Emergency Aortic Valve Replacement John H. Sanders, Jr, MD, Boston, Massachusetts Lawrence H. Cohn, MD, Boston Massachusetts James E. Dalen, MD,’ Boston, Massachusetts John J. Collins, Jr, MD, Boston, Massachusetts
Death as a result of valvular heart disease may occur as the final event of a prolonged course of deterioration or it may occur more suddenly and even without warning. Rapid hemodynamic deterioration is more often seen with isolated aortic valve dysfunction than with disease of other valves singly or in combination. The rapidity of progression in patients with symptomatic aortic stenosis has been well described [I] and is widely appreciated. Nonetheless, some patients progress to a point where life for even a few more hours seems highly unlikely; the malignant pace of left ventricular power failure may surprise both patient and physician. The following series of patients who underwent operation when they were apparently moribund with aortic stenosis or insufficiency illustrates that satisfactory results may be obtained with aortic valve replacement even under the most dire circumstances. Clinical Material Emergency isolated aortic valve replacement was performed in twenty-seven patients at the Peter Bent Brigham Hospital during the four years from July 1970 through June 1974. Patients’ ages ranged from thirtynine to eighty-one years (mean, 61.4 years). There were seventeen males and ten females. Indications for operation included low cardiac output, cardiac arrest, and intractable pulmonary edema. Patients with low cardiac output underwent emergency aortic valve replacement when inadequate peripheral perfusion resulted in confusion or oliguria with increasing blood urea level. Cardiac arrest was sudden and unexpected in only one patient, whereas in three patients, deterioration was evident for more than 30 minutes before loss of an organized heart beat. One patient had ventricular fibrillation at catheFrom the Departments of Surgery and Medicine, Harvard Medical School, and Peter Bent Brigham Hospital, Boston, Massachusetts. Reprint requests should be addressed to John H. Sanders, Jr, MD, Division of Thoracic and Cardiac Surgery, Peter Sent Brlgham Hospital, 721 Huntinoton Avenue. Boston. Massachusetts 02115. Presented at the Fifty-Sixth Annual Meeting of me New England Surgical Society, Portsmouth, New Hampshire, September 25-27, 1975. Present address: Department of Surgery, University of Massachusetts Medical School, Worcester, Massachusetts. l
vohlme 131. April 1976
terization. Pulmonary edema was considered intractable when aggressive medical therapy failed to halt deterioration in respiratory status or resulted in hypotension. Predominant aortic stenosis (AS) was present in eleven patients, combined aortic stenosis and insufficiency (AS-AI) in eleven patients, aortic insufficiency (AI) in four patients, and the combination of aortic insufficiency with ventricular septal defect (VSD) in one patient. Active endocarditis was present in all four patients with AI alone and in the single patient with AI-VSD. Duration of symptoms showed great variation ranging from a few days to twenty-two years from the date patients were Class I, according to the New York Heart Association (NYHA), until operation was performed. (Table I.) No patient in this series was entirely without symptoms prior to the catastrophic events which precipitated emergency surgery. Of the eleven patients with AS, ten had congestive heart failure for an average duration of seventeen months, seven had angina for an average of thirty-two months, and five had one or more episodes of syncope from a few hours to two years before operation. The eleven patients with AS-AI all had congestive heart failure with an average duration of twentysix months, and five had angina for an average of twenty months. None had syncope. The duration of congestive heart failure ranged from one month to fifteen years in patients with AS and two months to six years in those with AS-AI. The range for angina was three months to thirteen years in patients with AS and one month to six years in those with AS-Al. ‘The five patients with AZ had a much shorter duration of symptoms for congestive heart failure (2 weeks to 6 months, with a mean of 2.9 months). None had angina and only one had a syncopal episode occurring two weeks before operation, presumably as a result of embolization from a bacterial vegetation. Regarding immediate indications for surgery, eighteen patients (67 per cent) had intractable pulmonary edema (7 with AS, 7 with AS-AI, and 4 with AI), four patients (15 per cent) had low cardiac output (cardiogenie shock) (1 with AS, 2 with AS-AI, and 1 with AI) and five patients (19 per cent) had cardiac arrest (3 with AS and 2 with AS-AI). All patients with active endocarditis were in sinus rhythm at operation. The patients with AS or AS-AI showed sinus rhythm in eighteen instances, atria1 fibrillation in two, and atria1 flutter, bigeminy, and atria1 tachycardia in one each. Left anterior hemiblock was
