Cardiovascular disease is the primary cause of morbidity

Transcription

1 Valve Replacement in Patients on Chronic Renal Dialysis: Implications for Valve Prosthesis Selection William T. Brinkman, MD, Willis H. Williams, MD, Robert A. Guyton, MD, Ellis L. Jones, MD, and Joseph M. Craver, MD Division of Cardiothoracic Surgery, Joseph B. Whitehead Department of Surgery, Emory University School of Medicine, Atlanta, Georgia Cardiovascular disease is the primary cause of morbidity and mortality in patients requiring renal dialysis [1]. In 1998 the incidence (number of new cases reported) of end-stage renal disease (ESRD) in the United States was 86,825; prevalence (number of current cases) was 372,407. For the year 2010, projected incidence is 172,667; projected prevalence is 661,330 [1]. Chronic uremia, hypertension, volume overload, dyslipidemia, and anemia all associated with ESRD predispose patients to cardiac valvular abnormalities. Secondary hyperparathyroidism and the resultant increased calcium phosphate product accelerate calcification of cardiac valves and specialized conduction tissue [2]. Calcific degenerative cardiac valve disease predisposes to bacterial endocarditis in the presence of frequent vascular access-related bacteremia and compromised immunity associated with chronic renal failure [3]. Since the first valve replacement in a patient with ESRD in 1966 [4], debate has continued regarding technical and ethical aspects of valve replacement in these chronically ill and presumably high-risk patients anticipated to have a relatively short life expectancy. Most Presented at the Thirty-eighth Annual Meeting of The Society of Thoracic Surgeons, Fort Lauderdale, FL, Jan 28 30, Address reprint request to Dr Craver, The Emory Clinic, 1365 Clifton Rd, Atlanta, GA 30322; christy_jenkins@emoryhealthcare.org. Background. Reports are sparse describing heart valve replacement in patients with end-stage renal disease. This review assesses a 15-year experience and outcomes after valve replacement in patients on chronic preoperative renal dialysis. Methods. A computerized database, hospital records, and telephone contact provided outcome data for patients on chronic dialysis undergoing valve replacement between March 22, 1985, and October 13, 2000, in two hospitals. Results. Seventy-two patients underwent 95 valve procedures (74 operations). Ages ranged from 23 years to 84 years (mean, 57 years). Fifty-five aortic, 30 mitral, and 3 tricuspid valve replacements and 7 valvuloplasties were performed. Six of the 74 procedures were reoperative valve replacements. In the 46 patients with reliable long-term (greater than 30 days) follow-up data, significant bleeding or stroke was documented in 17 of 34 patients with a mechanical valve and 1 of 12 patients with a bioprosthetic valve. Overall survival (including two operative deaths) was 72.8% at 3 months, 65.4% at 6 months, 60.5% at 1 year, 39.8% at 2 years, 28.5% at 3 years, and 15.9% at 6 years (Kaplan-Meier). Type of valve implanted did not influence early and late survival. Conclusions. In this series of patients on chronic dialysis, survival appears to justify valve replacement. However, the sixfold higher incidence of late bleeding or stroke in patients on dialysis with a mechanical valve requiring warfarin suggests that bioprosthetic valves are the valve substitute of choice in patients on chronic dialysis. (Ann Thorac Surg 2002;74:37 42) 2002 by The Society of Thoracic Surgeons surgeons have believed that mechanical valves are superior to bioprosthetic valves in the setting of chronic renal failure because of accelerated bioprosthesis calcification and structural degeneration [5]. Supportive evidence is largely anecdotal [6 8]; recent reviews have cast doubt on previous assertions [9, 10]. Limited data, lack of consensus, and increasing recent and future referral of patients on chronic renal dialysis but needing cardiac valve operations prompted this review of our experience. Patients and Methods Seventy-two patients requiring chronic renal dialysis at the time of operation underwent cardiac valve replacement at Emory University Hospital or Crawford Long Hospital of Emory University between March 1985 and October Patients with acute renal failure not on chronic renal dialysis were excluded from this review. Information was obtained from pertinent medical records, the concurrently compiled computerized Emory University Cardiac Database, governmental vital statistics registries, and telephone follow-up with patients or their families. The decision regarding type of valve placed was made primarily on the basis of the expected survival of the patient. In general, patients who received a mechanical valve were deemed to have a probable survival greater than 5 years. Those with expected survivals less than by The Society of Thoracic Surgeons /02/$22.00 Published by Elsevier Science Inc PII S (02)

