Few studies have evaluated the optimal size of mechanical mitral prostheses. A patient–prosthesis mismatch (PPM) in which the prosthesis is proportionally too small compared with the body surface area (BSA) can result in significant mitral stenosis and poor clinical outcomes.1,2 Therefore, most surgeons prefer to place a larger prosthesis during mitral valve replacement (MVR).
However, a disproportionally large mitral annulus is common in Asia and developing countries where the incidence of advanced rheumatic valve disease is high.3 Additionally, the prevalence of rheumatic valve disease has increased in some countries with a previously low disease prevalence.4 A large mitral annulus is also associated with chronic heart failure, dilated left ventricle, enlarged left atrium, and poor left ventricular function. Occasionally, the mitral annulus is too large to insert a prosthesis consistent with the patient’s BSA (Figure 1).
There are many drawbacks in the use of excessively large mechanical prostheses. Some studies have suggested that a large, rigid structure is related to poorer late outcomes and reduced systolic function.5,6 The most adverse late outcomes related to mechanical prostheses are bleeding and thromboembolism, which are because of anticoagulation therapy and the large mechanical surface area (MSA) of the prosthesis activating the coagulation system.
Although there have been many studies reporting on the size of heart valve prostheses, using effective orifice area (EOA) or geometric orifice area (GOA), there have been no studies on the impact of three-dimensional (3D) MSA on mechanical valves. Because both the EOA and GOA refer to the hemodynamic performance of prosthetic valves, we assumed that the indexed value of 3D MSA may be predictive of the clinical outcome of mechanical valves. Thus, the aim of this study was to investigate the impact of large MSAs on the clinical outcomes of patients who have undergone isolated mechanical MVR.
Materials and Methods
The records of all patients aged >18 years who underwent MVR at Samsung Medical Center between January 1995 and December 2010 were retrospectively reviewed. Patients with acute coronary syndrome and those who underwent multivalve replacement were excluded. The Institutional Review Board of Samsung Medical Center, Sungkyunkwan University School of Medicine, Seoul, Korea, approved this study and waived the requirement for individual consent from patients or relatives.
In this study, the etiology of mitral valve disease was rheumatic heart disease (403 patients, 83%), infective endocarditis (42 patients, 9%), or degenerative valve disease (36 patients, 7%). Mitral valve pathologies included mitral regurgitation without stenosis (195 patients, 40%) and mitral stenosis with or without regurgitation (293 patients, 60%). Forty-five patients (9%) had a history of percutaneous mitral valvotomy, and 34 patients (7%) had undergone previous cardiac surgeries, including mitral valve repair (n = 14, 41%), MVR (n = 10, 30%), and open mitral commissurotomy (n = 8, 24%). The patient characteristics are summarized in Table 1.
Operative Techniques and MSA
MVR was performed using standard cardiopulmonary bypass with bicaval cannulation. The mitral annulus was measured after excision of the anterior mitral leaflet. The posterior mitral apparatus was preserved. Three types of mechanical mitral prosthetic valves were used: ATS (ATS Medical, Inc., Minneapolis, MN), St. Jude Medical (SJM; St. Jude Medical Inc., St. Paul, MN), and On-X (On-X Life Technologies Inc., Austin, TX). We preferred the On-X 25 mm valve for the patients with a small mitral annulus because it has a larger GOA. The details of the prosthesis data are shown in Table 1.
Estimation of the MSA of the Prosthetic Valves
Computed tomography (CT) images of each prosthesis were acquired using a dual-source CT system (SOMATOM Definition Flash, Siemens Medical Solution, Forchheim, Germany) with a 2 × 64 × 0.6 mm detector collimation. A noncontrast CT scan was conducted in accordance with the following parameters: 280-ms gantry rotation time, 100-kV tube potential, and 80 tube current. CT images were reconstructed with 0.6-mm section thickness and reconstruction increments using a sharp kernel.