495
Sanders
et al
TAbLE I Incidence and Duration of Symptoms ~______~___________________ __c-_~ ______-________ Angina Congestive Heart Failure Number of Patients AS (11 patients) AS-AI (11 patients) Al (5 patients)
--_-__
17.0 10 26.0 11 2.9 5 ___________________--__-___
observed in four patients. Twenty-three patients had electrocardiographic evidence of left ventricular hypertrophy, including four of the five patients with AI. Roentgenographic signs of left ventricular enlargement were found in twenty-five patients with evidence of interstitial pulmonary edema in twenty-six patients. Aor-
tic valve calcification was detected by fluoroscopy in nineteen of the twenty-two patients with AS or AS-AI and in none of the patients with active endocarditis and AI. Cardiac catheterization was performed in seventeen patients: eleven patients, catheterized electively, came to emergency valve replacement because of progressive pulmonary edema or hypotension, sometimes occurring two or three days after the study, and six patients were catheterized as an emergency to clarify their disease and went directly to the operating room. Three patients were catheterized more than three years before their ultimate clinical deterioration. Catheterization was not performed in ten patients because their condition was considered too critical. Catheterization was terminated prematurely in three patients because of intractable pulmonary edema in two and cardiac arrest in the other. Cardiac index in fifteen patients averaged 2.2 L/min/M2 with only two patients above 2.9 L/min/M2. Peak gradients across the aortic valve in the sixteen patients in whom gradients were obtained varied from 30 to 120 mm Hg, averaging ‘75 mm Hg. Calculated aortic valve area in the eleven patients in whom both the gradient and cardiac index were obtained averaged 0.43 cm2 with a range of 0.3 to 0.6 cm2. Coronary arteriograms were performed in nine patients in the AS and AS-AI groups,
all of whom had angina. In seven patients no significant obstructive lesion was noted. One patient had isolated right coronary obstruction in two areas and one had a 50 per cent middle third lesion in the left anterior descending coronary artery. Operations were performed using total cardiopulmonary bypass with a bubble oxygenator, moderate hemodilution (hematocrit, 18 to 26 per cent), and total body hypothermia (28 to 32’C). Coronary perfusion with the heart beating was used for myocardial protection in five patients early in this series. The remaining twenty-three patients were managed with electrically induced ventricular fibrillation and local cardiac hypothermia by a technic previously described [2]. Ischemia time in the hypothermia group ranged from 53 to 110 minutes, averaging 66 minutes.
496
Number of Patients
Mean Duration (months)
7 5 0
Mean Duration (months)
Number of Patients with Syncope
20.5 20.0 ------~
5 0 1
Starr-Edwards model #I200 valves were used in twelve patients, the Bjork-Shiley valve in eleven patients, Harken valves with cloth-covered struts in two patients, and a Braunwald-Cutter, porcine heterograft (Hancock), and Cutter-Smeloff valve in one patient each. Results
There were no in-hospital deaths. The length of hospitalization after surgery for the AS and AS-AI groups varied from nine to thirty-one days, averaging sixteen days. There was no correlation between the length of hospital stay and the technic of myocardial protection during valve replacement. Patients operated on for AI, all of whom had active endocarditis, had longer stays after surgery (29 to 109 days; mean, 56 days) because of the need for intravenous antibiotic therapy which was continued for four to six weeks after valve replacement. In the follow-up period of five to thirty-three months, no patient had recurrence of infection. There has been one late death at nine months of apparent myocardial infarction. One patient required reoperation to relieve a clotted prosthesis at twenty-one months. Nineteen of the twenty-two patients in the AS and AS-AI group are Class I (NYHA). One patient continues to be moderately limited with exertional dyspnea. All of the patients operated on for hemodynamic emergencies during active endocarditis are alive and are Class I (NYHA). Comments
Emergency aortic valve replacement for aortic insufficiency due to bacterial endocarditis has been described in a number of reports and reviews [3-61. E mer genc y va 1ve replacement for preterminal hemodynamic deterioration in patients with predominant aortic stenosis has been described less frequently [7,8]. Our series of surgically treated patients includes twenty-three patients whose predominant lesion was aortic stenosis. In eleven of these, there was mild to moderate aortic insuffi-
The American Journal of Surgery
Aortic Valve Replacement
