Prosthesis-Patient Mismatch Affects Survival After Aortic Valve Replacement
Background—Surgeons traditionally avoid the use of “small” aortic prostheses because of the potential for residual left ventricular outflow tract obstruction and persistent transvalvular gradients. This study examines the ratio between prosthetic valve size and the body surface area (BSA) of patients undergoing aortic valve replacement (AVR). We sought to determine the effect of potential “prosthesis-patient” mismatch on long-term survival.
Methods and Results—Follow-up was conducted on 2981 patients who underwent AVR with stented bioprostheses between 1976 and 1996. To account for differences between manufacturers’ labeled valve sizes, we calculated the ratio between the prosthetic valve effective orifice area (EOA) and the patient’s BSA (recorded for 2154 patients). The lowest decile in this cohort had a calculated EOA/BSA of <0.75 cm2/m2 (Small group, n=227) compared with the control group (n=1927), in whom the EOA/BSA ratio was >0.75 cm2/m2. Operative mortality was higher in the Small group (8% versus 5%, P=0.03). Actuarial survival at 12 years was 50±5% in the Small group compared with 49±2% in the control group (P=0.27). However, freedom from valve-related mortality was significantly lower in the Small group (75±5% versus 84±2%, P=0.004). Cox regression analysis determined age and NYHA functional class to be the multivariate predictors of overall mortality, whereas advanced age and EOA/BSA <0.75 cm2/m2 were found to be the predictors of valve-related mortality.
Conclusions—Prosthesis-patient mismatch results in significantly higher early and late mortality after bioprosthetic AVR. We recommend careful selection of stented bioprostheses to ensure an adequate ratio of EOA to BSA. An EOA/BSA ratio of >0.75 cm2/m2 may avoid residual left ventricular outflow tract obstruction and persistent transvalvular gradients. Careful prosthesis-patient matching will improve both early and late survival after AVR.
Prosthetic valve replacement represents a successful surgical therapy for patients with symptomatic aortic valve disease. Despite an increasing proportion of high-risk patients presenting for valve surgery, the morbidity and mortality of isolated aortic valve replacement (AVR) remain low.1 Small aortic annular size has been reported to be associated with increased operative mortality.1 2 3 However, considerable controversy remains regarding the effects of small aortic prostheses on long-term survival.4 5
Prosthesis-patient mismatch occurs when the effective orifice area (EOA) of an artificial valve is inadequate for the recipient. Prosthesis-patient mismatch can result in persistent left ventricular outflow tract obstruction and higher postoperative transvalvular gradients. This residual obstruction may prevent the regression of left ventricular mass index observed in most patients undergoing AVR for aortic stenosis. Earlier regression of left ventricular mass index has been suggested to improve long-term survival after AVR.6 7
Unfortunately, data are lacking in the surgical literature to accurately determine the effects of small prosthesis size on long-term survival. The controversy is further confounded by the differences in valve sizing among manufacturers.8 The measured internal or external diameter of a “19-mm” valve varies by up to 4 mm depending on the valve manufacturer. Thus, the measured dimensions of a 19-mm Carpentier-Edwards pericardial valve are similar to those of a 21-mm Hancock II bioprosthesis. We have previously demonstrated that apparent hemodynamic differences between valves of similar labeled sizes are no longer present when the comparison is adjusted to reflect true measured dimensions.9 Therefore, any investigation of the effect of valve size must be based on a uniform measurement that is applicable across valve manufacturers. In this study, we used the preoperative EOA indexed to body surface area (BSA) as a measure of prosthesis-patient mismatch5 and used this ratio to evaluate long-term survival after AVR.
The combined valvular registry at Toronto General Hospital and Vancouver General Hospital was reviewed for all patients who underwent isolated bioprosthetic AVR with the Hancock II bioprosthesis (n=670; Medtronic Inc) or the Carpentier-Edwards standard (n=420), pericardial (n=233), or supra-annular (n=1658) valve (Baxter Inc) with or without concomitant coronary artery bypass between 1982 and 1995 (Toronto) or between 1976 and 1996 (Vancouver). Preoperative measurements of calculated BSA were available for 2154 patients (72%), who form the basis of the survival analysis. Estimates of prosthetic EOA were obtained from the manufacturer’s specifications for each size valve before implantation. The range of valve sizes implanted is given in Table 1⇓.