To estimate the MSA of the prosthetic valves from the CT images, a valve was first segmented from the CT images using the thresholding technique (I > 76 HU). Next, 3D meshes were generated from the 3D volumetric valve images using the Iso2mesh function of the mesh generation toolbox.7 The mesh area was computed, and the MSA was calculated as the sum of the mesh areas. The CT images were analyzed using an in-house software written using MATLAB v.7.6 (The Mathworks Inc, Natick, MA). This process was repeated three times. The average values were used for the analysis.
Definitions and Follow-up
The MSA was measured using CT for each type and size of the new mitral prosthesis (Figure 2). Although there have been a few minor changes in the prosthetic sewing ring, no maker changed the design or ingredient of the blood-contacting mechanical components. The MSA index (MSAI, cm2/m2) was defined as the MSA divided by the BSA of the patients at the time of surgery. Anticoagulation was maintained via daily warfarin administration. The target international normalized ratio (INR) was consistent between 2.0 and 3.0 during the study period. The primary end-point of the study was death, regardless of the cause. The secondary end-points consisted of cardiac death and major adverse valve–related events, including structural or nonstructural valve dysfunction, thrombosis, embolism, bleeding, endocarditis, and reintervention. Anticoagulation-related events included thrombosis, embolism, and all bleeding events. All outcomes were defined in accordance with previously reported guidelines.8 Two-dimensional Doppler echocardiography was performed before discharge and 1, 3, and 5 years after surgery. Echocardiographic data, including left ventricular ejection fraction, left ventricular dimension, and mean prosthetic pressure gradient, were available for 393 patients (82%) at 1 year, 312 patients (69%) at 3 years, and 193 patients (45%) at 5 years postoperatively.
Data were obtained from the computerized medical records in the hospital database. Most patients (78%) regularly visited the outpatient clinic at the time of data collection. Additional data were obtained via telephone interview with the patients or families. The latest patient follow-up was January 2014. Survival status was confirmed for all patients using a national database. The mean follow-up or survival duration was 115.4 ± 56.21 months (0.7 to 228.8 months).
Measurements were expressed as means ± standard deviations or as frequencies and proportions. The patient characteristics associated with an increased risk of mortality were examined using univariate and multivariate Cox regression analyses. Clinically relevant variables and those with p values of <0.05 in the univariate analysis were entered into the multivariate analysis model. Survival and event-free survival curves were estimated using the Kaplan–Meier method. p values of <0.05 were considered statistically significant. The INR values were compared using averages during the follow-up period. Statistical analysis was performed using the SPSS version 21.0 (SPSS Inc., Chicago, IL).
Patient Characteristics and Operative Data
The mean patient age was 50.6 ± 11.1 years. The mean BSA and MSAI were 1.6 ± 0.17 m2 and 12.9 ± 1.7 cm2/m2, respectively. The patients were divided into two groups (low and high MSAI) according to the median MSAI (12.92 cm2/m2). The high MSAI group had higher indexed left ventricular end-diastolic dimension and smaller BSA than the low MSAI group. The other preoperative characteristics were not significantly different between the two groups (Table 2). The On-X valve was used most often (n = 170, 35%). Tricuspid repair with a modified Cox-Maze operation was the most common simultaneous procedure (Table 3). Tricuspid annuloplasty was performed using an annuloplasty ring (58%) or sutures (modified De Vega method, 42%).
Impact of MASA on the Clinical Outcomes
There was no difference in the INR between the two groups during the follow-up period (p = 0.94). Overall, early death was reported in 1% of the patients (Table 3). The most common postoperative events were related to anticoagulation issues (11%), such as thromboembolic and bleeding events (n = 29 and n = 27, respectively). The estimated 5 and 10 year survival rates were 95% and 93%, respectively. According to the Kaplan–Meier survival analysis (log rank method), the long-term survival, valve-related event-free survival, and thromboembolic event-free survival differed between the low and high MSAI groups. However, the long-term survival was identical, regardless of the type of prosthesis (Figure 3). The multivariate analysis showed that old age, New York Heart Association class III or IV, and high MSAI were independent predictors of death. With respect to thromboembolic events, the patients with a high MSAI and New York Heart Association class III or IV had an increased risk of mortality (Table 4).
