- Email: [email protected]
Postperfusion coronary ostial stenosis
Postperfusion coronary ostial stenosis Incidence and significance In a study of 117 consecutive aortic valve replacements in which selective coronary perfusion was routinely employed, four patients developed coronary ostial stenosis (3.5 per cent). Continuous selective coronary perfusion was performed by use of Mayo balloon catheters with individual measuring of pressure and flow. All 4 patients developed progressive symptoms of angina pectoris within 6 months of the original operation, after uneventful recoveries. All 4 were found to have lesions in the left main coronary ostium and required a coronary bypass. Two made uneventful recoveries and are presently asymptomatic, whereas 2 died in the immediate postoperative period. The literature indicates that the reported incidence of this complication varies from I to 5 per cent. Furthermore, the mortality rate for reoperation in these patients is higher than that for those undergoing uncomplicated coronary bypass without an antecedent procedure. Our experience confirms the lethal nature of this complication and the necessity for reoperation once the diagnosis is established. The development of these dangerous lesions must be taken into account in the prevailing controversy between the most effective methods of myocardial protection during aortic valve replacement.
Allen I. Midell, M.D.,* Arthur DeBoer, M.D.,* and Gustave Bermudez, M.D.,** Chicago, 111.
V-^urrent efforts to reduce the early mortality rate of aortic valve replacement center around selection of the preferable method of myocardial protection to use during operation. The two methods under the most intensive study are selective continuous coronary perfusion as compared to ischemic arrest with hypothermia (systemic and topical). 1-12 Central to the argument among surgeons who prefer ischemic arrest with hypothermia rests the documented occurrence of injuries to the coronary arteries during cannulation and perfusion.3' 13~18 Among the most important of these injuries is the development of coronary ostial stenosis. Although this syndrome is characterized by a remarkably consistent clinical pattern, its cause and frequency are uncertain. Establishing the true incidence of this lesion is essential to determining its significance. This From the Department of Surgery, University of Chicago, Pritzker School of Medicine, and the Columbus Hospital Medical Center, Chicago, 111. Received for publication Feb. 18, 1976. Accepted for publication May 4, 1976. *Senior Attending Cardiac Surgeon, Columbus Hospital Medical Center. **Director of Division of Cardiology, Columbus Hospital Medical Center.
80
is especially so in the light of recent favorable results being reported using ischemic arrest and hypothermia.1, 4,6,8-n Accordingly, having routinely used continuous coronary perfusion, we have surveyed our recent experience with aortic valve replacement to determine the incidence of this complication and have reviewed the experience of others who have reported on this complication. Methods This study consists of 117 consecutive aortic valve replacements performed from September, 1971, to July, 1974. The Starr-Edwards aortic valve prosthesis was used in 114 patients. The Bjork-Shiley prosthesis was used in 3 patients. Eighty-four patients underwent isolated aortic valve replacement and 33 underwent multiple valve replacement. To minimize variables in this study, we have excluded patients who underwent concomitant revascularization procedures, such as coronary bypass or ventricular aneurysmectomy along with aortic valve replacement. All operations were performed by means of moderate systemic hypotermia (32° C.) with a hemodiluted prime of Ringer's lactate solution. The oxygenators used were either rotating disc oxygenators or the standard bubble
Volume 72 Number 1 July, 1976
Postperfusion
coronary ostial stenosis
81
Fig. 1. Left anterior oblique view demonstrating severe stenosis of left main coronary artery in Case 1. oxygenators (Travenol, Galen, and Bentley). Flow rates varied during individual perfusion but, in general, were high enough to ensure a perfusion pressure of at least 50 mm. Hg and to maintain adequate urine output during the perfusion. Mayo balloon catheters with individual monitoring of pressure and flow were used in all cases to perfuse the coronary arteries. Each coronary ostium was cannulated and perfused continuously with blood at body temperature during the operation. Cannulas were inserted just prior to removal of the valve after the aorta was opened, and they were removed just before the aortotomy closure was completed. On occasion cannulation of the right coronary artery was not attempted or accomplished due to a rudimentary ostium. Most cannulas lay in the coronary arteries without fixation, but occasionally it was necessary to place a suture around the ostium to secure the cannula. The right coronary cannula was usually inserted several inches into the right coronary artery, but the left coronary cannula was inserted just past the ostium. Great care was taken to make certain that the cannula did not block off one of the main branches of the left coronary artery because of the uncertain length of the left main coronary artery. 10 The pressure in each cannula was maintained so as not to exceed 150 mm. Hg, if possible, and the flow generally varied from 75 to 200 c.c. per minute in each coronary artery. Results Of the 117 patients, 84 underwent isolated aortic valve replacement and 80 survived the immediate
postoperative period. Of the 33 who had multiple valve replacement, 23 survived postoperatively. Of the 103 survivors, 4 developed coronary ostial stenosis, an incidence of 3.8 per cent of the operative survivors and 3.5 per cent of the total number operated upon. Three of these 4 patients underwent isolated aortic valve replacement and one underwent a triple valve replacement. All the lesions were in the ostium of the left coronary artery, although the right coronary artery was perfused in 3 of the 4 cases. Two had aortic stenosis as the predominant aortic lesion. The remaining two had a mixed lesion of aortic stenosis and insuffiency. The average length of coronary perfusion in the entire series was 55 minutes. The longest was 2 hours, 10 minutes and the shortest, 35 minutes. The average length of coronary perfusion in the 4 patients developing the lesion was 65 minutes. A careful review of the operative record in all 4 patients revealed no particular problem related to the coronary perfusion or to the insertion of the cannulas. Case reports CASE 1. A 41-year-old man was admitted June 28, 1973. Cardiac catheterization revealed aortic stenosis and insufficiency, with a normal coronary arteriogram. The aortic valve was replaced on July 30, 1973, with a Starr-Edwards aortic valve prosthesis. Both coronary arteries were perfused uneventfully. The patient was readmitted on Nov. 14, 1973, with severe chest pain. Catheterization revealed a tight stenosis of the left main coronary artery (Fig. 1). A saphenous vein bypass to the left anterior descending was performed, with an uneventful recovery. The patient is presently asymptomatic. CASE 2. A 47-year-old white man was admitted to
82
Midell, DeBoer,
Bermudez
The Journal of Thoracic and Cardiovascular Surgery
Fig. 2. Right anterior oblique view demonstrating lesion of left main coronary artery in Case 3. Note that the strut of the aortic valve prosthesis almost obscures the lesion in this view.