2 38 BRINKMAN ET AL Ann Thorac Surg VALVE REPLACEMENT ON CHRONIC DIALYSIS 2002;74:37 42 years were considered primarily bioprosthetic candidates. Other factors such as an inability to tolerate warfarin anticoagulation and individual surgeon anecdotal experience also affected valve selection. Discharge anticoagulation management was documented for all 72 patients. During the course of the study (1986 to 2001) the target international normalized ratio for patients with a mechanical valve varied according to surgeon, type and location of prosthesis implanted, and presence of atrial fibrillation. Patients receiving bioprosthetic valves in either position were generally not given anticoagulants unless also in atrial fibrillation, with a target international normalized ratio of 2.3 to 2.6 U. Patients with mechanical valves in the aortic position were maintained with an international normalized ratio of 2.5 U. Any patient with a mechanical valve in the mitral or tricuspid position was maintained at an international normalized ratio of 3.3 to 3.5 U. Patients who received a mechanical valve in the aortic position plus an additional valve were maintained at an international normalized ratio target level of 3.0 to 3.5 U. Stroke was defined as a central nervous system deficit documented by a physician or evidence of a new hemorrhagic or embolic stroke identified by computerized tomography or magnetic resonance imaging of the brain. Both early stroke (occurring within 30 days of valve replacement) and late stroke (occurring more than 30 days after valve replacement) were considered. Major late bleeding (occurring more than 30 days after valve replacement) was defined as bleeding from any site requiring hospital admission or transfusion of at least 2 U of packed red blood cells. Complete follow-up was obtained for survival. Late complications were documented wherever possible without bias to the type of valve substitute implanted. An extensive effort was made to completely assess patient follow-up. Contact of dialysis centers was attempted, but was not consistently helpful. This population of patients was especially difficult to follow owing to several factors: (1) a transient population, (2) low socioeconomic status in general, and (3) poor compliance with dialysis and medical services. Patients in whom there were periods of time that the medical history could not be confirmed were not used for statistical analysis of nonfatal outcomes. Data were compiled and analyzed using Microsoft Access, Microsoft (Access and Excel) (Redmond, WA), StatView (Cary, NC), and S-Plus (Seattle, WA). Continuous data are presented as the range and, in parentheses, the mean standard deviation. When data were skewed, the median and interquartile range (IQR) are presented. Survival estimates were calculated using the Kaplan- Meier method. Dichotomous data were compared using Fisher s exact test with differences at p less than 0.05 considered significant. Results Patients ranged in age from 23 years to 84 years (mean, years). Patients receiving bioprosthetic valve substitutes ranged in age from 35 years to 84 years (mean, years), whereas patients receiving mechanical valves were younger, their ages ranging from 23 years to 76 years (mean, years; p 0.005). Valve replacements were performed in 38 men (53%) and 34 women (47%). Thirty-seven patients were African Americans (51%), 31 were white (43%), and 4 were racially unidentified (6%). Date of initial dialysis was available for 50 of the 72 patients (69%). Duration of preoperative dialysis for these 50 patients ranged from 9 days to 19.6 years (mean, years) with no significant difference between those patients receiving mechanical valves compared with those receiving bioprosthetic valves. Preoperative patient characteristics are summarized in Table 1. Causes of cardiac valve failure are listed in Table 2. For the 18 patients having bacterial endocarditisrelated disease, pathogens reported included Staphylococcus aureus (7 patients), coagulase-negative staphylococci (2 patients), Enterococcus sp. (1 patient), and Candida parapsilosis (1 patient). Bacteriologic data were not available for 7 patients. Ninety-five valve procedures, including 88 valve replacements, 3 valvuloplasties with ring annuloplasty, and 4 valvuloplasties without ring placement, were performed in 74 operations in 72 patients. Fifty-eight mechanical valves (66%) and 30 bioprosthetic valves (34%) were implanted. Forty-three were isolated aortic valve replacements (58%); 17 were isolated mitral valve replacements (23%); 11 were combined aortic and mitral replacements (15%); 2 were combined tricuspid and mitral replacements (3%); and 1 was a combined tricuspid and aortic valve replacement (1%). Concomitant coronary artery bypass grafting was performed in 23 of the 74 operations (31%). In these 23 patients, the number of