Hemodynamic Performance of the Mitral Prostheses
There was no difference in the mean left ventricular ejection fraction between the groups at the 1 year follow-up (56.27 ± 8.624 and 57.17 ± 8.886, respectively, p = 0.31). A total of 11 patients (3%, eight patients in the low MSAI group and three patients in the high MSAI group) had a mean pressure gradient across the mitral valve greater than 6 mm Hg. The number of patients with a mean transmitral pressure gradient of > 6 mm Hg was not significantly different between the groups (p = 0.138). A mean transmitral pressure gradient of > 6 mm Hg was not related to the survival (p = 0.467). The mean pressure gradient at the 1 year follow-up was not different between the groups (3.5 ± 1.18 and 3.4 ± 1.21, respectively, p = 0.289).
Although there have been many studies on the PPM of prosthetic aortic and mitral valves, its impact on the clinical outcomes is inconclusive.9–12 Previous studies have evaluated the effect of a small mitral prosthesis size on outcomes. Many surgeons routinely attempt to implant a large valve because small prosthetic valves can lead to abnormally increased transvalvular gradients. However, the mitral annulus can be excessively dilated by chronic heart disease and subsequent remodeling processes. If the mitral annulus is abnormally enlarged, implanting a prosthesis that fits exactly into the annulus may result in an excessively and unnecessarily large mitral prosthesis (Figure 1).
Prosthesis Size, MSA, and Clinical Implications
The size of a mechanical prosthesis can be represented by the manufacturer’s valve size, tissue annulus diameter, internal diameter, sewing ring diameter, GOA with or without leaflets, and EOA. Although the EOA is used to evaluate the hemodynamic performance of prosthetic valves, this parameter may be inconsistent. The GOA represents the actual size of the prosthesis8,13,14; however, most complications associated with mechanical mitral valves are related to the thrombotic properties of metallic components and to anticoagulation-related complications. We hypothesized that a 3D value, such as the MSA, might be an important predictor of the long-term outcomes of mechanical MVR. In our study, the participants in the high MSAI group experienced higher all-cause mortality and anticoagulation-related events, especially thromboembolic complications. Only a few differences were noted between the low and high MSAI groups, including sex, left ventricular end-diastolic dimension, BSA, and type of prosthesis. The multivariate analysis was performed to adjust for possible confounders. Although we adjusted for potential confounders, such as left ventricular end-diastolic dimension, atrial fibrillation at discharge, and left ventricular ejection fraction, a high MSAI was independently related to all-cause mortality and thromboembolic complications among the patients.
Clinical Implications and Mechanisms of Adverse Effects of Large Mechanical Mitral Prostheses
The increase in the poor outcomes associated with large mechanical mitral prostheses may be because of anticoagulation-related events. A large prosthesis in relation to the patient size will have a low pressure gradient and velocity across the valve. Blood with a low flow rate may increase the risk of thrombosis and embolism. Larger prostheses have a larger MSA, which increases the contact between blood and foreign bodies, thereby, activating the coagulation system. In the current study, there was an increase in the anticoagulation-related and thromboembolic events in the patients with a high MSAI. Thus, a high MSAI may increase the risk of thrombosis, embolism, and consequent death. Our results suggest that cardiac surgeons should consider patient size (e.g., BSA), type of prosthesis, and diameter of the mitral annulus when deciding on the appropriate size of mechanical prostheses for MVR. We also suggest that maintaining higher anticoagulation levels, such as 3.0 to 3.5, than our standard level (2.0 to 3.0) in such patients, may prevent thromboembolic complications.