i Fig. 3. Left anterior oblique view demonstrating left main coronary lesion in Case 4. Columbus Hospital on Aug. 10, 1973. Cardiac catheterization revealed severe aortic stenosis with a normal coronary arteriogram. On Aug. 27, 1973, an aortic valve replacement with a Starr-Edwards prosthesis was performed. Coronary perfusion of both arteries was unremarkable. After an uneventual recovery, the patient was readmitted on Dec. 10, 1973, with an obvious anginal syndrome. Catheterization revealed a normally functioning aortic valve prosthesis and severe narrowing of the ostium of the left main coronary artery (90 per cent). Saphenous vein bypass of the left anterior descending coronary artery was performed on Dec. 17, 1973, and resulted in fatal myocardial infarction on the day of the operation. CASE 3. A 49-year-old woman was admitted on May 5,
1974. Cardiac catheterization revealed aortic stenosis and insufficiency. The coronary arteriogram was normal. An aortic valve replacement was performed on May 13, 1974. Only the left coronary artery was perfused. After an uneventful recovery, she was readmitted on July 5, 1974, with severe angina, and a coronary arteriogram performed shortly thereafter showed 90 per cent narrowing of the left main coronary artery (Fig. 2). A coronary bypass to the left anterior descending was performed on July 6, 1974. After an uneventful recovery, the patient is presently asymptomatic. CASE 4. A 59-year-old woman was admitted on Dec. 19, 1974. Catheterization revealed aortic stenosis with mitral and tricuspid insufficiency. The coronary arteriogram was normal. A triple valve replacement was performed on Jan. 8,
Volume 72 Number 1 July, 1976
1975. The patient's recovery was complicated by severe congestive heart failure and pulmonary complications, but she eventually became well enough to be sent home after 8 weeks of hospitalization. She was readmitted on May 21, 1975, with severe recurrent chest pain. A coronary arteriogram revealed a severe ostial stenosis involving 90 per cent of the left main coronary artery (Fig. 3). A coronary artery bypass to the left anterior descending was performed, but the patient died on the fourteenth postoperative day from refractory congestive heart failure, pulmonary complications, and ventricular arrhythmia. Discussion The development of dangerous proximal lesions in the coronary arteries following cannulation, first described by Trimble and associates17 in 1959, has since been confirmed by Yates,18 Hazan,13 Lesage,14 and their co-authors, as well as by others. 15 ' 16 Patients in this review should be distinguished from those in whom malposition of the coronary cannulas results either in occlusion of one of the two branches of the left coronary artery or intimal dissection in a distal section. These developments usually result in early and massive infarction, which is entirely different from the lesion under discussion.10 Reports of coronary ostial stenosis are remarkably similar in their description of the onset and clinical pattern. In Trimble's17 original description, lesions were noted in both the right and left coronary ostia following the use of Mayo balloon catheters during aortic valve replacement. Pressures were not monitored in either cannula, and development of these lesions was attributed at the time to high perfusion pressures. One patient died at reoperation and the other 2 died without having another operation. In a subsequent communication Reed, Spencer, and colleagues16 reported an incidence of 5 per cent (6 patients) in 103 consecutive aortic valve replacements. Silicone rubber as well as self-inflating balloon catheters were used to perfuse the coronary arteries, but pressures were not monitored. The right coronary ostium was involved in 4 of the 6 patients. In 1973, Nakhjavan and associates15 reported on 3 patients in whom lesions developed in both coronary ostia following cannulation and perfusion. They did not mention what percentage of their total experience these lesions comprised, and they further speculated that the left coronary lesions might have been present prior to the original operation. More recently, Yates and associates18 reported on the largest number of patients in a single series developing coronary ostial stenosis, 8 out of 292 or 2.7 per cent. The interval between the original operation and the onset of angina varied up to 6 months. Left coronary
Postperfusion coronary ostial stenosis
83
artery stenosis developed in 6, both the right and left coronary arteries were involved in one, and the final patient had an isolated right coronary artery lesion. All required reoperation but only 5 of the 8 survived reoperation. Hazan, Rioux, and co-workers13 reported on 3 patients with left coronary artery stenosis following aortic valve replacement. Their incidence was 1.6 per cent of the total series. All required coronary artery bypass surgery and all survived. Lesage and colleagues14 have reported on 6 patients with iatrogenic coronary artery lesions following prosthetic replacement of the aortic valve but gave no indication of the incidence in their series. Our patients bore a striking resemblance to those in previously cited reports. AH 4 patients had selective coronary artery perfusion varying from 40 to 90 minutes, and a satisfactory perfusion was achieved in all 4 cases. In the immediate postoperative period no clinical or electrocardiographic evidence of ischemia or underlying injury to the coronary arteries was noted. However, within 20 weeks of the original operation, all 4 patients developed a severe, rapidly progressive anginal syndrome with electrocardiographic patterns of ischemia. In all 4 patients a saphenous vein bypass was required from the aorta to the left anterior descending coronary artery to achieve symptomatic relief. The remaining 99 survivors have not reported the development of angina since the operation, although 3 patients died suddenly after the first year of valve replacement. No autopsies were available. It is clear from a review of this subject that coronary ostial lesions may occur in both coronary ostia, although there seems to be a slight predisposition for the left coronary ostium. The lesions occur and develop in a relatively short period of time following the original operation. The initial complaint is invariably a progressive anginal syndrome accompanied by characteristic electrocardiographic changes. The cause of these lesions remains unproved, but the most acceptable theory is that they represent injuries to the ostium, either from insertion of the cannula or inflation of the balloon, resulting in an "intimal insult" with subsequent fibrosis and narrowing of the ostium.15' 23,24 Lesions are also observed in patients who are not perfused. This is thought to be due to repeated unsuccessful attempts to insert a cannula, resulting in injury to the ostium. Another explanation of this lesion is offered by Roberts and Morrow. 20-22 They postulate that turbulent flow around the newly placed prosthetic valve results in a hyperplastic reaction at the ostium. The fact that virtually no cases of coronary ostial
The Journal of
8 4 Midell, DeBoer, Bermudez
Thoracic and Cardiovascular Surgery
Table I. Comparative incidence of coronary ostial stenosis in previous reports Incidence Author
No. of cases
No.
