Abnormal calcium homeostasis in patients with end-stage renal failure results in dystrophic calcification. Valvular and perivalvular involvement in ESRD is most commonly manifested as mitral annular calcification and aortic valve calcification[10]. Both mitral and aortic valve calcification occur more frequently and at younger age in those with ESRD than in those with normal renal function[11]. Mitral valve and aortic valve replacement are indicated for severe symptomatic valve stenosis or regurgitation[11].
Valve replacement in hemodialysis patients presents a dilemma to the cardiac surgeon, who must balance the potential for accelerated prosthetic valve deterioration against the morbidity of anticoagulation when selecting a prosthesis[3, 5, 12]. There is much controversy regarding prosthetic valve selection in patients with ESRD requiring dialysis[6, 13, 14]. Like other studies[5, 9], we found accelerated calcification of bioprosthesis to be uncommon in patients on preoperative dialysis. Importantly, Kaplon et al. found no convincing evidence for accelerated calcification as a major cause of bioprosthetic valve failure and resultant adverse morbidity and mortality[10]. In fact, the limited sample size of this study probably does not prevent it from yielding a meaning conclusion that bioprosthesis could be a preferable choice for ESRD.
Given the lack of a clinical difference in morbidity or survival following either mechanical or bioprosthetic valve replacement, we believe that concern for accelerated calcific bioprosthesis degeneration should not play a role when choosing a valve for patients on dialysis. Rather, other variable such as patient age, gender, level of activity, and presence of infection should dictate valve selection, and not the diagosis of end-stage renal failure. We, however, did not observe differences in preoperative rates of systemic hypertension, smoking history, diabetes mellitus, endocarditis, or cardiac arrhythmias. In this study, therefore, we investigated the risk factors of hospital mortality in dialysis-dependent patients after undergoing cardiac surgery. Univariate analysis showed that preoperative lower LVEF was important independent predictors of hospital mortality. In addition, Kaplan-Meier analysis further accentuated the unacceptably high rates of complications and death with mechanical valves. Recent reports have demonstrated the effective use of tissue valves in dialysis patients, without increased mortality or reoperation compared to mechanical valves[13, 15].
Although life expectancy for patients with ESRD has gradually improved in the United States, mortality consistently exceeds 25% per year. Four-year survival of patients on hemodialysis or peritoneal dialysis is approximately 40%[16]. In the present study, the overall survival rate of dialysis patients after isolated valve replacement of 85.2% at 3 years and 55.9% at 5 years is clearly better than in previous reports[17]. This might be due to differences in concomitant procedures and postoperative anticoagulation therapy for those with mechanical valves. Although concomitant CABG was the independent predictors of hospital mortality and survival in the previous reports[5–7, 9], we excluded patients with CABG because we tried to clarify the long-term result of isolated valve replacement. Our analysis showed no significant difference in life expectancy or rate of reoperation after follow-up 5 years between patients receiving mechanical or biological prostheses, despite the greater proportion of elderly patients in the biological prosthesis group.
It is clear that the overall survival of ESRD patients was poor. In contrast to previous literature[14, 18], differences in survival between patients receiving bioprostheses or mechanical prostheses were related to age at operation and not to prosthesis type. In the present study, we have thus far found no survival difference between bioprostheses and mechanical prostheses patients. And age at operation was found to be no different between the two groups. There was no superiority of freedom from all valve-related complications and individual valve-related complications with mechanical prostheses or bioprostheses. Bioprostheses should not be contraindicated in ESRD patients given the observed rarity of accelerated calcification, as well as the poor intermediate-term survival[18, 19].
Chronic dialysis patients tend to have more hemorrhagic complications. Therefore, dialysis patients undergoing anticoagulation therapy may be at increased risk of these complications[5, 6]. The target value for anticoagulation therapy is generally lower in China than in Europe and the United States: INR is 1.8–2.5 in the case of dialysis patients in China. This is because minor bleeding complications (such as nasal bleeding or bleeding from cannulation sites) often occur among hemodialysis patients receiving warfarin therapy when their INR exceeds 2.5. During the mean follow-up of 44 ± 26 months, there were bleeding complications in 4 of 20 survivors, and no thromboembolic events occurred. Recently, lower-intensity anticoagulation therapy has been demonstrated to result in a lower rate of bleeding complications with bileaflet mechanical valves, without increasing the rate of thromboembolism[17].
The ACC/AHA guidelines were changed in 2006 to reflect the findings of a series of observational studies which showed no significant difference in the freedom from valve-related events including reoperation and late mortality between patients with dialysis-dependent RF receiving a mechanical or a biological prosthesis[20]. In addition to the above findings, a meta-analysis of the published literature demonstrates no survival difference following valve replacement with either bioprosthesis or mechanical prosthesis in patients with ESRD on dialysis. Some recent studies reported that early structural valve deterioration was uncommon, even in dialysis patients, but others showed that it occurred even in new-generation bioprosthetic valves[21, 22]. In fact, these reports also support our current study, which have similar outcomes than the present paper.
Study limitations
The present study was limited by a lack of randomization, as are all studies of valve replacement in this category of population. Furthermore, as no standardized protocols were used with regard to choice of prosthesis, the data presented herein are subject to individual surgeon biases. In addition, our survival analysis is based on a retrospective study design, and it is possible that selection bias for a particular choice of valve may have occurred.