initial operation, and this has also been repaired. Two additional patients have had transient episodes of neurologic symptoms which may represent emboli. They have been treated by adding aspirin and dipyridamole to their warfarin regimen. A final patient had a subdural hemorrhage around the cervical spinal cord despite a prothrombin time of 19 seconds (control 11 seconds).
8 20
The Journal of Thoracic and Cardiovascular Surgery
Mayer et al.
+\Jh
Fig. 4. Technique of buttressing coronary ostium suture line with Teflon felt. Discussion Surgical treatment for aneurysm of the ascending aorta with associated aortic insufficiency has undergone a gradual evolution over the past 20 years. Initial efforts were aimed at preventing rupture of the aneurysm by aneurysmorrhaphy, with the hope that reduction in aortic size would ameliorate the aortic regurgitation.1 Muller and associates13 subsequently reported successful aneurysmorrhaphy combined with a direct plastic repair of the aortic valve. However, Merendino's group16 reported disastrous results with this approach, especially in patients with the full Marfan syndrome. Bloodwell,17 Ferlic,18 and their co-workers reported experience with prosthetic aortic valve replacement and supracoronary prosthetic graft replacement of the aneurysm. Postoperative hemorrhage was a major complication in both of these series. Nasrallah and colleagues19 using a similar technique, reported a 20 percent operative mortality rate, the deaths primarily resulting from hemorrhage from suture lines. The supracoronary graft technique also permits aneurysm formation in the coronary sinuses, which remain affected by cystic medial necrosis. One of the patients in the present series had a previous supracoronary graft
and then a recurrent aneurysm developed below this graft (Fig. 1). Recent reports of aortic valve replacement and variations of supracoronary aneurysm grafting11' 12, 14 or aneurysmorrhaphy13 show slightly lower (9 to 12 percent) operative mortality rates. Because of the problems with these earlier techniques, several authors have adopted the technique of replacing the ascending aorta and aortic valve with a composite graft. The basic concept was put forward by Wheat and associates2 in 1964, although several groups have since added modifications.3-10 The major advantages are reduced intraoperative hemorrhage and operating time with a consequent lowering of operative mortality rates. 5 , 8 ' 1 0 ' 2 1 In our experience requirements for blood replacement were just over 4 units in elective cases, and the mortality rate for all cases was 6 percent. In addition, aneurysm formation in the aortic sinuses, which may occur with supracoronary graft techniques, is prevented by use of the composite technique. Among those utilizing composite grafts, different techniques of myocardial preservation have been employed. Coronary perfusion5' 7' 8 and topical myocardial hypothermia10 have been the most widely used. In the last 24 months we have employed a second infant
Volume 76 Number 6 December, 1978
heat exchanger to cool the coronary perfusion blood to 18 to 20° C. This technique allows local myocardial hypothermia without separate crystalloid solutions for coronary perfusion or pericardial irrigating systems. In addition, myocardial cooling may be more uniform and systemic rewarming can be accomplished while the heart is kept cold. The problem of restoring coronary flow has generally been handled by implanting the coronary ostia into the sides of the graft, 3-8 ' 10 although Zubiate and Kay20 utilized saphenous vein grafts. The present report contains two patients with pseudoaneurysms occurring at the site of implantation of the coronary ostia into the graft, both of which developed more than 1 year postoperatively. Several factors may contribute to the development of these pseudoaneurysms. It is likely that there is continual stress on the junction of the coronary ostia with the prosthetic graft owing to cardiac motion. Second, postoperative bleeding into the space between the aortic wall and the graft would tend to pull the coronary ostia (which remain attached to the aortic wall) away from the graft. Tissue factors may also play a role in pseudoaneurysm formation, especially in patients with the full Marfan syndrome or in those with acute dissections involving the coronary arteries. All of the coronary suture line dehiscences have occurred in the inferior position on the anastomosis, and in an attempt to prevent this complication we now advocate buttressing the coronary ostial suture line in this location with a Teflon felt buttress (Fig. 4). Using saphenous vein grafts to restore coronary perfusion may have the advantage of reduced tension on the suture line, but questions remain about the long-term fate of the vein grafts, especially for younger patients. Patients with the full Marfan syndrome appear more susceptible to late suture line problems, as two of the six Marfan patients in this report had such problems. One had pseudoaneurysms around both coronary ostia and the other has had dehiscences of mitral and aortic prostheses from the anulus. It is likely that these patients have a more severe connective tissue defect and are therefore more likely to have late complications. Finally, two patients in this series have had transient episodes of altered neurologic status suggesting embolic events. Our experience with the Lillehei-Kaster prosthesis has shown very low thromboembolism rates in the aortic position (two emboli in 3,645 patientmonths).21 The emboli in the patients who have a composite graft may result either from the valve prosthesis or from the graft, which occupies the entire ascending aorta and probably never is covered with endothelium. Our approach has been to add aspirin and dipyridamole
Replacement of aortic valve and ascending aorta
82 1