%
Yates et al. 18 Hazan et al.13 Reed et al.16 Nakhjavan et al.15 DeBoer and Midell2 Lesage et al.14
292 185 103 7 117 ?
8 3 6 3 4 6
2 1.6 5 7 3 7
Deaths on reoperation No.
%
3 35 0 0 0 2 50 None reoperated
stenosis have been reported in the group of patients in whom ischemic arrest and hypothermia were used casts considerable doubt on this theory. The diagnosis of coronary ostial stenosis is determined by repeat coronary angiograms. This should be performed as soon as the lesion is suspected. Great care should be taken during angiographic study, as repeated injections can produce a dangerous arrhythmia. The left anterior oblique view is considered the most effective in demonstrating these lesions, as the prosthesis often obscures the ostium in the right anterior oblique view (Fig. 2). Clearly, the only effective treatment is early reoperation: Either a saphenous vein bypass or an anastomosis between the internal mammary artery and the left anterior descending coronary artery is used for lesions in the left main coronary artery; a saphenous vein bypass to the right coronary artery is employed for lesions in the right coronary ostium. We do not find it necessary to place another bypass to the circumflex coronary artery in those with left main coronary lesions, as bidirectional flow through the graft seems to provide adequate perfusion to the circumflex coronary artery if there is not a second lesion in the left anterior descending distal to the circumflex orifice. We favor the use of a saphenous vein over the internal mammary artery, as the vein bypass assures the high flows necessary to provide adequate perfusion to both the left anterior descending and the circumflex coronary artery. We conclude from our review that the incidence of coronary ostial stenosis is low, varying from 1 to 5 per cent (Table I). Robicsek and associates10 suggest it is probably higher than reported, however. This may well be true if we assume that certain "unexpected sudden deaths" following aortic valve replacement are due to this disease progressing to its logical terminus and that
these deaths have not heretofore been considered as being caused by coronary ostial stenosis or, indeed, to any injury to the coronary arteries. They are usually attributed to ventricular arrhythmia, noniatrogenic myocardial infarction, or pulmonary embolus. Certainly data from more carefully followed patients must be available before any final conclusion can be drawn regarding the true incidence of this complication. The total incidence of coronary ostial stenosis in their over-all experience is often omitted by the reporting authors (Table I). The significance of ostial stenosis nevertheless exceeds that of its lower reported incidence, because the mortality rate of reoperation is substantial. Two of the 4 patients in our series died during reoperation and 3 of the 8 patients reoperated by Yates did not survive. Although it is not the purpose of this discussion to compare the merits of the various methods of myocardial protection during aortic valve replacement, nevertheless our findings have relevance to this controversy. We have previously reported our experience with aortic valve replacement and have credited a low operative mortality rate to, among other factors, continuous coronary perfusion.2 We are presently continuing the use of coronary perfusion in contrast to ischemic arrest, but we have replaced the Mayo balloon catheters with O-ring cannulas in the hope that this will minimize, or even eliminate, the development of coronary ostial stenosis. At this time we do not have enough experience to provide meaningful data. If efforts to eliminate this complication fail, then we must ask whether the continued development of this problem, with its rapid evolution and necessity for reoperation, can be tolerated in any large surgical experience with aortic valve replacement. Certainly, the development of these lesions, even with the low incidence as described, must be considered in any analysis of the over-all results of aortic valve replacement performed with coronary perfusion, even though the patient may survive the initial period of operation. The use of ischemic arrest and hypothermia has been advocated by Bloodwell,1 Shumway,12 Robicsek,9 and others23, 8 for some years. Recent clinical and experimental evidence from several medical centers seems to substantiate the hypothesis that the myocardium can tolerate long periods of anoxic arrest under hypothermia without incurring irreversible damage. 4,5 ' 23, 24 i^g ii m jt s 0 f toleration of ischemia have not yet been tightly defined, however. In a recent editorial, Robicsek9 stated that ischemic arrest and hypothermia is appropriate only to those who are "quick operators." The time required to replace the aortic valve varies
Volume 72 Number 1 July, 1976
Postperfusion
substantially in each individual case. Replacing the aortic valve in a large, dilated, noncalcified annulus often requires much less time than replacing a densely calcified valve in a narrow annulus, as is often seen in calcific aortic stenosis. Therefore, the final determinant of the preferred method of myocardial protection will ultimately depend on the skill of the individual surgeon. It is unlikely that one method will " w i n " in that it will be used exclusively over the other. Each surgeon must consider whether in his hands the incidence of injury and infarction from prolonged ischemic arrest equals or exceeds the incidence of iatrogenic coronary artery injury from cannulation and perfusion. REFERENCES 1 Bloodwell, R. D., Kidd, M. J., Hallman, G. L., Burdette, W. J., McMurtrey, M. J., and Cooley, D. A.: Cardiac Valve Replacement Without Coronary Artery Perfusion: Clinical and Laboratory Observations, in Brewer, L. A., editor: Prosthetic Heart Valves, Springfield, 111., 1969, Charles C Thomas, Publisher. 2 DeBoer, A., and Midell, A. I.: Isolated Aortic Valve Replacement: Analysis of Factors Influencing Survival After Replacement With the Starr-Edwards Prosthesis Ann. Thorac. Surg. 17: 360, 1974. 3 Fishman, N. H., Youker, J. E., and Benson, B. R.: Mechanical Injury to the Coronary Arteries During Operative Cannulation, Am. Heart J. 75: 26, 1968. 4 Griepp, R. B., Stinson, E. B., and Shumway, N. E.: Profound Local Hypothermia for Myocardial Protection During Open-Heart Surgery, J. THORAC. CARDIOVASC. SURG. 66: 731, 1973.