distal anastomoses ranged from one to five (mean, anastomoses; median, 1.0; IQR, 2.0). Those patients who received a bioprosthetic valve had more grafts than those receiving a mechanical valve (mean, ; median, 2.0; IQR, 2.0 versus mean, ; median, 1.0; IQR, 1.0; not significant). The aortic occlusion time ranged from 34 minutes to 225 minutes (mean, minutes; median, 85 minutes; IQR, 49 minutes). Time on cardiopulmonary bypass ranged from 50 minutes to 259 minutes (mean, minutes; median, 141 minutes; IQR, 61 minutes). Length of hospital stay for survivors of the operation ranged from 4 to 85 days (mean, 13.5 days; median, 9 days; IQR, 9.0 days). In-hospital postoperative complications included stroke in 5 patients; bleeding requiring exploration in 4; sepsis in 4; heart block requiring permanent pacemaker placement in 4; mediastinitis in 3; small bowel obstruction with ischemic bowel in 2; deep venous thrombosis in 2; myocardial infarction in 1; and pericarditis with effusion and tamponade requiring open pericardial fenestration in 1 patient. Operative mortality includes both (1) all deaths occurring during the hospitalization in which the operation was performed, even if after 30 days, and (2) those deaths occurring after discharge from the hospital, but within 30 days of the procedure... [11]. Two patients died on the day of operation, one in the operating room and the other a few hours later in the intensive care unit. Both cases involved patients with sepsis related to endocarditis

3 Ann Thorac Surg BRINKMAN ET AL 2002;74:37 42 VALVE REPLACEMENT ON CHRONIC DIALYSIS 39 Table 1. Preoperative Patient Characteristics Preoperative Characteristics All Patients Bioprosthetic Valves Mechanical Valves N 72 % N 29 % N 43 % p Value a NYHA class I II NS NYHA class III IV NYHA unknown Hypertension NS Smoking history NS Diabetes mellitus NS Age 60 years Previous myocardial infarction NS Previous open heart operation NS Previous coronary artery bypass NS grafting Previous angioplasty NS Previous stroke Cardiac arrhythmia NS Previous gastrointestinal bleed NS Renal transplant with rejection NS Obstructive pulmonary disease NS a By Fisher s exact test. Abbreviations: NS not significant; NYHA New York Heart Association. requiring intravenous pressor support and aggressive preoperative intensive care unit care. Ten other patients died later during the same hospitalization. Death occurred from 4 to 45 days postoperatively (mean, days; median, 9.0 days; IQR, 7.0 days). Sixty-two patients were discharged alive from the hospital, none of whom subsequently died within 30 days of operation. Accordingly, operative survival was 62 of 74 patients (84%). Operative survival, as previously defined, was 88% (45 of 51 patients) for isolated valve replacement and 78% (18 of 23 patients) for valve replacement with concomitant coronary artery bypass grafting (not significant). Operative survival for those patients receiving a mechanical valve substitute (38 of 45, 84%) was not different from those patients receiving a bioprosthetic implant (25 of 29, 86%; not significant). Late survival (Kaplan-Meier) [12] including deaths in the operating room, on the day of operation, in the hospital, or within the follow-up period of this review was 72.8% at 3 months, 65.4% at 6 months, 60.5% at 12 months, 39.8% at 24 months, 28.5% at 36 months, and 15.9% at 72 months (Fig 1). Neither type of valve implanted, sex, race, nor the presence of concomitant coronary artery bypass grafting was shown to have an Table 2. Causes of Valve Failure Cause of Valve Failure 74 Operations Percent Calcific degeneration Endocarditis Rheumatic Prolapse/myxomatous 5 7 Congenital 2 3 Aortic dissection 1 1 Ischemic dysfunction 1 1 Prosthetic dysfunction 1 1 Not reported 7 9 influence on survival in this small group of 72 patients having 74 operations. A trend toward improved survival for patients younger than 60 years at time of operation was noted, but statistical significance was not achieved. Many of these patients were returned to the care of their distant referring physician or cardiac care system after hospital discharge.accordingly, complete long-term follow-up data documenting complications other than death were available in only 46 (64%) of the 72 patients, of whom 34 had received a mechanical valve (74%) and 12 had received a prosthetic valve (26%). Outcomes data were obtained for all 72 patients for the first 30 postoperative days. There were five documented early strokes in the 72 patients (6.8%), two in the 45 patients with mechanical valves (4.4%) and three in the 29 patients with bioprosthetic valves (10.3%, not significant). None of the patients having an early (within 30 days) new stroke had reported a stroke before valve replacement. There were seven Fig 1. Kaplan-Meier survival after operation (solid line) with 95% confidence limits (dotted lines) above and below.

