Interact CardioVasc Thorac Surg 2007;6:462-469. doi:10.1510/icvts.2006.145532
© 2007 European Association of Cardio-Thoracic Surgery
Institutional report - Valves |
Prosthesis–patient mismatch with latest generation supra-annular prostheses. The beginning of the end?
Rafael García Fuster*,
Vanesa Estevez,
Ignacio Rodríguez,
Sergio Cánovas,
Oscar Gil,
Fernando Hornero and
Juan Martínez-León
University General Hospital of Valencia, Valencia, Spain
Received 27 September 2006;
received in revised form 13 March 2007;
accepted 15 March 2007
 Presented at the joint 20th Annual Meeting of the European Association for Cardio-thoracic Surgery and the 14th Annual Meeting of the European Society of Thoracic Surgeons, Stockholm, Sweden, September 10–13, 2006.
*Corresponding author. C/Artes Gráficas n°4, esc. izq. pta.3, 46010 Valencia, Spain. Tel.: +34 96 3622216; fax: +34 96 1972163.
E-mail address: rgfuster{at}terra.com (R. García Fuster).
 |
Abstract
|
|---|
Most studies about prosthesis–patient mismatch (PPM) were conducted before the introduction of new high-performance prostheses. Nowadays, PPM could become unfrequent. Our aim was to study the impact of new prostheses on PPM in comparison with previous experience. Prosthetic Indexed Effective Orifice Area (EOAi) was estimated in two historical cohorts. Group A: 339 patients undergoing AVR from Mar 94–Nov 01. Group B: 404 operated on during the last three years when latest generation prostheses were implanted. Incidence, determinants of PPM and clinical results were studied. Moderate PPM (EOAi  0.85 cm2/m2) was present in 38% and 19% (respective groups). Mean EOAi increased from 1.02±0.29 cm2/m2 to 1.11±0.27 cm2/m2. ‘Group B’ and ‘new prostheses’ were protective. Thirty-day mortality was 3.8% and 4.7% with higher rate in patients with increased left ventricular mass index (LVMI), especially if PPM was present: 14.7 vs. 2.1% (P<0.05) in Group A; 25.0 vs. 4.8% (P<0.05) in Group B (PPM vs. no-PPM). LVMI regression was impaired in these patients. Moderate PPM was an independent predictor of late cardiac mortality (OR: 3.38, 95% CI: 1.37–8.31; P<0.01). PPM is a prognostic factor for late cardiac death. Its impact on early mortality is only relevant in patients with high LVMI. Its incidence has decreased with the use of new prostheses.
Key Words: Prosthesis–patient mismatch; Aortic valve replacement; Mortality
|
1. Introduction
|
|---|
Prosthesis–patient mismatch (PPM) is presented when the prosthesis used for aortic valve replacement (AVR) is too small in relation to patient's body size. Several studies have demonstrated favorable results despite PPM. Others have found PPM as a strong predictor of mortality [1–3]. But most studies were conducted before the introduction of new supra-annular prostheses. Recently, prostheses design has evolved with lower profiles, thinner sewing rings and better hemodynamics [4]. Consequently, with the use of the latest generation prostheses moderate–severe PPM could become unfrequent.
Our aim was to evaluate the impact of these new prostheses on the incidence and effects of PPM.
|
2. Materials and methods
|
|---|
2.1. Study group
Two non-consecutive historical cohorts of patients who underwent AVR because of predominant aortic valve stenosis were evaluated. In Group A were included 339 patients from a previous study (Mar 1994–Nov 2001) [5]. Group B included 404 consecutive patients operated on during a recent period (Jan 2003–Dec 2005), starting in 2003 when new prostheses were first implanted (Table 1). Only isolated AVR were considered. CABG, redo AVR, emergent operation (carried out on referral before the beginning of the next working day) or other procedures were excluded. Aortic annulus enlargement and septal myectomy were carried out only for specific indications and were not considered. Active endocarditis was also excluded.
2.2. Prosthesis–patient mismatch
PPM estimation was based on the projected EOA index [6] using the in vitro EOA indexed to body surface area. Moderate PPM was considered as an EOA index 0.85 cm2/m2 and severe PPM as <0.65 cm2/m2. Values >0.85 cm2/m2 were defined as no PPM. Expected EOA for different prostheses are shown in Table 1.