This was a retrospective review of our clinical database conducted more than 15 years, during which time surgical procedures and valve selection changed. However, there was no difference in the rate of survival according to prosthesis type. Larger-scale studies investigating specific types of prostheses may be necessary to further investigate our hypothesis. A small percentage of patients (mean pressure gradient of ≥6 mm Hg 1 year after surgery, n = 11, 3%) in our study exhibited PPM. All patients in this study were Koreans with a relatively small BSA (mean BSA, 1.6 ± 0.17 m2); therefore, the impact of large mitral prostheses on patients is more evident in our study than in studies conducted in Western countries. Our study does not suggest the preferential use of small mitral prostheses.
In our study, a large mechanical mitral prosthesis MSA relative to patient BSA was a predictor of poor clinical outcomes. A high MSAI was associated with increased late death and thromboembolic complications. Both the BSA and size of the mitral annulus should be considered when selecting a mechanical mitral prosthesis for isolated MVRs. In addition, high levels of anticoagulation may be beneficial in patients with high MSAIs.
1. Pibarot P, Dumesnil JG. Prosthesis
-patient mismatch: definition, clinical impact, and prevention. Heart 2006.92: 10221029.
2. Li M, Dumesnil JG, Mathieu P, Pibarot P. Impact of valve prosthesis
-patient mismatch on pulmonary arterial pressure after mitral valve
replacement. J Am Coll Cardiol 2005.45: 10341040.
3. Marijon E, Mirabel M, Celermajer DS, Jouven X. Rheumatic heart disease. Lancet 2012.379: 953964.
4. Doukky R, Abusin SA, Bayissa YA, Kelly RF, Ansari AH. Rheumatic heart disease in modern urban America: a cohort study of immigrant and indigenous patients in Chicago. Int J Cardiol 2014.175: 178180.
5. Ulus AT, Tütün U, Budak B, et al. Evaluation of hemodynamic changes in patients with mitral valve
replacement using dobutamine stress echocardiography
. Anadolu Kardiyol Derg 2007.7: 397403.
6. Kestelli M, Zeybek R, Onem G, et al. Late effect of mechanical mitral valve
diameter on left ventricular function. Asian Cardiovasc Thorac Ann 1999.7: 3032.
7. Qianqian F, Boas DA. Tetrahedral mesh generation from volumetric binary and grayscale images, 2009, pp. in 2009 IEEE International Symposium on Biomedical Imaging: From Nano to Macro, Boston, Massachusetts, 11421145.
8. Blackstone EH, Cosgrove DM, Jamieson WR, et al. Prosthesis
size and long-term survival after aortic valve replacement. J Thorac Cardiovasc Surg 2003.126: 783796.
9. Blais C, Dumesnil JG, Baillot R, Simard S, Doyle D, Pibarot P. Impact of valve prosthesis
-patient mismatch on short-term mortality after aortic valve replacement. Circulation 2003.108: 983988.
10. Rao V, Jamieson WR, Ivanov J, Armstrong S, David TE. Prosthesis
-patient mismatch affects survival after aortic valve replacement. Circulation 2000.102: III59.
11. Okamura H, Yamaguchi A, Yoshizaki T, et al. Clinical outcomes and hemodynamics of the 19-mm Perimount Magna bioprosthesis in an aortic position: comparison with the 19-mm Medtronic Mosaic Ultra Valve. Circ J 2012.76: 102108.
12. Sato S, Fujita T, Shimahara Y, Hata H, Kobayashi J. Impact of prosthesis
-patient mismatch on late recurrence of atrial fibrillation after cryomaze procedure with mitral valve
replacement. Circ J 2014.78: 19081914.
13. Christakis GT, Buth KJ, Goldman BS, et al. Inaccurate and misleading valve sizing: a proposed standard for valve size nomenclature. Ann Thorac Surg 1998.66: 11981203.
14. Gillinov AM, Blackstone EH, Rodriguez LL. Prosthesis
-patient size: measurement and clinical implications. J Thorac Cardiovasc Surg 2003.126: 313316.