ciency as well as aortic stenosis, demonstrated either by the presence of a characteristic diastolic murmur or by aortic root angiogram. All five patients with dominant aortic insufficiency had active bacterial endocarditis and underwent emergency operation because it appeared they would not survive otherwise. In each patient in the entire series survival for even a few days longer was considered doubtful without valve replacement. It is of interest that the time from onset of congestive heart failure and angina pectoris to operation in the patients with AS and AS-AI closely approximates the expected span of life after development of these symptoms found by those studying the natural history of aortic stenosis [I]. Perhaps because of the relatively small number of patients with pure AS and with mixed AS-AI and the considerable individual variation in symptom duration, there was no apparent significant difference in the natural history of AS as compared with ASAI. Syncope occurred in only five patients with wide variation in the duration before surgery. In one patient increasing frequency of syncopal attacks to several each day was clearly life-threatening. Although sudden death is a well recognized hazard in patients with aortic stenosis, only one patient in this series had a totally unexpected cardiac arrest. This patient succumbed in the hospital parking lot while on his way to work and was resuscitated by the alert action of a passing anesthesiologist. Later questioning revealed that this patient had had angina for nearly a year but had not sought medical consultation. The remaining four patients with preoperative cardiac arrest were known to be critically ill. One developed ventricular fibrillation during cardiac catheterization and the others suffered cardiac arrest during efforts to correct intractable hypotension. In each instance it was possible to reestablish an organized rhythm although no blood pressure was obtainable in one patient at the beginning of operation. The occurrence of severe congestive heart failure in patients with predominant aortic stenosis presents a difficult and often near impossible therapeutic dilemma for the cardiologist. When diuretics are administered, the temporary loss of circulating blood volume accompanying any significant diuresis may produce a decrease in cardiac output sufficient to reduce cerebral and renal blood flow to dangerous levels. In the chronic situation the blood urea level increases and the patient becomes progressively weak and somnolent or confused. Myocardial &hernia may become significant and
volume 131, April 1976
further reduce the cardiac output, thus predisposing to the development of hazardous ventricular arrhythmias. More acutely, particularly with the use of intravenous diuretic agents, the course of events may be quite rapid, with “cardiac arrest” as the final catastrophe. It has been our practice to proceed with immediate valve replacement when diuresis causes significant hypotension or results in progressive azotemia and weakness, or when respiratory embarrassment increases despite therapy. Aortic stenosis may be suspected when an ejection systolic murmur is audible at the base of the heart, but in terminal cases the murmur may be very soft because of diminished cardiac output. A history of such a murmur should not be neglected. A characteristic pulse contour may be helpful and echocardiographic demonstration of calcification and limited mobility of the aortic leaflets is strong evidence for aortic stenosis. When these signs exist and the electrocardiogram shows left ventricular hypertrophy, the diagnosis is virtually certain. The surgical significance of angina pectoris is difficult to define considering our own experience and that reported by others [9,10]. The patients in this series, although having an appreciable incidence of angina, had relatively minimal coronary obstructive disease. Does this suggest that the presence of severe coronary obstructive disease predisposes to earlier death or earlier surgery? Perhaps. Certainly, we have seen a higher incidence of significant coronary artery disease in patients undergoing elective aortic valve replacement than was evident in this series of patients with end-stage hemodynamic deterioration. Preliminary cardiopulmonary bypass was not utilized in any patient. The usual anesthetic consisted of morphine, nitrous oxide, and oxygen. All cannulations were carried out within the chest using a single venous cannula in the right atrium and an arterial cannula in the ascending aorta. Since this can be accomplished, if necessary, in about eight to ten minutes, it was not considered necessary to utilize partial bypass before anesthetic induction. No other form of left ventricular assist was used either before or after operation in this series. The satisfactory recovery of all patients undergoing operation demonstrates that surgery should not be withheld in patients with critical aortic stenosis because the patient is “too ill.” Aortic valve replacement is, in fact, the only possible recourse for those patients who deteriorate while in the hospital on strict medical management programs.