4 40 BRINKMAN ET AL Ann Thorac Surg VALVE REPLACEMENT ON CHRONIC DIALYSIS 2002;74:37 42 documented late ( 30 days) new strokes in the 46 patients for whom long-term follow-up data were obtainable (15.2%), six in 34 patients with mechanical valves (17.6%) and one in the 12 patients with a bioprosthetic valve (8.3%, not significant). Of patients having a late new stroke, 3 (all of whom received a mechanical valve) had reported a preoperative stroke. The 1 patient with a bioprosthetic valve who had a late stroke had been discharged only on aspirin for anticoagulation. A definitive diagnosis of embolic versus hemorrhagic stroke could not be made from available data in most instances and so was not included in this review. There were 14 significant late bleeding episodes in the 46 patients for whom long-term follow-up data were available (30.4%), 13 in the 34 patients having mechanical valves (38.2%) and 1 in the 12 patients with a bioprosthetic valve (8.3%, p 0.07). All of the 13 patients with a mechanical valve who subsequently experienced a late bleed had been maintained on warfarin after valve replacement. The one patient with a bioprosthesis who experienced a late bleed had been maintained on aspirin after valve replacement. Time lapse to documented late bleeding ranged from 1.1 months to 7.4 years (mean, months; median, 10.5 months; IQR, 22.9 months) Four documented episodes of postoperative bacterial endocarditis were treated medically. Two of these 4 patients eventually died of endocarditis-related complications. Valve replacement had initially been indicated in these 4 patients by calcific degeneration, bacterial endocarditis, rheumatic disease, and myxomatous degeneration. Three early sternal wound infections were documented. One of these 3 patients died in the hospital of sepsis, pneumonitis, and multisystem organ failure. Mediastinitis was successfully treated in 2 of the 3 patients with debridement and muscle flap closure. During the secondary valve replacements, one mechanical and five bioprosthetic valves were excised. Four of the excised prosthetic valves were implanted before initiation of chronic renal dialysis and hence are not included in this review. The other two excised valves, a Hancock alpha amino oleic acid treated porcine bioprosthesis in the aortic position (originally implanted in a 67-year-old man) and a Carpentier-Edwards porcine bioprosthesis (originally implanted in a 53-year-old woman) in the mitral position, were placed after initiation of chronic renal dialysis and are included in this review. These two valves failed owing to structural dysfunction 15.2 months and 54.6 months after implantation, respectively, yielding a bioprosthetic valve excision rate of 2 excisions per 30 bioprosthetic implantations (7%). No mechanical prosthetic valves implanted in patients on chronic renal dialysis at the time of implantation required excision. Comment Although life expectancy for patients with ESRD has gradually improved in the United States, mortality consistently exceeds 25% per year [13]. Four-year survival of patients on hemodialysis or peritoneal dialysis is approximately 40%. Duration of preoperative dialysis for 50 patients in this review for whom data were available ranged from 9 days to 19.6 years (mean, years; median, 3.0 years; IQR, 4.54 years). Accordingly, even if these patients had not required heart valve replacement, anticipated 4-year survival would have been less than 40%. Cardiovascular disease is the most common cause of death in patients requiring chronic renal dialysis [1]. Of the patients in this review, at least 69% were in congestive heart failure (New York Heart Association class III or IV) before operation, and 24% reported a previous myocardial infarction. Significant cardiac arrhythmias had occurred in at least 18% and a previous stroke in 15% (Table 1). Stroke, the most common early postoperative complication, occurred in 5 patients, two in patients who received mechanical valves and three in patients who received bioprosthetic valves. There were seven additional late strokes, six in patients who received mechanical valves and one in a patient who received a bioprosthetic valve. Lucke and associates [10], in their retrospective review of 18 patients requiring preoperative chronic renal dialysis before cardiac valve