Follow-up was performed at the respective center’s valve surgery clinic and was complete in all but 2 patients (99.9%) with a mean follow-up of 74±49 months (range, 1 to 233 months). Morbidity and mortality were specified according to the Society of Thoracic Surgeons/American Association for Thoracic Surgery guidelines for reporting events after cardiac valvular operations.10 According to these guidelines, valve-related mortality is death caused by structural valve deterioration, nonstructural deterioration, valve thrombosis, embolism, bleeding event, valvular endocarditis, or death as a result of reoperation. Sudden, unexplained deaths are included as valve-related mortality. Operative mortality is defined as death within 30 days of operation or during the same hospitalization.
All data were analyzed with the SAS statistical software program (SAS Institute). Continuous data were compared by Student’s t test and are expressed as mean±SD. Categorical data were compared by χ2 analysis. Survival data were analyzed by Cox regression analysis. Prosthetic EOA was indexed to BSA to yield a measure of prosthesis-patient mismatch (EOA/BSA ratio). Similarly, the in vitro internal diameter (ID) of each valve was indexed to BSA to yield an alternative measurement of prosthesis-patient mismatch (ID/BSA). Variables entered into a multivariable model examining the predictors of long-term survival included age, sex, repeat operation, NYHA functional class, coronary artery disease, and a measure of valve size (EOA/BSA, ID/BSA, or size of 19 or 21 mm). Exact probability values are provided for each comparison to enable the reader to determine clinical and statistical significance.
Figure 1⇓ illustrates the relationship between valve size and patient BSA. An EOA/BSA ratio of 0.75 cm2/m2 represents the lowest decile in this patient population (n=227). Similarly, an ID/BSA ratio of <10 mm/m2 represents those patients with prosthesis mismatch. Overall operative mortality was 5.3%. Operative mortality in patients receiving a 19- or 21-mm valve was 9.3%, compared with 4.2% in those receiving a valve >21 mm (P=0.001). The operative mortality was also higher in those patients with an EOA/BSA ratio of <0.75 cm2/m2 (7.9% versus 4.6%, P=0.027). Table 2⇓ illustrates the demographic differences in patients with prosthesis mismatch (EOA/BSA ≤0.75 cm2/m2) and the remainder of the study group (n=1927). Patients with prosthesis mismatch were more likely to be women with larger BSAs. The multivariable predictors of operative mortality were age (OR 1.03, 95% CI 1.00 to 1.05), female sex (OR 1.75, 95% CI 1.19 to 2.58), and preoperative NYHA status (OR 2.85, 95% CI 2.00 to 4.05).
Figure 2⇓ compares overall survival in patients who received a 19- or 21-mm valve (manufacturer’s labeled size) and those who received a valve >21 mm in diameter. The survival at 12 years was 43±3% in patients with small prostheses compared with 50±2% in patients with larger valves (P=0.003). However, small valve size (19 or 21 mm) failed to be a multivariable predictor of valve-related mortality.
Figures 3⇓ and 4⇓ represent overall survival and freedom from valve-related mortality, respectively, in patients with and without prosthesis mismatch (EOA/BSA ≤0.75 cm2/m2). Overall survival at 12 years was 50±5% in patients with EOA/BSA ≤0.75 cm2/m2 compared with 49±2% in patients with EOA/BSA >0.75 cm2/m2 (P=0.27). The multivariable predictors for overall mortality were age (relative risk [RR] 1.06, 95% CI 1.04 to 1.07) and preoperative NYHA functional status (RR 1.25, 95% CI 1.09 to 1.49). The predictors of valve-related mortality were age (RR 1.07, 95% CI 1.04 to 1.09) and prosthesis mismatch (RR 1.46, 95% CI 0.95 to 2.24).
Figure 5⇓ illustrates overall survival in patients with an ID/BSA ratio >10 mm/m2 compared with patients with prosthesis mismatch (ID/BSA ratio ≤10 mm/m2). By this measure of relative valve size, there was a divergence in survival curves after 10 years (59±1% versus 28±5%, P=0.03).
For patients with aortic stenosis, there is a widely accepted correlation between the reduction of transvalvular gradients and clinical improvement after AVR.3 4 Persistently high postoperative gradients are thought to impair the regression of left ventricular mass observed in most patients after surgery.3 6 7 Complete abolition of transvalvular gradients is not possible because of the obstructive effects of the prosthetic valve stents and sewing ring. A 19-mm Carpentier-Edwards Perimount bioprosthesis has an estimated EOA of 1.21±0.31 cm2. For a patient with a BSA of 2.0 m2, implantation of this valve would lead to an EOA/BSA ratio of 0.6 cm2/m2. According to the natural history study performed by Pellikka and colleagues,11 a ratio <0.6 cm2/m2 represents severe aortic stenosis. In contrast, implantation of the same valve in a smaller patient with a BSA of 1.5 would lead to an EOA/BSA ratio of 0.8 cm2/m2, a value associated with only mild aortic stenosis. Therefore, small aortic valves (19 or 21 mm) may be nonobstructive provided that they are implanted in small patients with small hearts and a small aorta.