497
Sanders et al
Cardiac catheterization, with coronary angiography in patients with angina or a history of myocardial infarction, is useful for establishment of a diagnosis and to exclude additional valve disease or cardiomyopathy. However, surgeons and cardiologists should not hesitate to recommend operation without catheterization when the clinical diagnosis seems assured and when the critical nature of the illness makes catheterization hazardous. Summary
The aortic valve was replaced as an emergency in twenty-seven patients between July 1970 and December 1974. Twenty-two patients had critical aortic stenosis and five had acute aortic insufficiency. The indications for emergency surgery were cardiac arrest in five patients, low cardiac output in four patients, and medically intractable pulmonary edema in eighteen patients. Cardiac catheterization was not undertaken in ten persons because of their critical condition. The clinical diagnosis in these patients was supported by noninvasive maneuvers. No surgical deaths occurred. Twenty-five patients are well, active, and NYHA Class I at five to thirty-three months after surgery. There has been one late death and one patient has some residual exertional dyspnea. One patient required reoperation to relieve a clotted prosthetic valve. These results suggest that the patient with aortic valve disease may be offered a reasonable chance for survival, even when desperately ill. References 1. Cohen LS, Friedman WF, Braunwald E: Natural history of mild congenital aortic stenosis elucidated by serial hemodynamic studies. Am J Cardiol30: 1. 1972. 2. Cohn LH. Collins JJ Jr: Local cardiac hypothermia for myocardial protection. Ann Thorac Surg 17: 135, 1974. 3. Okies JE, Bradshaw MW, Williams TW Jr: Valve replacement in bacterial endocarditis. Chest 63: 898, 1973. 4. Wise JR Jr, Cleland WP, Hallidie-Smith KA, et al: Urgent aortic valve replacement for acute aortic regurgitation due to infective endocarditis. Lancet 2: 115, 1971. 5. Neville WE, Magno M, Foxworthy DT, et al: Emergency aortic valve replacement in bacterial endocarditis. J Thorac Cardiovasc Surg 6 1: 9 16, 197 1. 6. Manhas DR. Mohri H, Hessel EA II, et al: Experience with surgical management of primary infective endocarditis. Am Heart J 84: 738, 1972. 7. Stinson EB, Shumway NE: Emergency heart valve replacement. Calif Med 109: 441, 1968. 8. Hutter AM Jr, DeSanctis RW, Nathan MJ, et al: Aortic valve surgery as an emergency procedure. Circulation 41: 623, 1970. 9. Manchester JH, Amsterdam EA, Kemp HG, et al: Coronary artery disease in valvular heart disease (abstract). Circulation 38 (Suppl 6): 132, 1968. 10. Coleman EH, Soloff LA: Incidence of significant coronary ar-
498
tery disease in rheumatic valvular heart disease. Am J Cardiol25: 401, 1970.
Discussion Richard J. Cleveland (Boston, MA): I will direct my comments to those patients with long-standing AS or combined AS-AI. Unfortunately, all of us involved in cardiac surgery have frequently had disappointing longterm results in patients such as these. Most cardiac surgeons have found it possible to operate successfully on such patients; however the patients never really return to the level of activity which we had initially anticipated. There are several elegant studies published and our own clinical experience indicates that when a patient is allowed to reach the level of cardiac disability reported in this series, that irreversible myocardial damage is almost invariably present. This myocardial damage obviously adversely effects the long-term prognosis of the patient. I believe we should be more aggressive in the surgical management of aortic valve disease, and procrastination for any reason for a prolonged period of time after the onset of symptoms is not in a patient’s best interest. It is of interest that so many of these patients reported by the authors, who were essentially moribund, have returned to NYHA Class I. I think this is a bit unusual and does not reflect the general experience of others. It is incumbent upon the surgeon to continue to encourage medical colleagues to refer patients such as these much earlier in their clinical course. The surgeon will operate at a lower mortality and just as importantly, the patient will be restored to a meaningful existence after surgical correction. John H. Sanders, Jr (closing): Actually, we do not have as much hemodynamic information as we would like. Twenty-two of these patients did have evidence of significant cardiomegaly on chest x-ray examination, but most of these did not have ventriculograms. The diastolic volumes and ejection fractions are, therefore, difficult to assess. (Slide) The x-ray film on the left is of an individual in this series who had fairly marked cardiomegaly preoperatively. In film taken one month postoperatively, the heart size is diminished, the effusions are gone and the pulmonary vascular redistribution has begun to improve. It has been our experience that the patients who have had ventriculography have had ejection fractions of 20 to 30 per cent. These patients have done well in our experience. One speculation is that although the ejection fraction may be 20 per cent, many of these patients can generate fairly substantial intraventricular pressures, in the range of 180 mm Hg. The fact that a dilated ventricle is capable of generating a high pressure, even with a small ejection fraction, has led us to suspect that the ventricular function will improve when the gradient is removed.