replacement, reported seven postoperative strokes, all of which occurred in the 10 patients receiving mechanical valves. Kaplon and colleagues [9], in a review of 42 similar patients, reported only three strokes, two in patients with mechanical valves and one in a patient with a bioprosthetic valve. It should be noted that the patients who received a mechanical valve in this study had a higher preoperative stroke rate (p 0.054), and the patients receiving bioprostheses were older (p 0.041). We, however, did not observe differences in preoperative rates of systemic hypertension, smoking history, diabetes mellitus, endocarditis, or cardiac arrhythmias. None of the preoperative strokes in this review were associated with major morbidities (ie, persistent significant loss of motor function or cognitive function). Bacterial endocarditis was the primary indication for operation in 18 of the 74 operations in this review. End-stage renal disease and diabetes mellitus are known to predispose to endocarditis of both normal valves and structurally abnormal valves [14]. Mechanical valves are at higher risk for infection than are bioprostheses within the first 5 years after implantation. In the environment of infectious endocarditis the choice of valve type is not absolute but does favor the bioprosthetic at least in the first 5 years, a potentially important consideration in these patients with higher risk of endocarditis and limited anticipated survival [15]. When both stroke and significant bleeding are considered together as a composite morbid end point, statistical inference favors the use of a bioprosthetic rather than a mechanical valve. Of the 46 patients with adequate follow-up, only 1 of 12 patients with a bioprosthesis had a bleed or a stroke (8.3%); 17 of 34 patients with a mechanical valve had either a bleed or a stroke (50%; p 0.016, Fisher s exact test for differences; odds ratio, 6.02). In this review the overall operative mortality rate according to The Society of Thoracic Surgeons National

5 Ann Thorac Surg BRINKMAN ET AL 2002;74:37 42 VALVE REPLACEMENT ON CHRONIC DIALYSIS 41 Table 3. Causes of Death Cause N 45 % Cardiac failure Infectious Stroke 3 7 Hepatic failure Renal failure Bleeding Head trauma Complications of diabetes mellitus Multisystem organ failure Unknown 7 16 Adult Cardiac Database definition [11] was 12 of 74 patients (16.2%). Lucke and coworkers [[10] reported a 30-day mortality rate of 15.8% in patients on chronic renal dialysis receiving an isolated valve replacement. Horst and associates [16], in a meta-analysis of patients with ESRD undergoing cardiac operation, reported a 19.3% mortality for isolated valve replacement and a 39.5% mortality for valve replacement with concomitant coronary artery bypass grafting. In our review operative mortality was 6 of 52 patients (12%) for patients undergoing isolated valve replacement and 5 of 23 patients (22%) for those having valve replacement with concomitant coronary artery bypass grafting. Hertzog and colleagues [17] in their recent review of survival after valve replacement in patients on renal dialysis using the US Renal Data System database also confirmed our findings. Their analysis involving 5,858 patients reported an inhospital mortality rate of 20.7%. In addition, using a statistical model Edwards and coworkers [18] predicted an operative mortality of 17.1% after valve replacement operation in dialysis patients. The relatively low 2-year survival of 40% of patients in this review is consistent with previous reports [9, 10, 16]. Causes of death were primarily cardiovascular events (Table 3). However, considered in the context of a 33% 2-year survival for patients with congestive heart failure and ESRD requiring chronic dialysis alone, heart valve replacement seems reasonable in appropriately selected patients in whom important symptomatic improvement and even temporarily enhanced quality of life might be anticipated [19]. Although we found no important differences in survival relative to the type of valve implanted, our experience was consistent with similar reports in patients with ESRD [9, 10, 17]. Accordingly, factors other than mortality must be considered when choosing between a mechanical valve and a bioprosthetic valve for implantation in a patient requiring chronic renal dialysis. Patients with ESRD have