2.3. Echocardiography
Preoperative echocardiography was performed in all patients. Postoperatively it was obtained in 312 and 378 patients at the 1st month in groups A and B, respectively. In 190 and 320 patients, at least one echocardiogram was performed between six months to one year. Left ventricular mass index (LVMI) was calculated by the formula described by Devereux and used in a previous article [5]. The highest quartile of LVMI values in both sexes was considered as increased (Group A: >226 g/m2 in males and >216 g/m2 in females; Group B: >215 g/m2 in males and >198 g/m2 in females).
2.4. Operative technique
CPB standard techniques were used. Myocardial protection was similar during the study period: moderate hemodilution and multidose intermittent (antegrade/retrograde) cold blood cardioplegia. Prosthesis size and model were used to the surgeon's discretion. Prosthesis type was selected according to ACC/AHA guidelines [7]. Valve implantation was performed using horizontal mattress sutures reinforced with pledgets in a non-everting fashion.
2.5. Definitions
Previous stroke was defined as history of central neurologic deficit >72 h. Chronic renal failure (CRF) as creatinine  2 mg/dl. Chronic obstructive pulmonary disease (COPD) as the need for pharmacologic therapy for chronic pulmonary compromise or espirometry with moderate–severe obstruction. Urgent operation was considered when the procedure was performed during the hospital stay of an acute episode. Emergency status was excluded. Thirty-day mortality was defined as death within 30 days of operation regardless of the patient's geographic location.
2.6. Thirty-day mortality and follow-up
Thirty-day mortality was calculated considering the presence of PPM. Patients were examined on a regular basis 4–6 weeks postoperatively and every six months. Further data were obtained from hospital records or by telephone. Follow-up was too short and late survival analysis was not considered in Group B.
2.7. Statistical analysis
Statistical analysis was performed using SPSS statistical package 12.0 for Windows (SPSS Corp, Birmingham, AL, USA). Continuous variables were summarized as mean±S.D. if normally distributed, on the contrary, median and inter-quartile ranges were used. Nominal data were presented as frequencies and percentages. Continuous variables were compared by unpaired t-test. Mann–Whitney test was used if not normally distributed data. Associations among categorical variables were compared by Pearson's  2 test, continuity correction or two-sided Fisher exact test.
Univariate analysis of risk factors for PPM and 30-day mortality was performed. Several variables were initially considered (Table 4). All variables with P<0.10 were entered into a subsequent multivariate analysis. Stepwise logistic-regression was used to assess the independent impact of the risk factor on the occurrence of PPM and 30-day mortality (selection cut-off set at 0.05). Hosmer–Lemeshow goodness-of-fit statistic and the area under the ROC curve were calculated. Cox regression analysis was used to identify predictors for late death. Non-parametric estimates of death were determined using the Kaplan–Meier method. Differences in survival curves were analyzed with the log-rank test.
To control for selection bias in the comparison between patients with and without PPM, propensity score (PS) matching was performed. The PS was calculated for each patient from both cohorts using all the above-mentioned variables (Table 4). For every PPM patient, matching patients with the closest PS were identified from the larger pool of no PPM group (maximum allowable difference: <0.1). When two or more patients had the same PS, the match was chosen randomly. This left 100 PPM patients matched to an equal number of no PPM patients from both cohorts.
|
3. Results
|
|---|
3.1. Pre and operative data
Group B patients were older and had worse LV function, more symptoms and higher risk (EuroSCORE), reflecting a change in risk profile (Table 2). Rheumatic disease was more prevalent in Group A. Bioprostheses were implanted more frequently in Group B. Latest generation valves were exclusively implanted in these patients.
3.2. Incidence of PPM
Mean EOA index was lower in Group A: 1.02±0.29 cm2/m2 (IQR: 0.81–1.21) vs. 1.11±0.27 cm2/m2 (IQR: 0.90–1.31). Observed EOA by echocardiography showed a similar trend (Table 3). Fig. 1a,b shows the distribution of EOA index per valve size. Fig. 1c is referred to latest generation prostheses. Most of them were over the cut-off point considered to define moderate–severe PPM (cut-off lines in Fig. 1c,d). The incidence of PPM was higher in the previous cohort: 129 of 339 patients (38%) in Group A and only 77 of 404 patients (19%) in Group B (P<0.0001). Severe PPM was rare in Group B: 6.8% vs. 1.6% (P<0.0001). This reduction in PPM was especially marked with bioprostheses: 22.8% vs. 9.7% with mechanical (P<0.001) and 75.5% vs. 26.8% with bioprostheses (P<0.0001). PPM with latest generation prostheses was anecdotal.