Death as a result of valvular heart disease may occur as the final event of a prolonged course of deterioration or it may occur more suddenly and even without warning. Rapid hemodynamic deterioration is more often seen with isolated aortic valve dysfunction than with disease of other valves singly or in combination. The rapidity of progression in patients with symptomatic aortic stenosis has been well described [I] and is widely appreciated. Nonetheless, some patients progress to a point where life for even a few more hours seems highly unlikely; the malignant pace of left ventricular power failure may surprise both patient and physician. The following series of patients who underwent operation when they were apparently moribund with aortic stenosis or insufficiency illustrates that satisfactory results may be obtained with aortic valve replacement even under the most dire circumstances. Clinical Material Emergency isolated aortic valve replacement was performed in twenty-seven patients at the Peter Bent Brigham Hospital during the four years from July 1970 through June 1974. Patients’ ages ranged from thirtynine to eighty-one years (mean, 61.4 years). There were seventeen males and ten females. Indications for operation included low cardiac output, cardiac arrest, and intractable pulmonary edema. Patients with low cardiac output underwent emergency aortic valve replacement when inadequate peripheral perfusion resulted in confusion or oliguria with increasing blood urea level. Cardiac arrest was sudden and unexpected in only one patient, whereas in three patients, deterioration was evident for more than 30 minutes before loss of an organized heart beat. One patient had ventricular fibrillation at catheFrom the Departments of Surgery and Medicine, Harvard Medical School, and Peter Bent Brigham Hospital, Boston, Massachusetts. Reprint requests should be addressed to John H. Sanders, Jr, MD, Division of Thoracic and Cardiac Surgery, Peter Sent Brlgham Hospital, 721 Huntinoton Avenue. Boston. Massachusetts 02115. Presented at the Fifty-Sixth Annual Meeting of me New England Surgical Society, Portsmouth, New Hampshire, September 25-27, 1975. Present address: Department of Surgery, University of Massachusetts Medical School, Worcester, Massachusetts. l
vohlme 131. April 1976
terization. Pulmonary edema was considered intractable when aggressive medical therapy failed to halt deterioration in respiratory status or resulted in hypotension. Predominant aortic stenosis (AS) was present in eleven patients, combined aortic stenosis and insufficiency (AS-AI) in eleven patients, aortic insufficiency (AI) in four patients, and the combination of aortic insufficiency with ventricular septal defect (VSD) in one patient. Active endocarditis was present in all four patients with AI alone and in the single patient with AI-VSD. Duration of symptoms showed great variation ranging from a few days to twenty-two years from the date patients were Class I, according to the New York Heart Association (NYHA), until operation was performed. (Table I.) No patient in this series was entirely without symptoms prior to the catastrophic events which precipitated emergency surgery. Of the eleven patients with AS, ten had congestive heart failure for an average duration of seventeen months, seven had angina for an average of thirty-two months, and five had one or more episodes of syncope from a few hours to two years before operation. The eleven patients with AS-AI all had congestive heart failure with an average duration of twentysix months, and five had angina for an average of twenty months. None had syncope. The duration of congestive heart failure ranged from one month to fifteen years in patients with AS and two months to six years in those with AS-AI. The range for angina was three months to thirteen years in patients with AS and one month to six years in those with AS-Al. ‘The five patients with AZ had a much shorter duration of symptoms for congestive heart failure (2 weeks to 6 months, with a mean of 2.9 months). None had angina and only one had a syncopal episode occurring two weeks before operation, presumably as a result of embolization from a bacterial vegetation. Regarding immediate indications for surgery, eighteen patients (67 per cent) had intractable pulmonary edema (7 with AS, 7 with AS-AI, and 4 with AI), four patients (15 per cent) had low cardiac output (cardiogenie shock) (1 with AS, 2 with AS-AI, and 1 with AI) and five patients (19 per cent) had cardiac arrest (3 with AS and 2 with AS-AI). All patients with active endocarditis were in sinus rhythm at operation. The patients with AS or AS-AI showed sinus rhythm in eighteen instances, atria1 fibrillation in two, and atria1 flutter, bigeminy, and atria1 tachycardia in one each. Left anterior hemiblock was
495
Sanders
et al