to the anticoagulant regimen in those patients with episodes suggesting emboli. The low operative mortality rates associated with the composite graft method make this an attractive technique for ascending aortic aneurysms associated with aortic valvular insufficiency. The late results in this report suggest that significant problems may develop late, even in asymptomatic patients. Therefore, these patients must be followed closely, and we continue to perform routine postoperative supravalvular aortography and coronary angiography in all patients 6 to 12 months postoperatively. The appearance of new cardiac symptoms or changes in the cardiac silhouette on chest roentgenogram indicates the need for repeat catheterization. REFERENCES 1 Bahnson HT, Nelson AR: Cystic medial necrosis as a cause of localized aortic aneurysms amenable to surgical treatment. Ann Surg 144:519-529, 1956 2 Wheat MW, Wilson JR, Bartley TD: Successful replacement of the entire ascending aorta and aortic valve. JAMA 188:717-719, 1964 3 Bentall H, DeBono A: A technique for complete replacement of the ascending aorta. Thorax 3:338-339, 1968 4 Singh MP, Bentall HH: Complete replacement of the ascending aorta and the aortic valve for the treatment of aortic aneurysm. J THORAC CARDIOVASC SURG 63:218-
225, 1972 5 Helseth HK, Haglin JJ, Stenlund RR, Peterson CR, Gauger DW: Ascending aortic aneurysms with associated aortic regurgitation. Ann Thorac Surg 16:368-374, 1973 6 Helseth HK, Haglin JJ, Stenlund RR, Peterson CR: Evaluation of composite replacement of the aortic root and ascending aorta. Ann Thorac Surg 18:138-141, 1974 7 Edwards WS, Kerr AR: A safer technique for replacement of the entire ascending aorta and aortic valve. J THORAC CARDIOVASC SURG 59:837-839, 1970
8 Crosby IK, Ashcraft WC, Reed WA: Surgery of proximal aorta in Marfan's syndrome. J THORAC CARDIOVASC SURG 66:75-81, 1973
9 Wheat MW, Boruchow IB, Ramsey HW: Surgical treatment of aneurysms of the aortic root. Ann Thorac Surg 12:593-607, 1971 10 Kouchoukos NT, Karp RB, Lell WA: Replacement of the ascending aorta and aortic valve with a composite graft. Results in 25 patients. Ann Thorac Surg 24:140-148, 1977 11 Liddicoat JE, Bekassy SM, Rubio PA, Noon GP, De Bakey ME: Ascending aortic aneurysms. Circulation 52:Suppl 2:202-209, 1975 12 Najafi H, Dye WS, Javid H, Hunter JA, Goldin MD, Serry C: Aortic insufficiency secondary to aortic root aneurysm or dissection. Arch Surg 110:1401-1407, 1975 13 Groves LK: The surgical treatment of aneurysm of the
8 22
14
15 16
17
18 19
20
The Journal of Thoracic and Cardiovascular Surgery
Mayer et al.
ascending aorta with associated aortic valve insufficiency. Surg Clin North Am 55:1167-1174, 1975 Weldon CS, Ferguson TB, Ludbrook PA, McKnight RC: A new operation for far-advanced cystic medial necrosis of the aortic root. Ann Thorac Surg 23:499-506, 1977 Muller WH Jr, Dammann JF Jr, Warren WD: Surgical correction of cardiovascular deformities in Marfan's syndrome. Ann Surg 152:506-517, 1960 Merendino KA, Winterscheid LC, Dillard DH: Cystic medial necrosis with and without Marfan's syndrome. Surg Clin North Am 47:1403-1418, 1967 Bloodwell RB, Hallman GL, Cooley DA: Aneurysm of the ascending aorta with aortic valvular insufficiency. Arch Surg (Chicago) 92:588-599, 1966 Ferlic RM, Goott B, Edwards JE, Lillehei CW: Aortic valvular insufficiency associated with cystic medial necrosis. Ann Surg 165:1-9, 1967 Nasrallah AT, Cooley DA, Goussous Y, Hallman GL, Lufschanowski R, Leachman RD: Surgical experience in patients with Marfan's syndrome, ascending aortic aneurysm, and aortic regurgitation. Am J Cardiol 36:338-341, 1975 Zubiate P, Kay JH: Surgical treatment of aneurysm of the ascending aorta with aortic insufficiency and marked displacement of the coronary ostia. J THORAC CARDIOVASC SURG 71:415-421, 1976
21 Mayer JE, Pyle RB, Lindsay WG, Wang Y, Jorgensen CR, Nicoloff DM: Five year experience with LilleheiKaster prostheses in the aortic position. World J Surg (in press)
Discussion DR. H O V A L D K. H E L S E T H Minneapolis, Minn.
Dr. Mayer is to be complimented on a fine series of cases which historically have been difficult for the surgeon. We have enjoyed sharing experiences with Bentall's operation. At the Hennepin County Hospital, Dr. Haglin and I have used the composite graft in 32 patients with aortic root problems over the past 6V2 years. Save for one patient with giant cell aortitis, all had cystic medial necrosis. Twenty-three patients had an elective operation. Nine had acute ascending dissections which were operated upon as an emergency as soon as the diagnosis was made. Nine had true Marfan's syndrome. Four had associated mitral valve or coronary bypass. One patient with neglected aortic regurgitation died in the operating room. There was no instance of postoperative bleeding. Thirty-one patients are living without symptoms referable to the aortic root. No patient has been reoperated upon, and two thirds of the patients have had repeat aortography. Our conclusion is that Bentall's operation is probably the safest, shortest, most complete, and most durable operation for connective tissue problems in the aortic root. Our composite graft was generally prepared before or during bypass with a Dacron tube and Bj'ork valve. Annular sutures through
this rigid base sometimes created troublesome bleeding in the operating room; thus recently a manufactured composite has been designed with a soft outside sewing ring that allows the anulus to be buttressed to soft fabric. This provides a gentle cuff for attachment and, further, allows a somewhat subannular placement of the prosthesis itself. The coronary ostia are thus more adjacent to the tube. Attachment to the graft should be easier in those patients in whom the coronary ostia are not widely removed from the aortic anulus. This composite has been used in the last three patients and has allowed a bypass time of 75 minutes and a perfectly dry aortic root on heparin. Because the operation has been safe and because most patients with true Marfan's disease will die young from problems in the first 2 inches of the aorta, it seems correct to strongly encourage early evaluation and operation, with the hope of avoiding the problems of acute dissection. I would like to thank Dr. Bentall for his first report in 1968, in which he described this technique. DR. C L A R E N C E S. WELDON St. Louis, Mo.