View larger version (15K):
[in this window]
[in a new window]
|
Fig. 1. Distribution of EOA index values, (a) EOA index per valve size in Group A, (b) EOA index per valve size in Group B, (c) EOA index values in patients with latest generation prostheses, (d) EOA index values in patients without latest generation prostheses.
|
|
3.3. Determinants of PPM
Determinants of PPM were studied in the whole group (Table 4). Latest generation prostheses and Group B had a protective effect. Elderly patients, females, degenerative etiology and bioprosthesis were associated with PPM.
3.4. Thirty-day mortality: the effect of PPM
(a) In global groups: Thirty-two patients died within the first 30 days (4.3%): 13 patients from Group A (3.8%) and 19 from Group B (4.7%). Only a trend to higher mortality was observed with PPM: 5.4% vs. 2.9% (P=0.36) in Group A and 6.3% vs. 4.3% (P=0.46) in Group B. Cardiac mortality was also higher with PPM without statistical significance: 3.9% vs. 1.4% (P=0.28) in Group A and 3.8% vs. 1.5% (P=0.41) in Group B.
The most frequent cause of 30-day mortality was cardiac. Causes of death and factors influencing 30-day mortality are shown in Table 5. High LVMI was an important factor. Neither EOA index nor moderate PPM were significant.
(b) Influence of increased LVMI and PPM effect: a subgroup analysis. Patients with increased LVMI had a higher mortality: 8.4% vs. 2.3% (P<0.05) in Group A and 11.8% vs. 4.1% (P<0.05) in Group B. In those with PPM, this impact of LVMI on mortality was higher: 14.7% vs. 2.1% (P<0.05) and 25.0% vs. 4.8% (P<0.05) in the respective groups; but in no PPM-patients this difference was reduced: 4.1% vs. 2.5% (P=0.13) and 7.7% vs. 3.9% (P=0.35).
3.5. Follow-up
(a) Clinical status: Follow-up was complete in Group A: 83±29 months (4–12 years). In no PPM-patients NYHA I–II was more prevalent at the end: 158/174 (90.8%) vs. 81/99 (81.8%), P<0.05.
(b) Echocardiography: A similar behaviour was observed in both cohorts (Table 3). Absolute LVMI regression was higher in no PPM-patients and proportional to the preoperative LVMI value (higher regression if higher LVMI). In contrast, a lower regression was observed in PPM-patients, especially in those with higher LVMI. PPM impaired this proportional regression.
(c) Mortality: Fifty-five patients from Group A died after hospital discharge: 31/122 (25.4%) with PPM and 24/204 (11.8%) without PPM (P<0.001). A reduced survival from cardiac death was observed with PPM. When only non-cardiac death was considered, similar survival curves were obtained (Fig. 2). Causes of late death are shown in Fig. 2. ‘Moderate PPM’ and ‘EOA index’ were withheld as significant predictors for late cardiac mortality in contrast with ‘High LVMI’ (Table 6).
View larger version (12K):
[in this window]
[in a new window]
|
Fig. 2. (a–b) Free survival from cardiac (a) and non-cardiac death (b) in patients with and without prosthesis-patient mismatch (PPM). Only Group A was analyzed. (a) PPM vs. No PPM: log-rank: 18.52, df: 1, P<0.001, (b) PPM vs. No PPM: log-rank: 0.74, df: 1, P=0.38. Causes of late death: cardiac (valve and non-valve-related) in 35 patients: heart failure in 14, myocardial infarction in 5, sudden death in 7, thromboembolism in 3, anticoagulation-related bleeding in 5, endocarditis in 1; non-cardiac in 18 and unknown in 2 patients. (c–d) Free survival from cardiac (c) and from all-cause death (d) in patients with and without PPM after propensity score matching (patients from Group A and Group B), (c) PPM vs. No PPM: log-rank: 3.72, df: 1, P=0.05, (d) PPM vs. No PPM: log-rank: 1.61, df: 1, P=0.20.