TAbLE I Incidence and Duration of Symptoms ~______~___________________ __c-_~ ______-________ Angina Congestive Heart Failure Number of Patients AS (11 patients) AS-AI (11 patients) Al (5 patients)
--_-__
17.0 10 26.0 11 2.9 5 ___________________--__-___
observed in four patients. Twenty-three patients had electrocardiographic evidence of left ventricular hypertrophy, including four of the five patients with AI. Roentgenographic signs of left ventricular enlargement were found in twenty-five patients with evidence of interstitial pulmonary edema in twenty-six patients. Aor-
tic valve calcification was detected by fluoroscopy in nineteen of the twenty-two patients with AS or AS-AI and in none of the patients with active endocarditis and AI. Cardiac catheterization was performed in seventeen patients: eleven patients, catheterized electively, came to emergency valve replacement because of progressive pulmonary edema or hypotension, sometimes occurring two or three days after the study, and six patients were catheterized as an emergency to clarify their disease and went directly to the operating room. Three patients were catheterized more than three years before their ultimate clinical deterioration. Catheterization was not performed in ten patients because their condition was considered too critical. Catheterization was terminated prematurely in three patients because of intractable pulmonary edema in two and cardiac arrest in the other. Cardiac index in fifteen patients averaged 2.2 L/min/M2 with only two patients above 2.9 L/min/M2. Peak gradients across the aortic valve in the sixteen patients in whom gradients were obtained varied from 30 to 120 mm Hg, averaging ‘75 mm Hg. Calculated aortic valve area in the eleven patients in whom both the gradient and cardiac index were obtained averaged 0.43 cm2 with a range of 0.3 to 0.6 cm2. Coronary arteriograms were performed in nine patients in the AS and AS-AI groups,
all of whom had angina. In seven patients no significant obstructive lesion was noted. One patient had isolated right coronary obstruction in two areas and one had a 50 per cent middle third lesion in the left anterior descending coronary artery. Operations were performed using total cardiopulmonary bypass with a bubble oxygenator, moderate hemodilution (hematocrit, 18 to 26 per cent), and total body hypothermia (28 to 32’C). Coronary perfusion with the heart beating was used for myocardial protection in five patients early in this series. The remaining twenty-three patients were managed with electrically induced ventricular fibrillation and local cardiac hypothermia by a technic previously described [2]. Ischemia time in the hypothermia group ranged from 53 to 110 minutes, averaging 66 minutes.
496
Number of Patients
Mean Duration (months)
7 5 0
Mean Duration (months)
Number of Patients with Syncope
20.5 20.0 ------~
5 0 1
Starr-Edwards model #I200 valves were used in twelve patients, the Bjork-Shiley valve in eleven patients, Harken valves with cloth-covered struts in two patients, and a Braunwald-Cutter, porcine heterograft (Hancock), and Cutter-Smeloff valve in one patient each. Results
There were no in-hospital deaths. The length of hospitalization after surgery for the AS and AS-AI groups varied from nine to thirty-one days, averaging sixteen days. There was no correlation between the length of hospital stay and the technic of myocardial protection during valve replacement. Patients operated on for AI, all of whom had active endocarditis, had longer stays after surgery (29 to 109 days; mean, 56 days) because of the need for intravenous antibiotic therapy which was continued for four to six weeks after valve replacement. In the follow-up period of five to thirty-three months, no patient had recurrence of infection. There has been one late death at nine months of apparent myocardial infarction. One patient required reoperation to relieve a clotted prosthesis at twenty-one months. Nineteen of the twenty-two patients in the AS and AS-AI group are Class I (NYHA). One patient continues to be moderately limited with exertional dyspnea. All of the patients operated on for hemodynamic emergencies during active endocarditis are alive and are Class I (NYHA). Comments
Emergency aortic valve replacement for aortic insufficiency due to bacterial endocarditis has been described in a number of reports and reviews [3-61. E mer genc y va 1ve replacement for preterminal hemodynamic deterioration in patients with predominant aortic stenosis has been described less frequently [7,8]. Our series of surgically treated patients includes twenty-three patients whose predominant lesion was aortic stenosis. In eleven of these, there was mild to moderate aortic insuffi-
The American Journal of Surgery
Aortic Valve Replacement