This elegant series of patients hardly needs support, but I would like to show some evidence that this operation does in fact succeed where operations that are less extensive fail. [(Slide)] This patient had an aortic dissection from cystic medial necrosis that produced aortic insufficiency and an eventual aortic rupture into the pericardium. At operation, the dissection site was resected, the layers of the aorta sewn together, and then the aorta was re-anastomosed, so that the aortic was resuspended. The result was very good for about 24 hours, at which time there was a sudden recurrence of severe aortic insufficiency. This, of course, resulted from a recurrence of the dissection at the suture line when the stitches pulled out of the intimal layer of the aorta. Dr. Bentall's operation was then done with a complete resection of the aortic root. We used a composite graft of prosthetic aorta and a porcine valve. Because the dissection now extended into the coronary arteries, the proximal portions of both coronary arteries were resected and coronary continuity was reestablished by the use of saphenous vein segments connected between the graft and the proximal portion of both the right and the main left coronary arteries. This postoperative study [Slide] was done about 1V4 years after the operation and shows excellent filling of both coronary arteries through the saphenous vein segments. Another patient, operated upon in 1970 for a dissection from cystic medial necrosis, was initially treated by replacement of the aortic valve for aortic insufficiency and again by division of the aorta, suturing of the layers of the aorta, and anastomosis. He entered the hospital in 1978 with bizarre neurologic symptoms and personality changes. Study indicated that the extracranial cerebral circulation was almost completely occluded, because an enormous aneurysm had developed from a recurrence of the dissection at the suture line. This aneurysm was so large that we established a femoral-femoral bypass and lowered the body temperature to 18° C. so that we could shut off his circulation to enter the
Volume 76 Number 6 December, 1978
chest. Repair was done with a composite graft and this time with transplantation of the coronary arteries into the graft, as was originally described by Dr. Bentall. [Slide] This is a postoperative aortogram. We make our composite grafts with porcine prostheses. The indentation represents a suture line in the graft, which we made in order to get at the anastomosis between the left coronary artery and the graft, which was bleeding a little bit after the operation had been completed. Of the two techniques for coronary reconnection, using saphenous vein segments offers certain advantages. First, it allows all the diseased aortic tissue around the coronary ostia to be removed; this tissue is sometimes troublesome to sew. In addition, it puts the coronary graft anastomosis anteriorly where it can be managed easily. In closing, I would like to remind people that this is not the end of a very big problem. Mr. Yagub in England, in his extensive series of patients with Marfan's disease and cystic medial necrosis, has demonstrated quite clearly that although these patients survive repair of the ascending aortic disruption, they eventually have aneurysms in the transverse aortic arch, in the descending thoracic aorta, and indeed sometimes in the abdominal aorta. DR. HUGH H. B E N T A L L London, England
I was delighted to see the excellent results reported by Dr. Mayer using an alternative prosthesis to the one which we use. We used the Starr-Edwards prosthesis for almost all our patients. Since we have too few cases to show actuarially, I have chosen to show them historically, from the first case in 1965. This man worked until 7 years after the operation, when he died from a rupture of an aneurysm 3 cm. down the left coronary artery; the rupture appeared to be quite unrelated to the operative technique. We have now operated upon 23 patients. There were four hospital deaths, which have already been reported, and five late deaths. Another patient died from rupture of an abdominal aortic aneurysm. For the remaining patients, there is now a maximum follow-up time of 11 years and a mean follow-up time in the survivors of 6'/2 years. We continue to do this operation in the manner in which it was introduced, with one exception. Cardioplegia now offers an easier means of operating and has reduced operative time.
Replacement
of aortic valve and ascending aorta
823
In England, we separate Marfan's disease from medial necrosis and we differentiate Marfan's disease into overt and forme fruste types. Four of our patients had syphilis, one had yaws, and one had a mycotic aneurysm following bacterial infection in a rheumatic valve. I use a little trick which may help in the anastomosis of the coronary artery. I use a hot wire cautery for cutting the little hole for anastomosis to the coronary ostium. This melts the Dacron slightly and, therefore, seals the end of the fibers and gives a firm ring for suturing. DR. C H R I S T I A N E. C A B R O L Paris, France
We have treated 21 such patients with two deaths. Like Dr. Mayer, we used an aortic tubular graft with a valve sewn in it, but we have employed two technical modifications to the technique described by Dr. Mayer. The first one is done to obviate problems with anastomosing the coronary ostia to the graft. It consists in placing a small Dacron tube between each coronary ostium and the aortic graft in a horseshoe fashion in order to avoid tension on the coronary anastomosis. The second modification is sort of a trick. After having covered the graft with the remnants of the diseased aortic wall, if there is evidence of persistent bleeding between the aortic tissue and the graft (which is not uncommon in this operation), we create a small fistula between the aortic sac and the right atrial appendage, which are in juxtaposition. This small left-to-right shunt closes spontaneously within hours in all cases, relieving all tension and hematoma around the graft. D R . M A Y E R (Closing) I would like to thank all the discussers for their comments. Dr. Helseth has had a continuing influence on us in Minneapolis and we certainly appreciate his significant experience. The new composite graft which is being commercially made is a very attractive advance and we intend to utilize it. Several authors have commented on problems with the coronary arteries. We obviously have had our problems as well. We appreciate Dr. Cabrol's very interesting solution to the problems of bleeding between the graft and the aortic wall and think it may well offer some additional options. We also recognize Mr. Bentall's contribution in making everyone aware of his development of this type of procedure.