5 Hurley, E. J., Lower, R. R., Dong, E., Pillsbury, R. C , and Shumway, N. E.: Clinical Experience With Local Hypothermia in Elective Cardiac Arrest, J. THORAC. CARDIOVASC. SURG. 47: 50,
1964.
6 Iyengar, S. R. K., Ramchaud, S., Charrette, E. J. P., Iyengar, C. K. S., and Lynn, R. B.: Anoxic Cardiac Arrest: An Experimental and Clinical Study of Its Effects, J. THORAC CARDIOVASC. SURG. 66: 722,
1973.
7 Levitsky, S., Sloane, R. E., Mullin, E. M., Mclntosh, C. L., and Morrow, A. G.: Normothermic Myocardial Anoxia: Effects on the Canine Heart With Left Ventricular Outflow Obstruction, Ann. Thorac. Surg. 11: 229, 1971. 8 Pupelli, D. F., Blank, R. H., Connar, R. G., Bessone, L. N., Sbar, S., and Stevenson, A.: Fifty-two Consecutive Aortic Valve Replacements Employing Local Deep Hypothermia, Ann. Thorac. Surg. 17: 360, 1974. 9 Robicsek, F.: Local Cardiac Hypothermia in Lieu of Coronary Perfusion (Editorial), Ann. Thorac. Surg. 19: 487, 1975.
coronary ostial stenosis
85
10 Robicsek, F., Tarn, W., Doughterty, H. K., and Mullen, D. C : Myocardial Protection During Open Heart Surgery, Ann. Thorac. Surg. 10: 341, 1970. 11 Sapsford, R. N., Blackstone, E. H., Kirklin, J. W., Karp, R. B., Kouchoukos, N. T., Pacifico, A. D., Roe, C. R., and Bradley, E. L.: Coronary Perfusion Versus Cold Ischemic Arrest During Aortic Valve Surgery, Circulation 49: 1190, 1974. 12 Shumway, N. E., Lower, R. R., and Staffer, R. C : Selective Hypothermia of the Heart in Anoxic Cardiac Arrest, Surg. Gynecol. Obstet. 109: 750, 1959. 13 Hazan, E., Rioux, C , Dequirot, A., and Mathey, J.: Postperfusion Stenosis of the Common Left Coronary Artery, J. THORAC. CARDIOVASC. SURG. 69: 703,
1975.
14 Lesage, C. H., Jr., Vogel, J. H., and Blount, S. G., Jr.: Iatrogenic Coronary Occlusive Disease in Patients With Prosthetic Heart Valves, Am. J. Cardiol. 26: 123, 1970. 15 Nakhjavan, F. K., Maranhao, V., and Goldberg, H.: Iatrogenic Stenosis of the Proximal Portion of the Coronary Arteries, Am. Heart J. 83: 318, 1972. 16 Reed, G. E., Spencer, F. C , Boyd, A. D., Engelman, R. M., and Glassman, E.: Late Complications of Intraoperative Coronary Artery Perfusion, Circulation 48: 80, 1973 (Suppl. III). 17 Trimble, A. S., Bigelow, W. G., Wigle, E. D., and Silver, M. D.: Coronary Ostial Stenosis, J. THORAC CARDIOVASC SURG. 57: 792,
1969.
18 Yates, J. D., Kirsh, M. M., Sodeman, T. M., Walton, J. S., Jr., and Bryner, M. F.: Coronary Ostial Stenosis: A Complication of Aortic Valve Replacement, Circulation 49: 530, 1974. 19 Kronzon, I., Deutsch, R., and Glassman, E.: Length of Left Main Coronary Artery: Its Relations to the Pattern of Coronary Arterial Distribution, Am. J. Cardiol. 34: 787, 1974. 20 Roberts, W. C , and Morrow, A. G.: Late Postoperative Pathologic Findings After Cardiac Valve Replacement, Circulation 35: 68, 1967 (Suppl. I). 21 Roberts, W. C., and Morrow, A. G.: Anatomic Studies of Hearts Containing Caged-Ball Prosthetic Valves, Johns Hopkins Med. J. 121: 271, 1967. 22 Roberts, W. C , and Morrow, A. G.: Causes of Early Postoperative Death Following Cardiac Valve Replacement: Clinicopathologic Correlations in 64 Patients Studied at Necropsy, J. THORAC CARDIOVASC SURG.