ciency as well as aortic stenosis, demonstrated either by the presence of a characteristic diastolic murmur or by aortic root angiogram. All five patients with dominant aortic insufficiency had active bacterial endocarditis and underwent emergency operation because it appeared they would not survive otherwise. In each patient in the entire series survival for even a few days longer was considered doubtful without valve replacement. It is of interest that the time from onset of congestive heart failure and angina pectoris to operation in the patients with AS and AS-AI closely approximates the expected span of life after development of these symptoms found by those studying the natural history of aortic stenosis [I]. Perhaps because of the relatively small number of patients with pure AS and with mixed AS-AI and the considerable individual variation in symptom duration, there was no apparent significant difference in the natural history of AS as compared with ASAI. Syncope occurred in only five patients with wide variation in the duration before surgery. In one patient increasing frequency of syncopal attacks to several each day was clearly life-threatening. Although sudden death is a well recognized hazard in patients with aortic stenosis, only one patient in this series had a totally unexpected cardiac arrest. This patient succumbed in the hospital parking lot while on his way to work and was resuscitated by the alert action of a passing anesthesiologist. Later questioning revealed that this patient had had angina for nearly a year but had not sought medical consultation. The remaining four patients with preoperative cardiac arrest were known to be critically ill. One developed ventricular fibrillation during cardiac catheterization and the others suffered cardiac arrest during efforts to correct intractable hypotension. In each instance it was possible to reestablish an organized rhythm although no blood pressure was obtainable in one patient at the beginning of operation. The occurrence of severe congestive heart failure in patients with predominant aortic stenosis presents a difficult and often near impossible therapeutic dilemma for the cardiologist. When diuretics are administered, the temporary loss of circulating blood volume accompanying any significant diuresis may produce a decrease in cardiac output sufficient to reduce cerebral and renal blood flow to dangerous levels. In the chronic situation the blood urea level increases and the patient becomes progressively weak and somnolent or confused. Myocardial &hernia may become significant and
volume 131, April 1976
further reduce the cardiac output, thus predisposing to the development of hazardous ventricular arrhythmias. More acutely, particularly with the use of intravenous diuretic agents, the course of events may be quite rapid, with “cardiac arrest” as the final catastrophe. It has been our practice to proceed with immediate valve replacement when diuresis causes significant hypotension or results in progressive azotemia and weakness, or when respiratory embarrassment increases despite therapy. Aortic stenosis may be suspected when an ejection systolic murmur is audible at the base of the heart, but in terminal cases the murmur may be very soft because of diminished cardiac output. A history of such a murmur should not be neglected. A characteristic pulse contour may be helpful and echocardiographic demonstration of calcification and limited mobility of the aortic leaflets is strong evidence for aortic stenosis. When these signs exist and the electrocardiogram shows left ventricular hypertrophy, the diagnosis is virtually certain. The surgical significance of angina pectoris is difficult to define considering our own experience and that reported by others [9,10]. The patients in this series, although having an appreciable incidence of angina, had relatively minimal coronary obstructive disease. Does this suggest that the presence of severe coronary obstructive disease predisposes to earlier death or earlier surgery? Perhaps. Certainly, we have seen a higher incidence of significant coronary artery disease in patients undergoing elective aortic valve replacement than was evident in this series of patients with end-stage hemodynamic deterioration. Preliminary cardiopulmonary bypass was not utilized in any patient. The usual anesthetic consisted of morphine, nitrous oxide, and oxygen. All cannulations were carried out within the chest using a single venous cannula in the right atrium and an arterial cannula in the ascending aorta. Since this can be accomplished, if necessary, in about eight to ten minutes, it was not considered necessary to utilize partial bypass before anesthetic induction. No other form of left ventricular assist was used either before or after operation in this series. The satisfactory recovery of all patients undergoing operation demonstrates that surgery should not be withheld in patients with critical aortic stenosis because the patient is “too ill.” Aortic valve replacement is, in fact, the only possible recourse for those patients who deteriorate while in the hospital on strict medical management programs.