John E. Mayer, Jr., M.D. (by invitation), W. G. Lindsay, M.D. (by invitation), Y. Wang, M.D. (by invitation), C. R. Jorgensen, M.D. (by invitation), and D. M. Nicoloff, M . D . , Minneapolis, Minn.
V_^ystic medial necrosis primarily affects the ascending aorta and sinuses of Valsalva, the result being aneurysum formation and secondary aortic insufficiency. Surgical treatment for this condition has evolved through several stages since the initial report of Bahnson and Nelson 1 in 1956. In 1964 Wheat and Colleagues 2 reported replacement of the aortic valve and ascending aorta. In that procedure the coronary ostia were left attached to the rim of aortic wall, and a graft to replace the ascending aorta was sewn to the skirt of an aortic valve prosthesis and rim of aortic wall to encompass the coronary ostia. Bentall and DeBono 3 described the technique in which the ascending aorta and aortic valve are replaced with a composite unit of tubular qraft sewn to the aortic anulus with anastomosis of the coronary ostia to the sides of the graft. Since those reports, others have reported generally favorable experience with this technique or slight modifica-
From the Departments of Surgery and Medicine, University of Minnesota Hospitals, Minneapolis, Minn. 55455. Read at the Fifty-eighth Annual Meeting of The American Association for Thoracic Surgery, New Orleans, La., May 8, 9, and 10, 1978. Address for reprints: D. M. Nicoloff, M.D., Department of Surgery, University of Minnesota Hospitals, Minneapolis, Minn. 55455. 816
tions. 4 - 1 0 However, disagreement exists as to whether the composite graft technique represents optimal treatment for aortic root aneurysms with aortic insufficiency,11-14 and there is insufficient long-term follow-up information. The present report reviews a 4Vi year experience with composite grafts with data on postoperative catheterization and long-term follow-up. Patients Sixteen patients ranging in age from 13 to 61 years underwent replacement of the aortic valve and ascending aorta with a composite graft from July, 1973, through April, 1978. All had aneurysms of the ascending aorta and all but two had associated massive aortic insufficiency. One patient had an aneurysm involving the anulus and aortic root without aortic insufficiency, and a second patient had calcific aortic stenosis (94 mm. Hg gradient). Six patients had clinical features of the Marfan syndrome, one of whom had previously had a supracoronary aortic graft inserted for a dissection. Histologic proof of cystic medial necrosis was present in the seven cases in which a biopsy of the aortic wall was submitted. In all cases the aneurysm was typical of that associated with cystic medial necrosis. Acute or semi-acute (less than 2 weeks) dissections were present in five patients. Only one patient was asymptomatic
0022-5223/78/120816+08S00.80/0 © 1978 The C. V. Mosby Co.
Volume 76 Number 6 December, 1978
Replacement of aortic valve and ascending aorta
"\Jfr
^J&
817
+\Jh-
Decompress LV . ,, LC artery A \ y / / - y perfusion
iC artery perfusion
^Mt
Fig. 1. Technique of composite replacement (see text). LC, Left coronary, RC, Right coronary, LV, Left ventricle. prior to operation, with five in New York Heart Association Class II and eight in Class III. The two patients with acute dissections were placed in Class IV. Surgical technique (Fig. 1) In all cases cardiopulmonary bypass with moderate systemic hypothermia (30° C.) and arterial return through a femoral artery cannula was utilized. Coronary perfusion was used in all cases (Fig. 1, c). Since February, 1976, (Six patients), a second infant heat exchanger has been added to the coronary perfusion
lines to lower the temperature of the coronary blood to 18 to 20° C. The technique of composite replacement is similar to that described by others3' 5- 7 but differs in that a Lillehei-Kaster pivoting disc prosthesis has been exclusively utilized in this series (Fig. 2). Valve sizes are listed in Table I. The prosthesis is sewn into the proximal end of a tightly woven Dacron graft,* and the graft-valve composite unit is then sewn to the aortic anulus with 2-0 Mersilene mattress sutures reinforced *USCI, Division of C. R. Bard, Inc., Billerica, Mass.
The Journal of Thoracic and Cardiovascular Surgery
8 1 8 Mayer et al.
Table I. Valve sizes (Lillehei-Kaster) Size
No.
18A 20A 22A 25A
2 2 4 8
Table II. Associated procedures Procedure
No.