54: 792, 1969. 23 Angell, W. W., Rikkers, L., Dong, E., Jr., and Shumway, N. E.: Organ Viability With Hypothermia, J. THORAC CARDIOVASC SURG. 58: 619,
1969.
24 Buja, L. M., Ferrans, V. J., and Levitsky, S.: Occurrence of Intra-mitochondrial Glycogen in Cananine Myocardium After Prolonged Anoxic Cardiac Arrest, J. Mol. Cell. Cardiol. 4: 237, 1972.
Allen I. Midell, M.D.,* Arthur DeBoer, M.D.,* and Gustave Bermudez, M.D.,** Chicago, 111.
V-^urrent efforts to reduce the early mortality rate of aortic valve replacement center around selection of the preferable method of myocardial protection to use during operation. The two methods under the most intensive study are selective continuous coronary perfusion as compared to ischemic arrest with hypothermia (systemic and topical). 1-12 Central to the argument among surgeons who prefer ischemic arrest with hypothermia rests the documented occurrence of injuries to the coronary arteries during cannulation and perfusion.3' 13~18 Among the most important of these injuries is the development of coronary ostial stenosis. Although this syndrome is characterized by a remarkably consistent clinical pattern, its cause and frequency are uncertain. Establishing the true incidence of this lesion is essential to determining its significance. This From the Department of Surgery, University of Chicago, Pritzker School of Medicine, and the Columbus Hospital Medical Center, Chicago, 111. Received for publication Feb. 18, 1976. Accepted for publication May 4, 1976. *Senior Attending Cardiac Surgeon, Columbus Hospital Medical Center. **Director of Division of Cardiology, Columbus Hospital Medical Center.
80
is especially so in the light of recent favorable results being reported using ischemic arrest and hypothermia.1, 4,6,8-n Accordingly, having routinely used continuous coronary perfusion, we have surveyed our recent experience with aortic valve replacement to determine the incidence of this complication and have reviewed the experience of others who have reported on this complication. Methods This study consists of 117 consecutive aortic valve replacements performed from September, 1971, to July, 1974. The Starr-Edwards aortic valve prosthesis was used in 114 patients. The Bjork-Shiley prosthesis was used in 3 patients. Eighty-four patients underwent isolated aortic valve replacement and 33 underwent multiple valve replacement. To minimize variables in this study, we have excluded patients who underwent concomitant revascularization procedures, such as coronary bypass or ventricular aneurysmectomy along with aortic valve replacement. All operations were performed by means of moderate systemic hypotermia (32° C.) with a hemodiluted prime of Ringer's lactate solution. The oxygenators used were either rotating disc oxygenators or the standard bubble
Volume 72 Number 1 July, 1976
Postperfusion
coronary ostial stenosis
81
Fig. 1. Left anterior oblique view demonstrating severe stenosis of left main coronary artery in Case 1. oxygenators (Travenol, Galen, and Bentley). Flow rates varied during individual perfusion but, in general, were high enough to ensure a perfusion pressure of at least 50 mm. Hg and to maintain adequate urine output during the perfusion. Mayo balloon catheters with individual monitoring of pressure and flow were used in all cases to perfuse the coronary arteries. Each coronary ostium was cannulated and perfused continuously with blood at body temperature during the operation. Cannulas were inserted just prior to removal of the valve after the aorta was opened, and they were removed just before the aortotomy closure was completed. On occasion cannulation of the right coronary artery was not attempted or accomplished due to a rudimentary ostium. Most cannulas lay in the coronary arteries without fixation, but occasionally it was necessary to place a suture around the ostium to secure the cannula. The right coronary cannula was usually inserted several inches into the right coronary artery, but the left coronary cannula was inserted just past the ostium. Great care was taken to make certain that the cannula did not block off one of the main branches of the left coronary artery because of the uncertain length of the left main coronary artery. 10 The pressure in each cannula was maintained so as not to exceed 150 mm. Hg, if possible, and the flow generally varied from 75 to 200 c.c. per minute in each coronary artery. Results Of the 117 patients, 84 underwent isolated aortic valve replacement and 80 survived the immediate
postoperative period. Of the 33 who had multiple valve replacement, 23 survived postoperatively. Of the 103 survivors, 4 developed coronary ostial stenosis, an incidence of 3.8 per cent of the operative survivors and 3.5 per cent of the total number operated upon. Three of these 4 patients underwent isolated aortic valve replacement and one underwent a triple valve replacement. All the lesions were in the ostium of the left coronary artery, although the right coronary artery was perfused in 3 of the 4 cases. Two had aortic stenosis as the predominant aortic lesion. The remaining two had a mixed lesion of aortic stenosis and insuffiency. The average length of coronary perfusion in the entire series was 55 minutes. The longest was 2 hours, 10 minutes and the shortest, 35 minutes. The average length of coronary perfusion in the 4 patients developing the lesion was 65 minutes. A careful review of the operative record in all 4 patients revealed no particular problem related to the coronary perfusion or to the insertion of the cannulas. Case reports CASE 1. A 41-year-old man was admitted June 28, 1973. Cardiac catheterization revealed aortic stenosis and insufficiency, with a normal coronary arteriogram. The aortic valve was replaced on July 30, 1973, with a Starr-Edwards aortic valve prosthesis. Both coronary arteries were perfused uneventfully. The patient was readmitted on Nov. 14, 1973, with severe chest pain. Catheterization revealed a tight stenosis of the left main coronary artery (Fig. 1). A saphenous vein bypass to the left anterior descending was performed, with an uneventful recovery. The patient is presently asymptomatic. CASE 2. A 47-year-old white man was admitted to
82
Midell, DeBoer,
Bermudez
The Journal of Thoracic and Cardiovascular Surgery
Fig. 2. Right anterior oblique view demonstrating lesion of left main coronary artery in Case 3. Note that the strut of the aortic valve prosthesis almost obscures the lesion in this view.