Figure 6⇓ illustrates the EOA/BSA ratio of Carpentier-Edwards Perimount valves as a function of the patient’s BSA. To achieve an EOA/BSA ratio of >0.75 cm2/m2, a 21-mm valve should be implanted only in a patient with a BSA of <1.6 m2. Similarly, a 23-mm valve should be implanted in patients with BSAs of <2.0 m2.
Unfortunately, extrapolating Figure 6 to 21- or 23-mm valves from different manufacturers is difficult because of the inconsistencies in valve sizing and labeling.8 Accurate estimates of maximal theoretical EOAs are required for each individual valve type and size to avoid prosthesis-patient mismatch. Although no prospective randomized data are available to support the decision to avoid prosthesis-patient mismatch, there is an abundance of long-term observational data highly suggestive of the detrimental effects of residual postoperative gradients.2 3 4 6
A 30-year observational study by He and colleagues2 demonstrated that AVR in the small aortic root is satisfactory in patients with small BSA. However, in the subset of patients with concomitant coronary artery bypass surgery, prosthesis-patient mismatch proved to be a negative determinant of long-term survival. These patients are at high risk for sudden cardiac death due to a combination of left ventricular hypertrophy and coronary artery disease. Persistently elevated left ventricular mass in these patients, as a result of a small prosthesis with obstructive gradients, may increase the risk of sudden death. Data from the hypertension literature support the concept of lowering the risk of sudden cardiac death by facilitating the regression of left ventricular hypertrophy, particularly in patients with concomitant coronary artery disease.12 13 14
The results of this study are in contrast to other reports that demonstrate no meaningful effect of prosthesis size on long-term survival.2 15 However, there are several important characteristics of the present study. The combined database of Toronto General Hospital and Vancouver General Hospital allowed for sufficient statistical power to detect a 10% difference in survival at 12 years. Furthermore, there was little effect of indexed valve size on overall mortality, only on valve-related mortality. This suggests that prosthesis-patient mismatch poses a risk for valve-related events but that this risk may be overshadowed by the generally poor survival after AVR. The use of an indexed valve size corrects for patient size and for differences among manufacturers, a confounding variable present in most previous studies examining the effect of valve size on long-term outcomes.2 3 6 15 Clearly, a 19-mm valve (from any manufacturer) would be perfectly appropriate in the 80-year-old woman with a BSA of 1.5 m2. Even a larger patient in this age group would be unlikely to suffer detrimental effects of prosthesis-patient mismatch. Our results indicate that survival advantages occur only after postoperative year 7.
A significant limitation of this study is the lack of follow-up echocardiographic data. Most bioprostheses undergo degenerative changes with time. It is therefore conceivable that a patient who received a valve with an EOA/BSA ratio of 0.9 cm2/m2 may demonstrate deterioration with time such that a follow-up echocardiographic study yields an EOA/BSA ratio in the stenotic range (<0.8 cm2/m2). Conversely, early remodeling may increase the EOA to negate a potential prosthesis-patient mismatch.7 In addition, differences in the rates of structural valve deterioration between bioprostheses may also influence patient survival.16 17 18
The controversy regarding the long-term effects of prosthesis-patient mismatch will probably continue until a large, prospective randomized trial is performed. Unfortunately, this trial will most likely never be completed. A more practical study would involve a prospective, long-term, echocardiographic analysis of the changes in indexed valvular EOA with time and its impact on patient survival. A prospective demonstration of the negative impact of prosthesis-patient mismatch may influence surgeons to accept higher perioperative morbidity by performing more complex procedures, such as an annular enlargement19 or a stentless valve implantation.20 21 It is unclear whether the concept of prosthesis-patient mismatch even exists in stentless prostheses (homograft, autograft, or xenograft).