Coronary artery bypass Mitral valve replacement Arch aneurysm graft
1 1 1
was utilized to keep systolic blood pressure below 120 mm. Hg. All patients were initially treated with Coumadin with recommendations to maintain prothrombin times at two times control. Fig. 2. Lillehei-Kaster aortic prosthesis sewn into tightly woven Dacron graft. with Vi6 inch Teflon felt pledgets (Fig. 1, c). The coronary ostia are sewn to side holes in the graft with continuous 5-0 polypropylene (Prolene) suture (Fig. 1, d). After the completion of the coronary anastomoses, a slit is made in the anterior wall of the graft through which coronary perfusion cannulas are inserted to provide coronary perfusion during the completion of the distal aortic anastomoses. In all but one case the distal anastomosis was to the aorta proximal to the innominate artery. When the distal anastomosis is completed, the coronary cannulas are removed and the slit in the graft is repaired (Fig. 1, e). Air is evacuated from the graft and left ventricle by inserting an air bleed button into the graft, and the aortic cross-clamp is removed. The wall of the aneurysm is trimmed and sewn over the graft to aid hemostasis (Fig. \,f). Three patients underwent additional procedures at the time when the composite graft was inserted (Table II). In one patient the mitral valve was replaced because of severe mitral insufficiency owing to prolapsing leaflets. A second patient had saphenous vein grafts to the anterior descending and right coronary arteries for atherosclerotic lesions. In the third patient the graft was extended to a point distal to the left subclavian take-off because the aneurysm involved the aortic arch. The three arch vessels were anastomosed to the graft as a unit by preserving the aortic wall between their ostia and then suturing the island of aortic wall to the graft. In the immediate postoperative period, nitroprusside
Postoperative follow-up Thirteen of the 16 patients returned for elective postoperative cardiac catheterization. In seven patients a catheter was passed across the atrial septum and into the left ventricle and a second catheter was placed in the ascending aorta so that peak systolic transvalvular gradients could be measured. Supra valvular aortography and coronary arteriography were performed in all patients. Further follow-up has been obtained in our clinics or by telephone contact with the patient or the referring physician. Average length of follow-up is 26 months (range 2 to 51 months). Results Early (Table III). Fifteen of the 16 patients survived the operative procedure and the early postoperative period. The single early death occurred in a patient operated upon for an acute dissection who had renal failure postoperatively. During hemodialysis, massive bleeding occurred around the right coronary orifice, and the patient died. A second patient had transient renal failure and an empyema after a transfusion reaction but survived. A severe ventricular arrhythmia controlled only by Bretylium occurred intraoperatively in one patient who survived, and one patient with an acute dissection had an intraoperative myocardial infarction. The remainder of the patients had uncomplicated courses. No patient was re-explored for postoperative bleeding. The average cardiopulmonary bypass time for elective cases without additional procedures was 115 minutes. The average bypass time including pa-
Volume 76 Number 6 December, 1978
819
Replacement of aortic valve and ascending aorta
Fig. 3. Contrast outlines left coronary artery and small pseudoaneurysm arising from the inferior margin of the suture line. Table m . Early complications (four patients) Complication Operative death Hemorrhage Renal failure Empyeme Myocardial infarction Severe arrhythmia
No. 1(%) 1 2 1 1 1
tients with acute dissections and additional procedures was 161 minutes. Late (Table IV). Of the 15 survivors, two have died in the late postoperative period. Eleven of the survivors are in N.Y.H.A. Class I, and two are in Class II. One late death occurred in a 13-year-boy whose preoperative left ventricular end-diastolic pressure was 38 mm. Hg. He died 5 months postoperatively of end-stage cardiac failure. Repeat catheterization 4 weeks prior to death had shown a 4 mm. Hg peak gradient across the prosthesis, patent coronary anastomoses, and no aortic regurgitation. A second patient died 12 months postoperatively. Routine angiography had shown a small pseudoaneurysm at the distal suture line and patency of the two saphenous vein grafts. Coumadin was discontinued because of the pseudoaneurysm, but the patient remained asymptomatic until she died suddenly at home. No autopsy was performed. Two patients have had pseudoaneurysms form at the anastomoses between the graft and the coronary arteries. One patient with the Marfan syndrome has aneurysms at both ostia which
Table IV. Late complications (seven patients) Complication Death Coronary ostia pseudoaneurysm Aortic paravalvular leak Distal suture-line pseudoaneurysm Transient neurologic changes (? emboli) Spinal cord hemorrhage
No.
2 2 1 1 2 1
were discovered 14 months postoperatively on a routine catheterization (Fig. 3). She is scheduled for reoperation in the near future. The second patient began having chest pain 3 ^ years postoperatively, and a left coronary pseudoaneurysm was found by angiography. A previous angiogram 7 months postoperatively was normal. He underwent reoperation and successful repair of the small suture-line dehiscence. One additional patient with the full Marfan syndrome was found to have a mitral paravalvular leak 10 months postoperatively; the leak was repaired by resuture of the dehisced area of the suture line. An aortogram was normal at this time. This same patient had an aortic paraprosthetic dehiscence 20 months after the initial operation, and this has also been repaired. Two additional patients have had transient episodes of neurologic symptoms which may represent emboli. They have been treated by adding aspirin and dipyridamole to their warfarin regimen. A final patient had a subdural hemorrhage around the cervical spinal cord despite a prothrombin time of 19 seconds (control 11 seconds).
8 20
The Journal of Thoracic and Cardiovascular Surgery
Mayer et al.