i Fig. 3. Left anterior oblique view demonstrating left main coronary lesion in Case 4. Columbus Hospital on Aug. 10, 1973. Cardiac catheterization revealed severe aortic stenosis with a normal coronary arteriogram. On Aug. 27, 1973, an aortic valve replacement with a Starr-Edwards prosthesis was performed. Coronary perfusion of both arteries was unremarkable. After an uneventual recovery, the patient was readmitted on Dec. 10, 1973, with an obvious anginal syndrome. Catheterization revealed a normally functioning aortic valve prosthesis and severe narrowing of the ostium of the left main coronary artery (90 per cent). Saphenous vein bypass of the left anterior descending coronary artery was performed on Dec. 17, 1973, and resulted in fatal myocardial infarction on the day of the operation. CASE 3. A 49-year-old woman was admitted on May 5,
1974. Cardiac catheterization revealed aortic stenosis and insufficiency. The coronary arteriogram was normal. An aortic valve replacement was performed on May 13, 1974. Only the left coronary artery was perfused. After an uneventful recovery, she was readmitted on July 5, 1974, with severe angina, and a coronary arteriogram performed shortly thereafter showed 90 per cent narrowing of the left main coronary artery (Fig. 2). A coronary bypass to the left anterior descending was performed on July 6, 1974. After an uneventful recovery, the patient is presently asymptomatic. CASE 4. A 59-year-old woman was admitted on Dec. 19, 1974. Catheterization revealed aortic stenosis with mitral and tricuspid insufficiency. The coronary arteriogram was normal. A triple valve replacement was performed on Jan. 8,
Volume 72 Number 1 July, 1976
1975. The patient's recovery was complicated by severe congestive heart failure and pulmonary complications, but she eventually became well enough to be sent home after 8 weeks of hospitalization. She was readmitted on May 21, 1975, with severe recurrent chest pain. A coronary arteriogram revealed a severe ostial stenosis involving 90 per cent of the left main coronary artery (Fig. 3). A coronary artery bypass to the left anterior descending was performed, but the patient died on the fourteenth postoperative day from refractory congestive heart failure, pulmonary complications, and ventricular arrhythmia. Discussion The development of dangerous proximal lesions in the coronary arteries following cannulation, first described by Trimble and associates17 in 1959, has since been confirmed by Yates,18 Hazan,13 Lesage,14 and their co-authors, as well as by others. 15 ' 16 Patients in this review should be distinguished from those in whom malposition of the coronary cannulas results either in occlusion of one of the two branches of the left coronary artery or intimal dissection in a distal section. These developments usually result in early and massive infarction, which is entirely different from the lesion under discussion.10 Reports of coronary ostial stenosis are remarkably similar in their description of the onset and clinical pattern. In Trimble's17 original description, lesions were noted in both the right and left coronary ostia following the use of Mayo balloon catheters during aortic valve replacement. Pressures were not monitored in either cannula, and development of these lesions was attributed at the time to high perfusion pressures. One patient died at reoperation and the other 2 died without having another operation. In a subsequent communication Reed, Spencer, and colleagues16 reported an incidence of 5 per cent (6 patients) in 103 consecutive aortic valve replacements. Silicone rubber as well as self-inflating balloon catheters were used to perfuse the coronary arteries, but pressures were not monitored. The right coronary ostium was involved in 4 of the 6 patients. In 1973, Nakhjavan and associates15 reported on 3 patients in whom lesions developed in both coronary ostia following cannulation and perfusion. They did not mention what percentage of their total experience these lesions comprised, and they further speculated that the left coronary lesions might have been present prior to the original operation. More recently, Yates and associates18 reported on the largest number of patients in a single series developing coronary ostial stenosis, 8 out of 292 or 2.7 per cent. The interval between the original operation and the onset of angina varied up to 6 months. Left coronary
Postperfusion coronary ostial stenosis
83
artery stenosis developed in 6, both the right and left coronary arteries were involved in one, and the final patient had an isolated right coronary artery lesion. All required reoperation but only 5 of the 8 survived reoperation. Hazan, Rioux, and co-workers13 reported on 3 patients with left coronary artery stenosis following aortic valve replacement. Their incidence was 1.6 per cent of the total series. All required coronary artery bypass surgery and all survived. Lesage and colleagues14 have reported on 6 patients with iatrogenic coronary artery lesions following prosthetic replacement of the aortic valve but gave no indication of the incidence in their series. Our patients bore a striking resemblance to those in previously cited reports. AH 4 patients had selective coronary artery perfusion varying from 40 to 90 minutes, and a satisfactory perfusion was achieved in all 4 cases. In the immediate postoperative period no clinical or electrocardiographic evidence of ischemia or underlying injury to the coronary arteries was noted. However, within 20 weeks of the original operation, all 4 patients developed a severe, rapidly progressive anginal syndrome with electrocardiographic patterns of ischemia. In all 4 patients a saphenous vein bypass was required from the aorta to the left anterior descending coronary artery to achieve symptomatic relief. The remaining 99 survivors have not reported the development of angina since the operation, although 3 patients died suddenly after the first year of valve replacement. No autopsies were available. It is clear from a review of this subject that coronary ostial lesions may occur in both coronary ostia, although there seems to be a slight predisposition for the left coronary ostium. The lesions occur and develop in a relatively short period of time following the original operation. The initial complaint is invariably a progressive anginal syndrome accompanied by characteristic electrocardiographic changes. The cause of these lesions remains unproved, but the most acceptable theory is that they represent injuries to the ostium, either from insertion of the cannula or inflation of the balloon, resulting in an "intimal insult" with subsequent fibrosis and narrowing of the ostium.15' 23,24 Lesions are also observed in patients who are not perfused. This is thought to be due to repeated unsuccessful attempts to insert a cannula, resulting in injury to the ostium. Another explanation of this lesion is offered by Roberts and Morrow. 20-22 They postulate that turbulent flow around the newly placed prosthetic valve results in a hyperplastic reaction at the ostium. The fact that virtually no cases of coronary ostial
The Journal of
8 4 Midell, DeBoer, Bermudez
Thoracic and Cardiovascular Surgery
Table I. Comparative incidence of coronary ostial stenosis in previous reports Incidence Author
No. of cases
No.