|
|
3.6. PPM impact on survival after propensity score matching
After matching, PPM-patients only experienced a trend to higher 30-day mortality (6% vs. 4%, P=0.51). More PPM-patients died from cardiac causes at follow-up: 13/94 (13.8%) vs. 8/96 (8.3%), P=0.22, from all-cause death; 11/94 (11.7%) vs. 4/96 (4.2%), P=0.05, from cardiac death. Survival curves are shown in Fig. 2.
|
4. Discussion
|
|---|
PPM was defined by Rahimtoola as being present when the EOA of the prosthesis being implanted is less than that of the individual's normal native valve [8]. This situation is a direct consequence because the sewing ring, stents or leaflet housing will always reduce the available orifice area. The exception may be stentless valves implanted as root replacement (like homografts-autografts). However, the controversy remains from which extend such a reduction becomes relevant with a potential impact on early and late mortality, functional recovery and regression of ventricular hypertrophy.
Several studies have shown that severe PPM is negligible with modern-type valves [1, 9]. But these studies were published before the introduction of the latest generation supra-annular prostheses, most of them first implanted in recent years. With these new valves hemodynamic outcomes are even better. Thus, it can be expected that the incidence of PPM will decrease even more.
Latest generation bioprostheses have been designed for a totally supra-annular seating, which allows a precise alignment of the valve orifice to the patient's tissue annulus resulting in a maximization of flow [10, 11]. In a recent study, the supra-annular implant decreased the incidence of PPM from 50% to 34% (P<0.0001) [12]. Similarly, mechanical bileaflet valves with enhanced inner diameter may offer superior hemodynamic properties [13].
In our study, all these new prostheses have been predominantly implanted in the recent cohort resulting in a higher mean EOA index and a lower incidence of PPM. Group B and latest-generation prostheses were ‘protective’ factors for PPM. In contrast, elderly female patients with degenerative disease in which a bioprosthesis is implanted are typically at high risk. In these patients, PPM can largely be avoided using a preventive strategy [6].
Recent studies have documented higher early and long-term mortality with PPM [2, 14]. According to Rahimtoola, the problem with them is the use of manufacturer stated orifice size, different criteria for grading severity of PPM and not determining the causes of death [15]. We have used the projected EOA according to Pibarot [3, 6], without significant differences with observed values. These reference values have the advantage that can be attributed in retrospect in the absence of available echocardiographic data. In our patients we have observed a different trend in early vs. late and cardiac vs. non-cardiac mortality in relation to PPM. Its detrimental effect was predominant on late cardiac death. Other cause mortality was not affected and only a trend to higher 30-day mortality was observed. The deleterious effect of PPM was more evident in certain subgroups of patients (i.e. those with high LVMI). Thus, the impact of PPM on outcomes should be best addressed by subgroup analysis.
In conclusion, our study underlines that PPM may be an important prognostic factor for late cardiac death. A higher 30-day mortality was only observed in patients with increased LVMI and PPM. In survivors, subsequent LVMI regression was particularly impaired in these patients in contrast to no PPM-patients. Fortunately, a lower incidence of PPM has been observed in the recent cohort, especially with the latest generation prostheses. Future studies about new valves with detailed postoperative echocardiographic analyses must assess their real impact.
4.1. Limitations of the study
A retrospective, non-randomized single-center analysis over a long period and with a large variety of prostheses is subjected to the effects of selection bias. Although a propensity score method was used the sample size was reduced after matching. Clinical follow-up was too short and survival analysis was not considered in Group B. We could not address the impact of PPM on exercise capacity and quality of life. Future studies must address these issues.