+\Jh
Fig. 4. Technique of buttressing coronary ostium suture line with Teflon felt. Discussion Surgical treatment for aneurysm of the ascending aorta with associated aortic insufficiency has undergone a gradual evolution over the past 20 years. Initial efforts were aimed at preventing rupture of the aneurysm by aneurysmorrhaphy, with the hope that reduction in aortic size would ameliorate the aortic regurgitation.1 Muller and associates13 subsequently reported successful aneurysmorrhaphy combined with a direct plastic repair of the aortic valve. However, Merendino's group16 reported disastrous results with this approach, especially in patients with the full Marfan syndrome. Bloodwell,17 Ferlic,18 and their co-workers reported experience with prosthetic aortic valve replacement and supracoronary prosthetic graft replacement of the aneurysm. Postoperative hemorrhage was a major complication in both of these series. Nasrallah and colleagues19 using a similar technique, reported a 20 percent operative mortality rate, the deaths primarily resulting from hemorrhage from suture lines. The supracoronary graft technique also permits aneurysm formation in the coronary sinuses, which remain affected by cystic medial necrosis. One of the patients in the present series had a previous supracoronary graft
and then a recurrent aneurysm developed below this graft (Fig. 1). Recent reports of aortic valve replacement and variations of supracoronary aneurysm grafting11' 12, 14 or aneurysmorrhaphy13 show slightly lower (9 to 12 percent) operative mortality rates. Because of the problems with these earlier techniques, several authors have adopted the technique of replacing the ascending aorta and aortic valve with a composite graft. The basic concept was put forward by Wheat and associates2 in 1964, although several groups have since added modifications.3-10 The major advantages are reduced intraoperative hemorrhage and operating time with a consequent lowering of operative mortality rates. 5 , 8 ' 1 0 ' 2 1 In our experience requirements for blood replacement were just over 4 units in elective cases, and the mortality rate for all cases was 6 percent. In addition, aneurysm formation in the aortic sinuses, which may occur with supracoronary graft techniques, is prevented by use of the composite technique. Among those utilizing composite grafts, different techniques of myocardial preservation have been employed. Coronary perfusion5' 7' 8 and topical myocardial hypothermia10 have been the most widely used. In the last 24 months we have employed a second infant
Volume 76 Number 6 December, 1978
heat exchanger to cool the coronary perfusion blood to 18 to 20° C. This technique allows local myocardial hypothermia without separate crystalloid solutions for coronary perfusion or pericardial irrigating systems. In addition, myocardial cooling may be more uniform and systemic rewarming can be accomplished while the heart is kept cold. The problem of restoring coronary flow has generally been handled by implanting the coronary ostia into the sides of the graft, 3-8 ' 10 although Zubiate and Kay20 utilized saphenous vein grafts. The present report contains two patients with pseudoaneurysms occurring at the site of implantation of the coronary ostia into the graft, both of which developed more than 1 year postoperatively. Several factors may contribute to the development of these pseudoaneurysms. It is likely that there is continual stress on the junction of the coronary ostia with the prosthetic graft owing to cardiac motion. Second, postoperative bleeding into the space between the aortic wall and the graft would tend to pull the coronary ostia (which remain attached to the aortic wall) away from the graft. Tissue factors may also play a role in pseudoaneurysm formation, especially in patients with the full Marfan syndrome or in those with acute dissections involving the coronary arteries. All of the coronary suture line dehiscences have occurred in the inferior position on the anastomosis, and in an attempt to prevent this complication we now advocate buttressing the coronary ostial suture line in this location with a Teflon felt buttress (Fig. 4). Using saphenous vein grafts to restore coronary perfusion may have the advantage of reduced tension on the suture line, but questions remain about the long-term fate of the vein grafts, especially for younger patients. Patients with the full Marfan syndrome appear more susceptible to late suture line problems, as two of the six Marfan patients in this report had such problems. One had pseudoaneurysms around both coronary ostia and the other has had dehiscences of mitral and aortic prostheses from the anulus. It is likely that these patients have a more severe connective tissue defect and are therefore more likely to have late complications. Finally, two patients in this series have had transient episodes of altered neurologic status suggesting embolic events. Our experience with the Lillehei-Kaster prosthesis has shown very low thromboembolism rates in the aortic position (two emboli in 3,645 patientmonths).21 The emboli in the patients who have a composite graft may result either from the valve prosthesis or from the graft, which occupies the entire ascending aorta and probably never is covered with endothelium. Our approach has been to add aspirin and dipyridamole
Replacement of aortic valve and ascending aorta
82 1