In summary, the effect of indexed EOA on patient survival after AVR remains controversial. The data from this study indicate a negative impact on long-term valve-related mortality in a subgroup of patients with severe prosthesis mismatch. This finding is particularly striking given that fewer patients in the mismatched group were in preoperative NYHA functional class III or IV, suggesting that they should have had superior long-term survival. Further examinations of the effect of valve size on long-term survival should strive to eliminate the confounding variables of patient size and valve design. A sufficiently powered prospective study should demonstrate a small, yet clinically significant, survival benefit in young patients who receive an adequately matched aortic prosthesis.
- Copyright © 2000 by American Heart Association
Rao V, Christakis GT, Weisel RD, et al. Changing pattern of valve surgery. Circulation. 1996;94(suppl II):II-113–II-120.
He GW, Grunkemeier GL, Gately JL, et al. Up to thirty-year survival after aortic valve replacement in the small aortic root. Ann Thorac Surg. 1995;59:1056–62.
Gonzalez-Juanetey JR, Garcia-Acuna JM, Vega Fernandez M, et al. Influence of the size of aortic valve prostheses on hemodynamics and change in left ventricular mass: implications for the surgical management of aortic stenosis. J Thorac Cardiovasc Surg. 1996;112:273–280.
Pibarot P, Honos GN, Durand LG, et al. The effect of prosthesis-patient mismatch on aortic bioprosthetic valve hemodynamic performance and patient clinical status. Can J Cardiol. 1996;12:379–387.
McDonald ML, Daly RC, Schaff HV, et al. Hemodynamic performance of small aortic valve bioprostheses: is there a difference? Ann Thorac Surg. 1997;63:362–366.
Jin XY, Zhang Z-M, Gibson DG, et al. Effects of valve substitute on changes in left ventricular function and hypertrophy after aortic valve replacement. Ann Thorac Surg. 1996;62:683–690.
Christakis GT, Joyner C, Fremes SE, et al. Left ventricular mass regression early following aortic valve replacement. Ann Thorac Surg. 1996;62:1084–1089.
Christakis GT, Buth KJ, Goldman BS, et al. Inaccurate and misleading valve sizing: a proposed standard for valve size nomenclature. Ann Thorac Surg. 1998;66:1198–1203.
Rao V, Christakis GT, Sever J, et al. A novel comparison of stentless vs. stented valves in the small aortic root. J Thorac Cardiovasc Surg. 1999;117:431–438.
Edmunds LH, Clark RE, Cohn LH, et al. Guidelines for reporting morbidity and mortality after cardiac valvular operations. Ann Thorac Surg. 1996;62:932–935.
Pellikka PA, Nishimura RA, Bailey KR, et al. The natural history of adults with asymptomatic, hemodynamically significant aortic stenosis. J Am Coll Cardiol. 1990;15:1012–1017.
Bikkina M, Larson MG, Levy D. Asymptomatic ventricular arrhythmias and mortality risk in subjects with left ventricular hypertrophy. J Am Coll Cardiol. 1993;22:1111–1116.
Levy D. Clinical significance of left ventricular hypertrophy: insights from the Framingham Study. J Cardiovasc Pharmacol. 1991;17(suppl 2):S1–S6.
Ghali JK, Liao Y, Simmons B, et al. The prognostic role of left ventricular hypertrophy in patients with or without coronary artery disease. Ann Intern Med. 1992;117:831–836.
Sawant D, Singh AK, Feng WC, et al. Nineteen millimeter aortic St. Jude Medical heart valve prosthesis: up to sixteen years’ follow-up. Ann Thorac Surg. 1997;63:964–970.
Jamieson WRE, Burr LH, Tyers GFO, et al. Carpentier-Edwards standard and supra-annular porcine bioprostheses: 10 year comparison of structural valve deterioration. J Heart Valve Dis. 1994;3:59–65.
Barratt-Boyes BG, Jaffe WM, Ko PH, et al. The zero pressure fixed Medtronic Intact porcine valve: an 8.5 year review. J Heart Valve Dis. 1993;2:604–611.
David TE, Armstrong S, Sun Z, et al. The Hancock II bioprosthesis at 12 years. Ann Thorac Surg. 1998;66:S95–S98.
Sommers KE, David TE. Aortic valve replacement with patch enlargement of the aortic annulus. Ann Thorac Surg. 1997;63:1608–1612.
Sintek CF, Fletcher AD, Khonsari S. Stentless porcine aortic root: valve for choice for the elderly patient with small aortic root? J Thorac Cardiovasc Surg. 1995;109:871–876.
David TE, Puschmann R, Ivanov J, et al. Aortic valve replacement with stentless and stented porcine valves: a case-match study. J Thorac Cardiovasc Surg. 1998;116:236–241.