%
Yates et al. 18 Hazan et al.13 Reed et al.16 Nakhjavan et al.15 DeBoer and Midell2 Lesage et al.14
292 185 103 7 117 ?
8 3 6 3 4 6
2 1.6 5 7 3 7
Deaths on reoperation No.
%
3 35 0 0 0 2 50 None reoperated
stenosis have been reported in the group of patients in whom ischemic arrest and hypothermia were used casts considerable doubt on this theory. The diagnosis of coronary ostial stenosis is determined by repeat coronary angiograms. This should be performed as soon as the lesion is suspected. Great care should be taken during angiographic study, as repeated injections can produce a dangerous arrhythmia. The left anterior oblique view is considered the most effective in demonstrating these lesions, as the prosthesis often obscures the ostium in the right anterior oblique view (Fig. 2). Clearly, the only effective treatment is early reoperation: Either a saphenous vein bypass or an anastomosis between the internal mammary artery and the left anterior descending coronary artery is used for lesions in the left main coronary artery; a saphenous vein bypass to the right coronary artery is employed for lesions in the right coronary ostium. We do not find it necessary to place another bypass to the circumflex coronary artery in those with left main coronary lesions, as bidirectional flow through the graft seems to provide adequate perfusion to the circumflex coronary artery if there is not a second lesion in the left anterior descending distal to the circumflex orifice. We favor the use of a saphenous vein over the internal mammary artery, as the vein bypass assures the high flows necessary to provide adequate perfusion to both the left anterior descending and the circumflex coronary artery. We conclude from our review that the incidence of coronary ostial stenosis is low, varying from 1 to 5 per cent (Table I). Robicsek and associates10 suggest it is probably higher than reported, however. This may well be true if we assume that certain "unexpected sudden deaths" following aortic valve replacement are due to this disease progressing to its logical terminus and that
these deaths have not heretofore been considered as being caused by coronary ostial stenosis or, indeed, to any injury to the coronary arteries. They are usually attributed to ventricular arrhythmia, noniatrogenic myocardial infarction, or pulmonary embolus. Certainly data from more carefully followed patients must be available before any final conclusion can be drawn regarding the true incidence of this complication. The total incidence of coronary ostial stenosis in their over-all experience is often omitted by the reporting authors (Table I). The significance of ostial stenosis nevertheless exceeds that of its lower reported incidence, because the mortality rate of reoperation is substantial. Two of the 4 patients in our series died during reoperation and 3 of the 8 patients reoperated by Yates did not survive. Although it is not the purpose of this discussion to compare the merits of the various methods of myocardial protection during aortic valve replacement, nevertheless our findings have relevance to this controversy. We have previously reported our experience with aortic valve replacement and have credited a low operative mortality rate to, among other factors, continuous coronary perfusion.2 We are presently continuing the use of coronary perfusion in contrast to ischemic arrest, but we have replaced the Mayo balloon catheters with O-ring cannulas in the hope that this will minimize, or even eliminate, the development of coronary ostial stenosis. At this time we do not have enough experience to provide meaningful data. If efforts to eliminate this complication fail, then we must ask whether the continued development of this problem, with its rapid evolution and necessity for reoperation, can be tolerated in any large surgical experience with aortic valve replacement. Certainly, the development of these lesions, even with the low incidence as described, must be considered in any analysis of the over-all results of aortic valve replacement performed with coronary perfusion, even though the patient may survive the initial period of operation. The use of ischemic arrest and hypothermia has been advocated by Bloodwell,1 Shumway,12 Robicsek,9 and others23, 8 for some years. Recent clinical and experimental evidence from several medical centers seems to substantiate the hypothesis that the myocardium can tolerate long periods of anoxic arrest under hypothermia without incurring irreversible damage. 4,5 ' 23, 24 i^g ii m jt s 0 f toleration of ischemia have not yet been tightly defined, however. In a recent editorial, Robicsek9 stated that ischemic arrest and hypothermia is appropriate only to those who are "quick operators." The time required to replace the aortic valve varies
Volume 72 Number 1 July, 1976
Postperfusion
substantially in each individual case. Replacing the aortic valve in a large, dilated, noncalcified annulus often requires much less time than replacing a densely calcified valve in a narrow annulus, as is often seen in calcific aortic stenosis. Therefore, the final determinant of the preferred method of myocardial protection will ultimately depend on the skill of the individual surgeon. It is unlikely that one method will " w i n " in that it will be used exclusively over the other. Each surgeon must consider whether in his hands the incidence of injury and infarction from prolonged ischemic arrest equals or exceeds the incidence of iatrogenic coronary artery injury from cannulation and perfusion. REFERENCES 1 Bloodwell, R. D., Kidd, M. J., Hallman, G. L., Burdette, W. J., McMurtrey, M. J., and Cooley, D. A.: Cardiac Valve Replacement Without Coronary Artery Perfusion: Clinical and Laboratory Observations, in Brewer, L. A., editor: Prosthetic Heart Valves, Springfield, 111., 1969, Charles C Thomas, Publisher. 2 DeBoer, A., and Midell, A. I.: Isolated Aortic Valve Replacement: Analysis of Factors Influencing Survival After Replacement With the Starr-Edwards Prosthesis Ann. Thorac. Surg. 17: 360, 1974. 3 Fishman, N. H., Youker, J. E., and Benson, B. R.: Mechanical Injury to the Coronary Arteries During Operative Cannulation, Am. Heart J. 75: 26, 1968. 4 Griepp, R. B., Stinson, E. B., and Shumway, N. E.: Profound Local Hypothermia for Myocardial Protection During Open-Heart Surgery, J. THORAC. CARDIOVASC. SURG. 66: 731, 1973.