to the anticoagulant regimen in those patients with episodes suggesting emboli. The low operative mortality rates associated with the composite graft method make this an attractive technique for ascending aortic aneurysms associated with aortic valvular insufficiency. The late results in this report suggest that significant problems may develop late, even in asymptomatic patients. Therefore, these patients must be followed closely, and we continue to perform routine postoperative supravalvular aortography and coronary angiography in all patients 6 to 12 months postoperatively. The appearance of new cardiac symptoms or changes in the cardiac silhouette on chest roentgenogram indicates the need for repeat catheterization. REFERENCES 1 Bahnson HT, Nelson AR: Cystic medial necrosis as a cause of localized aortic aneurysms amenable to surgical treatment. Ann Surg 144:519-529, 1956 2 Wheat MW, Wilson JR, Bartley TD: Successful replacement of the entire ascending aorta and aortic valve. JAMA 188:717-719, 1964 3 Bentall H, DeBono A: A technique for complete replacement of the ascending aorta. Thorax 3:338-339, 1968 4 Singh MP, Bentall HH: Complete replacement of the ascending aorta and the aortic valve for the treatment of aortic aneurysm. J THORAC CARDIOVASC SURG 63:218-
225, 1972 5 Helseth HK, Haglin JJ, Stenlund RR, Peterson CR, Gauger DW: Ascending aortic aneurysms with associated aortic regurgitation. Ann Thorac Surg 16:368-374, 1973 6 Helseth HK, Haglin JJ, Stenlund RR, Peterson CR: Evaluation of composite replacement of the aortic root and ascending aorta. Ann Thorac Surg 18:138-141, 1974 7 Edwards WS, Kerr AR: A safer technique for replacement of the entire ascending aorta and aortic valve. J THORAC CARDIOVASC SURG 59:837-839, 1970
8 Crosby IK, Ashcraft WC, Reed WA: Surgery of proximal aorta in Marfan's syndrome. J THORAC CARDIOVASC SURG 66:75-81, 1973
9 Wheat MW, Boruchow IB, Ramsey HW: Surgical treatment of aneurysms of the aortic root. Ann Thorac Surg 12:593-607, 1971 10 Kouchoukos NT, Karp RB, Lell WA: Replacement of the ascending aorta and aortic valve with a composite graft. Results in 25 patients. Ann Thorac Surg 24:140-148, 1977 11 Liddicoat JE, Bekassy SM, Rubio PA, Noon GP, De Bakey ME: Ascending aortic aneurysms. Circulation 52:Suppl 2:202-209, 1975 12 Najafi H, Dye WS, Javid H, Hunter JA, Goldin MD, Serry C: Aortic insufficiency secondary to aortic root aneurysm or dissection. Arch Surg 110:1401-1407, 1975 13 Groves LK: The surgical treatment of aneurysm of the
8 22
14
15 16
17
18 19
20
The Journal of Thoracic and Cardiovascular Surgery
Mayer et al.
ascending aorta with associated aortic valve insufficiency. Surg Clin North Am 55:1167-1174, 1975 Weldon CS, Ferguson TB, Ludbrook PA, McKnight RC: A new operation for far-advanced cystic medial necrosis of the aortic root. Ann Thorac Surg 23:499-506, 1977 Muller WH Jr, Dammann JF Jr, Warren WD: Surgical correction of cardiovascular deformities in Marfan's syndrome. Ann Surg 152:506-517, 1960 Merendino KA, Winterscheid LC, Dillard DH: Cystic medial necrosis with and without Marfan's syndrome. Surg Clin North Am 47:1403-1418, 1967 Bloodwell RB, Hallman GL, Cooley DA: Aneurysm of the ascending aorta with aortic valvular insufficiency. Arch Surg (Chicago) 92:588-599, 1966 Ferlic RM, Goott B, Edwards JE, Lillehei CW: Aortic valvular insufficiency associated with cystic medial necrosis. Ann Surg 165:1-9, 1967 Nasrallah AT, Cooley DA, Goussous Y, Hallman GL, Lufschanowski R, Leachman RD: Surgical experience in patients with Marfan's syndrome, ascending aortic aneurysm, and aortic regurgitation. Am J Cardiol 36:338-341, 1975 Zubiate P, Kay JH: Surgical treatment of aneurysm of the ascending aorta with aortic insufficiency and marked displacement of the coronary ostia. J THORAC CARDIOVASC SURG 71:415-421, 1976
21 Mayer JE, Pyle RB, Lindsay WG, Wang Y, Jorgensen CR, Nicoloff DM: Five year experience with LilleheiKaster prostheses in the aortic position. World J Surg (in press)
Discussion DR. H O V A L D K. H E L S E T H Minneapolis, Minn.
Dr. Mayer is to be complimented on a fine series of cases which historically have been difficult for the surgeon. We have enjoyed sharing experiences with Bentall's operation. At the Hennepin County Hospital, Dr. Haglin and I have used the composite graft in 32 patients with aortic root problems over the past 6V2 years. Save for one patient with giant cell aortitis, all had cystic medial necrosis. Twenty-three patients had an elective operation. Nine had acute ascending dissections which were operated upon as an emergency as soon as the diagnosis was made. Nine had true Marfan's syndrome. Four had associated mitral valve or coronary bypass. One patient with neglected aortic regurgitation died in the operating room. There was no instance of postoperative bleeding. Thirty-one patients are living without symptoms referable to the aortic root. No patient has been reoperated upon, and two thirds of the patients have had repeat aortography. Our conclusion is that Bentall's operation is probably the safest, shortest, most complete, and most durable operation for connective tissue problems in the aortic root. Our composite graft was generally prepared before or during bypass with a Dacron tube and Bj'ork valve. Annular sutures through
this rigid base sometimes created troublesome bleeding in the operating room; thus recently a manufactured composite has been designed with a soft outside sewing ring that allows the anulus to be buttressed to soft fabric. This provides a gentle cuff for attachment and, further, allows a somewhat subannular placement of the prosthesis itself. The coronary ostia are thus more adjacent to the tube. Attachment to the graft should be easier in those patients in whom the coronary ostia are not widely removed from the aortic anulus. This composite has been used in the last three patients and has allowed a bypass time of 75 minutes and a perfectly dry aortic root on heparin. Because the operation has been safe and because most patients with true Marfan's disease will die young from problems in the first 2 inches of the aorta, it seems correct to strongly encourage early evaluation and operation, with the hope of avoiding the problems of acute dissection. I would like to thank Dr. Bentall for his first report in 1968, in which he described this technique. DR. C L A R E N C E S. WELDON St. Louis, Mo.