typically been referred for cardiac valve replacement only in extremis, when life is acutely threatened by infectious endocarditis or ventricular function has deteriorated to levels that threaten salvage. The fact that more than two thirds of patients referred for valve replacement were suffering from advanced congestive heart failure (New York Heart Association class III or IV) suggests that physicians may Fig 2. Survival after valve replacement, stratified by type of valve implanted. No influence of valve type on survival was demonstrated in this review. consider referral for valve replacement as a last resort. Indeed, such behavior is consistent with the highly conservative referral philosophy that dominated the early years of heart valve replacement operations when reported mortality and morbidity were both exceedingly high among those end-stage patients who were then being referred for operation. Perhaps earlier referral for cardiac valve repair or replacement while it can still be performed under elective rather than emergent circumstances might offer these patients with ESRD as for other patients an increased probability of early and even late survival, lower morbidity, and a somewhat more tolerable quality of life [20]. These data demonstrate more major morbid events including stroke and significant late bleeding in patients on chronic renal dialysis who have received mechanical valves requiring anticoagulation with warfarin. Type of valve implanted did not influence survival (Fig 2). Postoperative survival of these chronically ill patients with multisystem disease was limited; by 2 years more than half of the patients had died. Explantation of a bioprosthetic valve was required in only 2 of 30 patients (6.7%). Modern bioprosthetic and homograft valve fabrication technology, chemically induced resistance to degenerative mineralization, improved management of secondary hyperparathyroidism, and evolving renal dialysis techniques are anticipated to reduce early structural bioprosthesis failure. These trends and the survival and morbidity data from this retrospective review support the bioprosthesis as the valve substitute of choice in patients who require both cardiac valve replacement and chronic renal dialysis. References 1. US Renal Data System Annual Data Report. National Institutes of Health, National Institute of Diabetes and Digestive and Kidney Diseases 2001; adr.htm. 2. Jain M, D Cruz I, Kapthalia S, Goldberg A. Mitral annulus calcification as a manifestation of secondary hyperparathy-

6 42 BRINKMAN ET AL Ann Thorac Surg VALVE REPLACEMENT ON CHRONIC DIALYSIS 2002;74:37 42 roidism in chronic renal failure. Circulation 1980;62(Suppl): Goldblum S, Reed W. Host defenses and immunologic alterations associated with chronic hemodialysis. Ann Intern Med 1980;93: Lansing A, Leb D, Berman L. Cardiovascular surgery in end-stage renal failure. JAMA 1968;204: Mehta S, Pae WE. Complications of cardiac surgery. In: Edmunds LH, ed. Cardiac surgery in the adult. New York: McGraw-Hill, 1997: Ko W, Krieger KH, Isom W. Cardiopulmonary bypass procedures in dialysis patients. Ann Thorac Surg 1993;55: Monson BK, Wickstrom PH, Haglin JJ, Francis G, Comty CM, Helseth MK. Cardiac operation and end stage renal disease. Ann Thorac Surg 1980;30: Lamberti JJ, Wainer BH, Fisher KA, Karunaratne HB, Al- Sadir J. Calcific stenosis of the porcine heterograft. Ann Thorac Surg 1978;28: Kaplon RJ, Cosgrove DM, Gillinov M, Lytle BW, Blackstone EH, Smedira NG. Cardiac valve replacement in patients on dialysis: influences of prosthesis on survival. Ann Thorac Surg 2000;70: Lucke JC, Samy RN, Atkins BZ, et al. Results of valve replacement with mechanical and biological prostheses in chronic renal dialysis patients. Ann Thorac Surg 1997;64: Society of Thoracic Surgeons. Adult Cardiac Database Core Data Elements Summary, Version : Data analysis products division, Estimating survival. In: MathSoft S-Plus 2000 guide to statistics, volume 2. Seattle: MathSoft, Inc., 1999: Pastan S, Bailey J. Dialysis therapy. N Engl J Med 1998;338: Strom BL, Abrutyn E, Berlin JA, et al. Risk factors for infective endocarditis: oral hygiene and nondental exposures. Circulation 2000;102: Mylonakis E, Calderwood SB. Infective endocarditis in adults. N Engl J Med 2001;345: Horst M, Melhorn U, Hoerstrup SP, Suedkamp M, Ranier de