5 Hurley, E. J., Lower, R. R., Dong, E., Pillsbury, R. C , and Shumway, N. E.: Clinical Experience With Local Hypothermia in Elective Cardiac Arrest, J. THORAC. CARDIOVASC. SURG. 47: 50,
1964.
6 Iyengar, S. R. K., Ramchaud, S., Charrette, E. J. P., Iyengar, C. K. S., and Lynn, R. B.: Anoxic Cardiac Arrest: An Experimental and Clinical Study of Its Effects, J. THORAC CARDIOVASC. SURG. 66: 722,
1973.
7 Levitsky, S., Sloane, R. E., Mullin, E. M., Mclntosh, C. L., and Morrow, A. G.: Normothermic Myocardial Anoxia: Effects on the Canine Heart With Left Ventricular Outflow Obstruction, Ann. Thorac. Surg. 11: 229, 1971. 8 Pupelli, D. F., Blank, R. H., Connar, R. G., Bessone, L. N., Sbar, S., and Stevenson, A.: Fifty-two Consecutive Aortic Valve Replacements Employing Local Deep Hypothermia, Ann. Thorac. Surg. 17: 360, 1974. 9 Robicsek, F.: Local Cardiac Hypothermia in Lieu of Coronary Perfusion (Editorial), Ann. Thorac. Surg. 19: 487, 1975.
coronary ostial stenosis
85
10 Robicsek, F., Tarn, W., Doughterty, H. K., and Mullen, D. C : Myocardial Protection During Open Heart Surgery, Ann. Thorac. Surg. 10: 341, 1970. 11 Sapsford, R. N., Blackstone, E. H., Kirklin, J. W., Karp, R. B., Kouchoukos, N. T., Pacifico, A. D., Roe, C. R., and Bradley, E. L.: Coronary Perfusion Versus Cold Ischemic Arrest During Aortic Valve Surgery, Circulation 49: 1190, 1974. 12 Shumway, N. E., Lower, R. R., and Staffer, R. C : Selective Hypothermia of the Heart in Anoxic Cardiac Arrest, Surg. Gynecol. Obstet. 109: 750, 1959. 13 Hazan, E., Rioux, C , Dequirot, A., and Mathey, J.: Postperfusion Stenosis of the Common Left Coronary Artery, J. THORAC. CARDIOVASC. SURG. 69: 703,
1975.
14 Lesage, C. H., Jr., Vogel, J. H., and Blount, S. G., Jr.: Iatrogenic Coronary Occlusive Disease in Patients With Prosthetic Heart Valves, Am. J. Cardiol. 26: 123, 1970. 15 Nakhjavan, F. K., Maranhao, V., and Goldberg, H.: Iatrogenic Stenosis of the Proximal Portion of the Coronary Arteries, Am. Heart J. 83: 318, 1972. 16 Reed, G. E., Spencer, F. C , Boyd, A. D., Engelman, R. M., and Glassman, E.: Late Complications of Intraoperative Coronary Artery Perfusion, Circulation 48: 80, 1973 (Suppl. III). 17 Trimble, A. S., Bigelow, W. G., Wigle, E. D., and Silver, M. D.: Coronary Ostial Stenosis, J. THORAC CARDIOVASC SURG. 57: 792,
1969.
18 Yates, J. D., Kirsh, M. M., Sodeman, T. M., Walton, J. S., Jr., and Bryner, M. F.: Coronary Ostial Stenosis: A Complication of Aortic Valve Replacement, Circulation 49: 530, 1974. 19 Kronzon, I., Deutsch, R., and Glassman, E.: Length of Left Main Coronary Artery: Its Relations to the Pattern of Coronary Arterial Distribution, Am. J. Cardiol. 34: 787, 1974. 20 Roberts, W. C , and Morrow, A. G.: Late Postoperative Pathologic Findings After Cardiac Valve Replacement, Circulation 35: 68, 1967 (Suppl. I). 21 Roberts, W. C., and Morrow, A. G.: Anatomic Studies of Hearts Containing Caged-Ball Prosthetic Valves, Johns Hopkins Med. J. 121: 271, 1967. 22 Roberts, W. C , and Morrow, A. G.: Causes of Early Postoperative Death Following Cardiac Valve Replacement: Clinicopathologic Correlations in 64 Patients Studied at Necropsy, J. THORAC CARDIOVASC SURG.
54: 792, 1969. 23 Angell, W. W., Rikkers, L., Dong, E., Jr., and Shumway, N. E.: Organ Viability With Hypothermia, J. THORAC CARDIOVASC SURG. 58: 619,
1969.
24 Buja, L. M., Ferrans, V. J., and Levitsky, S.: Occurrence of Intra-mitochondrial Glycogen in Cananine Myocardium After Prolonged Anoxic Cardiac Arrest, J. Mol. Cell. Cardiol. 4: 237, 1972.