This elegant series of patients hardly needs support, but I would like to show some evidence that this operation does in fact succeed where operations that are less extensive fail. [(Slide)] This patient had an aortic dissection from cystic medial necrosis that produced aortic insufficiency and an eventual aortic rupture into the pericardium. At operation, the dissection site was resected, the layers of the aorta sewn together, and then the aorta was re-anastomosed, so that the aortic was resuspended. The result was very good for about 24 hours, at which time there was a sudden recurrence of severe aortic insufficiency. This, of course, resulted from a recurrence of the dissection at the suture line when the stitches pulled out of the intimal layer of the aorta. Dr. Bentall's operation was then done with a complete resection of the aortic root. We used a composite graft of prosthetic aorta and a porcine valve. Because the dissection now extended into the coronary arteries, the proximal portions of both coronary arteries were resected and coronary continuity was reestablished by the use of saphenous vein segments connected between the graft and the proximal portion of both the right and the main left coronary arteries. This postoperative study [Slide] was done about 1V4 years after the operation and shows excellent filling of both coronary arteries through the saphenous vein segments. Another patient, operated upon in 1970 for a dissection from cystic medial necrosis, was initially treated by replacement of the aortic valve for aortic insufficiency and again by division of the aorta, suturing of the layers of the aorta, and anastomosis. He entered the hospital in 1978 with bizarre neurologic symptoms and personality changes. Study indicated that the extracranial cerebral circulation was almost completely occluded, because an enormous aneurysm had developed from a recurrence of the dissection at the suture line. This aneurysm was so large that we established a femoral-femoral bypass and lowered the body temperature to 18° C. so that we could shut off his circulation to enter the
Volume 76 Number 6 December, 1978
chest. Repair was done with a composite graft and this time with transplantation of the coronary arteries into the graft, as was originally described by Dr. Bentall. [Slide] This is a postoperative aortogram. We make our composite grafts with porcine prostheses. The indentation represents a suture line in the graft, which we made in order to get at the anastomosis between the left coronary artery and the graft, which was bleeding a little bit after the operation had been completed. Of the two techniques for coronary reconnection, using saphenous vein segments offers certain advantages. First, it allows all the diseased aortic tissue around the coronary ostia to be removed; this tissue is sometimes troublesome to sew. In addition, it puts the coronary graft anastomosis anteriorly where it can be managed easily. In closing, I would like to remind people that this is not the end of a very big problem. Mr. Yagub in England, in his extensive series of patients with Marfan's disease and cystic medial necrosis, has demonstrated quite clearly that although these patients survive repair of the ascending aortic disruption, they eventually have aneurysms in the transverse aortic arch, in the descending thoracic aorta, and indeed sometimes in the abdominal aorta. DR. HUGH H. B E N T A L L London, England
I was delighted to see the excellent results reported by Dr. Mayer using an alternative prosthesis to the one which we use. We used the Starr-Edwards prosthesis for almost all our patients. Since we have too few cases to show actuarially, I have chosen to show them historically, from the first case in 1965. This man worked until 7 years after the operation, when he died from a rupture of an aneurysm 3 cm. down the left coronary artery; the rupture appeared to be quite unrelated to the operative technique. We have now operated upon 23 patients. There were four hospital deaths, which have already been reported, and five late deaths. Another patient died from rupture of an abdominal aortic aneurysm. For the remaining patients, there is now a maximum follow-up time of 11 years and a mean follow-up time in the survivors of 6'/2 years. We continue to do this operation in the manner in which it was introduced, with one exception. Cardioplegia now offers an easier means of operating and has reduced operative time.
Replacement
of aortic valve and ascending aorta
823
In England, we separate Marfan's disease from medial necrosis and we differentiate Marfan's disease into overt and forme fruste types. Four of our patients had syphilis, one had yaws, and one had a mycotic aneurysm following bacterial infection in a rheumatic valve. I use a little trick which may help in the anastomosis of the coronary artery. I use a hot wire cautery for cutting the little hole for anastomosis to the coronary ostium. This melts the Dacron slightly and, therefore, seals the end of the fibers and gives a firm ring for suturing. DR. C H R I S T I A N E. C A B R O L Paris, France
We have treated 21 such patients with two deaths. Like Dr. Mayer, we used an aortic tubular graft with a valve sewn in it, but we have employed two technical modifications to the technique described by Dr. Mayer. The first one is done to obviate problems with anastomosing the coronary ostia to the graft. It consists in placing a small Dacron tube between each coronary ostium and the aortic graft in a horseshoe fashion in order to avoid tension on the coronary anastomosis. The second modification is sort of a trick. After having covered the graft with the remnants of the diseased aortic wall, if there is evidence of persistent bleeding between the aortic tissue and the graft (which is not uncommon in this operation), we create a small fistula between the aortic sac and the right atrial appendage, which are in juxtaposition. This small left-to-right shunt closes spontaneously within hours in all cases, relieving all tension and hematoma around the graft. D R . M A Y E R (Closing) I would like to thank all the discussers for their comments. Dr. Helseth has had a continuing influence on us in Minneapolis and we certainly appreciate his significant experience. The new composite graft which is being commercially made is a very attractive advance and we intend to utilize it. Several authors have commented on problems with the coronary arteries. We obviously have had our problems as well. We appreciate Dr. Cabrol's very interesting solution to the problems of bleeding between the graft and the aortic wall and think it may well offer some additional options. We also recognize Mr. Bentall's contribution in making everyone aware of his development of this type of procedure.