Vivie E. Cardiac surgery in patients with end-stage renal disease: 10-year experience. Ann Thorac Surg 2000;69: Herzog CA, Ma JZ, Collins AJ. Long-term survival of dialysis patients in the United States with prosthetic heart valves: should the ACC/AHA practice guidelines on valve selection be modified? Circulation 2002;105: Edwards FH, Peterson ED, Koombs LP, et al. Prediction of operative mortality after valve replacement surgery. Am J Coll Cardiol 2001;37: Parfrey PS, Griffiths SM, Harnett JD, et al. Outcome of congestive heart failure, dilated cardiomyopathy, hypertrophic hyperkinetic disease, and ischemic heart disease in dialysis patients. Am J Nephrol 1990;10: Christiansen S, Claus M, Phillipp T, Reidemeister JC. Cardiac surgery in patients with end-stage renal failure. Clin Nephrol 1997;48: DISCUSSION DR KEVIN A. ACCOLA (Orlando, FL): That was an excellent presentation. We also agree with your conclusions in regard to the utilization of bioprosthetic valves in chronic renal dialysis patients because of the complications and risks associated with anticoagulation in this population. My questions to you are as follows: 1. How do you deal with patients who are in atrial fibrillation? After analyzing your results, does the presence of atrial fibrillation alter your approach to these patients in regard to valve selection? 2. My second question pertains to anticoagulation of mitral valves. At Emory do you anticoagulate bioprosthetic valves or are you just treating them with antiplatelet agents? DR BRINKMAN: Thank you for those good questions. Regarding atrial fibrillation, there were three patients in the bioprosthetic valve group who had dilated left atria and abnormalities of atrial function that required Coumadin use. The others were not on Coumadin, just an occasional antiplatelet agent such as aspirin. As far as routine anticoagulation for the mitral valve, in this population it was not done, it was not the policy at Emory. DR ARTHUR J. CRUMBLEY III (Charleston, SC): With structural valve deterioration of bioprosthetic valves from calcification being a real concern in these patients, have you considered looking at parathormone levels in them routinely, even if they do not have clinically evident hyperparathyroidism, so that perhaps subclinical levels would affect a bioprosthetic valve when it might not become otherwise clinically evident? DR BRINKMAN: That is a very good question. Secondary hyperparathyroidism is a severe problem with people with end-stage renal disease, and we did not look into parathormone values in these patients, although that might be a good way to see how well patients are being maintained on dialysis and if they may require a parathyroidectomy in the future. DR HORMAZ AZAR (Norfolk, VA): I enjoyed your paper. Did you try to analyze patient survival based on the location of the valve, in other words, aortic versus the mitral group? I know they were small numbers, but did you get a sense of that at all, particularly as the mitral regurgitation patients on dialysis could be handled with a little tighter dialysis regime in most cases? DR BRINKMAN: We did not break down the survival between the aortic and mitral valve groups because of the small patient numbers for that reason. So I cannot give you a good answer on that. But we are continuing to accumulate patients in this database, and we will probably do that in the future. DR MARK MARBEY (Kalamazoo, MI): I occasionally see younger patients who are being listed for transplantation to fix heart problems before their renal transplantation, so I am wondering whether that would change your strategy at all. If you had a younger patient in whom you might otherwise be using a mechanical valve in, if he did not have the renal failure problems, who is going to be getting a transplant presumably within the next year or so, would that change your strategy? DR BRINKMAN: That is a very good question. We specifically did not address that issue. In a person with an expected good long-term survival, such as a young person getting a renal transplant, you would need to consider the morbidity of a repeat valve operation. This might be an exception to our review presented here. Because his physiology would be changed at a young age, he might be a better candidate for a mechanical valve, but again that may be an exception.