497
Sanders et al
Cardiac catheterization, with coronary angiography in patients with angina or a history of myocardial infarction, is useful for establishment of a diagnosis and to exclude additional valve disease or cardiomyopathy. However, surgeons and cardiologists should not hesitate to recommend operation without catheterization when the clinical diagnosis seems assured and when the critical nature of the illness makes catheterization hazardous. Summary
The aortic valve was replaced as an emergency in twenty-seven patients between July 1970 and December 1974. Twenty-two patients had critical aortic stenosis and five had acute aortic insufficiency. The indications for emergency surgery were cardiac arrest in five patients, low cardiac output in four patients, and medically intractable pulmonary edema in eighteen patients. Cardiac catheterization was not undertaken in ten persons because of their critical condition. The clinical diagnosis in these patients was supported by noninvasive maneuvers. No surgical deaths occurred. Twenty-five patients are well, active, and NYHA Class I at five to thirty-three months after surgery. There has been one late death and one patient has some residual exertional dyspnea. One patient required reoperation to relieve a clotted prosthetic valve. These results suggest that the patient with aortic valve disease may be offered a reasonable chance for survival, even when desperately ill. References 1. Cohen LS, Friedman WF, Braunwald E: Natural history of mild congenital aortic stenosis elucidated by serial hemodynamic studies. Am J Cardiol30: 1. 1972. 2. Cohn LH. Collins JJ Jr: Local cardiac hypothermia for myocardial protection. Ann Thorac Surg 17: 135, 1974. 3. Okies JE, Bradshaw MW, Williams TW Jr: Valve replacement in bacterial endocarditis. Chest 63: 898, 1973. 4. Wise JR Jr, Cleland WP, Hallidie-Smith KA, et al: Urgent aortic valve replacement for acute aortic regurgitation due to infective endocarditis. Lancet 2: 115, 1971. 5. Neville WE, Magno M, Foxworthy DT, et al: Emergency aortic valve replacement in bacterial endocarditis. J Thorac Cardiovasc Surg 6 1: 9 16, 197 1. 6. Manhas DR. Mohri H, Hessel EA II, et al: Experience with surgical management of primary infective endocarditis. Am Heart J 84: 738, 1972. 7. Stinson EB, Shumway NE: Emergency heart valve replacement. Calif Med 109: 441, 1968. 8. Hutter AM Jr, DeSanctis RW, Nathan MJ, et al: Aortic valve surgery as an emergency procedure. Circulation 41: 623, 1970. 9. Manchester JH, Amsterdam EA, Kemp HG, et al: Coronary artery disease in valvular heart disease (abstract). Circulation 38 (Suppl 6): 132, 1968. 10. Coleman EH, Soloff LA: Incidence of significant coronary ar-
498
tery disease in rheumatic valvular heart disease. Am J Cardiol25: 401, 1970.
Discussion Richard J. Cleveland (Boston, MA): I will direct my comments to those patients with long-standing AS or combined AS-AI. Unfortunately, all of us involved in cardiac surgery have frequently had disappointing longterm results in patients such as these. Most cardiac surgeons have found it possible to operate successfully on such patients; however the patients never really return to the level of activity which we had initially anticipated. There are several elegant studies published and our own clinical experience indicates that when a patient is allowed to reach the level of cardiac disability reported in this series, that irreversible myocardial damage is almost invariably present. This myocardial damage obviously adversely effects the long-term prognosis of the patient. I believe we should be more aggressive in the surgical management of aortic valve disease, and procrastination for any reason for a prolonged period of time after the onset of symptoms is not in a patient’s best interest. It is of interest that so many of these patients reported by the authors, who were essentially moribund, have returned to NYHA Class I. I think this is a bit unusual and does not reflect the general experience of others. It is incumbent upon the surgeon to continue to encourage medical colleagues to refer patients such as these much earlier in their clinical course. The surgeon will operate at a lower mortality and just as importantly, the patient will be restored to a meaningful existence after surgical correction. John H. Sanders, Jr (closing): Actually, we do not have as much hemodynamic information as we would like. Twenty-two of these patients did have evidence of significant cardiomegaly on chest x-ray examination, but most of these did not have ventriculograms. The diastolic volumes and ejection fractions are, therefore, difficult to assess. (Slide) The x-ray film on the left is of an individual in this series who had fairly marked cardiomegaly preoperatively. In film taken one month postoperatively, the heart size is diminished, the effusions are gone and the pulmonary vascular redistribution has begun to improve. It has been our experience that the patients who have had ventriculography have had ejection fractions of 20 to 30 per cent. These patients have done well in our experience. One speculation is that although the ejection fraction may be 20 per cent, many of these patients can generate fairly substantial intraventricular pressures, in the range of 180 mm Hg. The fact that a dilated ventricle is capable of generating a high pressure, even with a small ejection fraction, has led us to suspect that the ventricular function will improve when the gradient is removed.
