Therapy and Prevention Valvular Heart Disease

Between 1961 and 1984, 91 patients underwent simultaneous triple valve replacement at the Mayo Clinic. Of the 273 prosthetic valves used, 77% were Starr-Edwards. Perioperative (30 day) mortality was 24% to 27% between 1962 and 1974 and 7% between 1975 and 1983 (p = .17). In patients with NYHA class IV symptoms, perioperative mortality was 44%, and in those with milder symptoms, it was 8% (p < .0001). The median follow-up was 7.5 years (range, 6 weeks to 20 years). Cumulative survival, which was calculated taking into consideration perioperative mortality, was 64% at 1 year, 55% at 5 years, 40% at 10 years, and 25% at 15 years. Multivariate analysis identified preoperative functional class and age as predictors of late survival. Among causes of late mortality were sudden death in 32.5%, congestive heart failure in 15%, thromboembolism in 12.5%, prosthetic valve dysfunction in 7.5%, and infective endocarditis in 5%. Late complications included systemic emboli in 42% (embolic rate, 12.3 events per 100 patient-years), bleeding in 22%, myocardial infarction in 16%, and infective endocardititis in 6%. Eight patients required reoperation for prosthetic valve dysfunction, and 12 patients had permanent pacemakers. Ofthe 29 patients still alive, 79% are in NYHA class I or II. In summary, perioperative mortality after triple valve replacement appears to be declining; long-term survival in 30 day survivors is similar to that after single valve replacement and excellent symptomatic improvement can be obtained, although morbidity is high. Circulation 72, No. 1, 130-137, 1985. OVER the past 10 years, reports from many centers have defined the early and late results of cardiac valve replacement for most categories of patients. '-9 The outcome of triple valve replacement (aortic, mitral, and tricuspid valve replacement) is less well defined, primarily because of the relatively small numbers of patients who have undergone the operation. With the trend toward conservative tricuspid valve surgery coupled with the development of more predictable methods of valve repair and the declining incidence of rheumatic heart disease in the United States, triple valve replacement is less frequently required than it was previously. Nonetheless, rheumatic heart disease with involvement of multiple valves remains a serious problem in many countries.'0 Accordingly, we report our experience in 91 consecutive patients with rheumatic heart disease who underwent simultaneous triple valve From the Mayo Clinic and Mayo Foundation, Rochester, MN. Address for correspondence: Bernard J. Gersh, M.B.,Ch.B., D.Phil., Mayo Clinic, 200 First St. SW, Rochester, MN 55905. Received Nov. 12, 1984; revision accepted March 22, 1985. Presented in part at the 32nd Annual Scientific Session, American College of Cardiology, New Orleans, March 20 to 24, 1983. 130 replacement at the Mayo Clinic. This is the largest reported series and is particularly notable for the long follow-up period, which spanned 20 years (median duration of follow-up, 7.5 years in perioperative survivors). During this period, major changes in the preoperative, intraoperative, and postoperative practice of valve surgery have occurred. Particular attention has been focused on factors influencing operative risk and late survival, on the causes of late death and late complications, and on the accumulated knowledge derived from this prolonged learning experience. Our earlier experience with triple valve replacement has been reported. 1"' 12 Material and methods We reviewed the records of all 91 patients with rheumatic heart disease who underwent aortic, mitral, and tricuspid valve replacement during a single operation at the Mayo Clinic from 1962 through 1984. For perspective, this cohort represents only 1% of the total 8359 patients who underwent cardiac valve replacement during the same period. At our institution, the frequency of triple valve replacement has progressively declined: 78 operations were perforned from 1962 through 1974, whereas only 14 triple valve replacements were done on 13 CIRCULATION by gest on A ril 0, 2017 http://ciajournals.org/ D ow nladed from THERAPY AND PREVENTION-VALVULAR HEART DISEASE patients from 1975 through 1984. One patient had undergone triple-valve replacement in 1970, but all three cloth-covered Starr-Edwards valves were replaced in 1984 because of hemolysis. Clinical data on the 91 patients are summarized in table 1. The severity of preoperative symptoms appears to have changed. Between 1962 and 1974, 50% of patients were in New York Heart Association (NYHA) class IV before surgery and 50% were in NYHA class II or III; of those undergoing operation from 1975 through 1984, only 8% were in NYHA class IV and 92% were in class II or III (p = .024). The predominant valve lesion was classified according to clinical findings, preoperative and intraoperative intracardiac pressure recordings, and operative inspection (table 2). Preoperatively 86% of patients were in atrial fibrillation, one patient was in paced rhythm, and one patient had episodic sinus arrest. The remaining patients (14%) were in sinus rhythm. All patients had rheumatic heart disease, and 23 patients had 35 previous cardiac operations. Eight prior operations involved the aortic valve, 24 involved the mitral valve (21 mitral commissurotomies), and three were performed for tricuspid valve disease. Four patients had undergone previous valve replacements (including the patient with a prior triple valve replacement). During the interval of this study, a variety of prosthetic heart valves were used, and these are listed in table 3. Most (77%) were Starr-Edwards prostheses. Surgical techniques. Methods of extracorporeal circulation and surgical techniques evolved throughout the two decades of this study. All operations were performed through a median sternotomy, and high-flow (2.0 to 2.4 liters/min/m2) hypothermic extracorporeal circulation was used. For the first 84 patients in this series, myocardial protection was achieved by continuous coronary artery perfusion; in the last seven patients, cold potasrium cardioplegia was used during aortic cross-clamping. Anticoagulation with sodium warfarin was begun in all patients as soon as possible after operation, generally on the second or third postoperative day. Follow-up. Current information on all patients was obtained by questionnaires to both patients and referring physicians, by telephone contact, or, in patients who had been seen before the end of the present review, by Mayo Clinic records. If possible, the cause of death was ascertained from copies of death certificates or from information supplied by the patient's physician or family. The diagnosis of thromboembolism was made by a history of stroke, transient cerebrovascular ischemic attacks, or myocardial infarction in patients younger than 45 years or in those with documented normal coronary arteries. Splenic, renal, and pulmonary emboli were diagnosed in a few patients in whom adequate documentation was present, including those in whom a diagnosis was made at autopsy. All patient contacts included specific questions related to possible thromboembolic episodes. TABLE 1 Preoperative characteristics of patients