|
References
|
|---|
- Medalion B, Blackstone EH, Lytle BW, White J, Arnold JH, Cosgrove DM. Aortic valve replacement: is valve size important? J Thorac Cardiovasc Surg 2000; 119:963–974.[Abstract/Free Full Text]
- Walther T, Rastan A, Falk V, Lehmann S, Garbade J, Funkat AK, Mohr FW, Gummert JF. Patient prosthesis mismatch affects short- and long-term outcomes after aortic valve replacement. Eur J Cardiothorac Surg 2006; 30:15–19.[Abstract/Free Full Text]
- Flameng W, Meuris B, Herijgers P, Herregods MC. Prosthesis-patient mismatch is not clinically relevant in aortic valve replacement using the Carpentier-Edwards Perimount valve. Ann Thorac Surg 2006; 82:530–536.[Abstract/Free Full Text]
- Bottio T, Caprili L, Casarotto D, Gerosa G. Small aortic annulus: the hydrodynamic performances of 5 commercially available bileaflet mechanical valves. J Thorac Cardiovasc Surg 2004; 128:457–462.[Abstract/Free Full Text]
- Fuster RG, Montero Argudo JA, Albarova OG, Sos FH, Lopez SC, Codoner MB, Buendia Minano JA, Albarran IR. Patient-prosthesis mismatch in aortic valve replacement: really tolerable? Eur J Cardiothorac Surg 2005; 27:441–449.[Abstract/Free Full Text]
- Pibarot P, Dumesnil JG, Cartier PC, Metras J, Lemieux M. Patient-prosthesis mismatch can be predicted at the time of operation. Ann Thorac Surg 2001; 71:S265–268.[CrossRef][Medline]
- Bonow RO, Carabello B, de Leon AC Jr, Edmunds LH Jr, Fedderly BJ, Freed MD, Gaasch WH, McKay CR, Nishimura RA, O'Gara PT, O'Rourke RA, Rahimtoola SH, Ritchie JL, Cheitlin MD, Eagle KA, Gardner TJ, Garson A Jr, Gibbons RJ, Russell RO, Ryan TJ, Smith SC Jr. Guidelines for the management of patients with valvular heart disease: executive summary: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines (Committee on management of patients with valvular heart disease). Circulation 1998; 98:1949–1984.[Free Full Text]
- Rahimtoola SH. The problem of valve prosthesis-patient mismatch. Circulation 1978; 58:20–24.[Abstract/Free Full Text]
- Izzat MB, Kadir I, Reeves B, Wilde P, Bryan AJ, Angelini GD. Patient-prosthesis mismatch is negligible with modern small-size aortic valve prostheses. Ann Thorac Surg 1999; 68:1657–1660.[Abstract/Free Full Text]
- Botzenhardt F, Eichinger WB, Bleiziffer S, Guenzinger R, Wagner IM, Bauernschmitt R, Lange R. Hemodynamic comparison of bioprostheses for complete supra-annular position in patients with small aortic annulus. J Am Coll Cardiol 2005; 45:2054–2060.[Abstract/Free Full Text]
- Totaro P, Degno N, Zaidi A, Youhana A, Argano V. Carpentier-Edwards Perimount Magna bioprosthesis: a stented valve with stentless performance? J Thorac Cardiovasc Surg 2005; 130:1668–1674.[Abstract/Free Full Text]
- Badano LP, Pavoni D, Musumeci S, Frassani R, Gianfagna P, Baldassi M, Tursi V, Mazzaro E, Zakja E, Fioretti PM, Livi U. Stented bioprosthetic valve hemodynamics: is the supra-annular implant better than the intra-annular? J Heart Valve Dis 2006; 15:238–246.[Medline]
- Albes JM, Hartrumpf M, Rudolph V, Krempl T, Huttemann E, Vollandt R, Wahlers T. Are mechanical valves with enhanced inner diameter advantageous in the small sized aortic annulus. Ann Thorac Surg 2003; 76:1564–1570.[Abstract/Free Full Text]
- Mohty-Echahidi D, Malouf JF, Girard SE, Schaff HV, Grill DE, Enriquez-Sarano ME, Miller FA Jr. Impact of prosthesis-patient mismatch on long-term survival in patients with small St Jude Medical mechanical prostheses in the aortic position. Circulation 2006; 113:420–426.[Abstract/Free Full Text]
- Rahimtoola SH. Is severe valve prosthesis-patient mismatch (VP-PM) associated with a higher mortality? Eur J Cardiothorac Surg 2006; 30:1.[Free Full Text]
This article has been cited by other articles:
|
|
|
|
 
S. Nozohoor, J. Nilsson, C. Luhrs, A. Roijer, and J. Sjogren
Influence of Prosthesis-Patient Mismatch on Diastolic Heart Failure After Aortic Valve Replacement
Ann. Thorac. Surg.,
April 1, 2008;
85(4):
1310 - 1317.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
Top
Abstract
1. Introduction
