Abstract
Resection of the chordopapillary apparatus during mitral valve replacement has been associated with a negative impact on survival. Mitral valve replacement with the preservation of the mitral valve apparatus has been associated with better outcomes, but surgeons remain refractory to its use. To determine if there is any real difference in preservation vs non-preservation of mitral valve apparatus during mitral valve replacement in terms of outcomes, we performed a systematic review and meta-analysis using MEDLINE, EMBASE, CENTRAL/CCTR, SciELO, LILACS, Google Scholar and reference lists of relevant articles to search for clinical studies that compared outcomes (30-day mortality, postoperative low cardiac output syndrome or 5-year mortality) between preservation vs non-preservation during mitral valve replacement from 1966 to 2011. The principal summary measures were odds ratios (ORs) with 95% confidence interval and P-values (that will be considered statistically significant when <0.05). The ORs were combined across studies using a weighted DerSimonian–Laird random-effects model. The meta-analysis was completed using the software Comprehensive Meta-Analysis version 2 (Biostat Inc., Englewood, NJ, USA). Twenty studies (3 randomized and 17 non-randomized) were identified and included a total of 3835 patients (1918 for mitral valve replacement preservation and 1917 for mitral valve replacement non-preservation). There was significant difference between mitral valve replacement preservation and mitral valve replacement non-preservation groups in the risk of 30-day mortality (OR 0.418, P <0.001), postoperative low cardiac output syndrome (OR 0.299, P <0.001) or 5-year mortality (OR 0.380, P <0.001). No publication bias or important heterogeneity of effects on any outcome was observed. In conclusion, we found evidence that argues in favour of the preservation of mitral valve apparatus during mitral valve replacement.
Keywords: Meta-analysis, Heart valve prosthesis implantation, Chordae tendineae
INTRODUCTION
Rationale
Resection of the chordopapillary apparatus during mitral valve replacement has been associated with a negative impact on survival and, on the contrary, more favourable outcomes have been observed with the use of subvalvular preservation techniques. However, most studies do not have probabilistic samples with sufficient sizes to demonstrate results with good power and/or good confidence intervals, which makes results be seen as not conclusive [1–20].
Our meta-analysis attempts to determine if there is any real difference in preservation vs non-preservation of mitral valve apparatus during mitral valve replacement in terms of outcomes.
Objectives
We performed a meta-analysis of studies to compare preservation (mitral valve replacement with preservation [MVR-P]) with non-preservation (mitral valve replacement with non-preservation [MVR-NP]) of mitral valve apparatus during mitral valve replacement, according to the preferred reporting items for systematic reviews and meta-analyses (PRISMA) statement [21].
MATERIALS and METHODS
Eligibility criteria
Using the population, intervention, comparison, outcome, study (PICOS) strategy, studies were considered if: (i) population comprised patients undergoing mitral valve replacement; (ii) compared outcomes between MVR-P and MVR-NP; (iii) outcomes studied included 30-day mortality, postoperative low cardiac output syndrome (LCOS) or 5-year mortality; (4) were prospective or retrospective or randomized or non-randomized studies.
Information sources
The following databases were used (from 1966 to 2011): MEDLINE, EMBASE, Cochrane Central Register of Controlled Trials (CENTRAL/CCTR), ClinicalTrials.gov, SciELO (Scientific Electronic Library Online), LILACS (Literatura Latino-Americana e do Caribe em Ciências da Saúde – The Latin American and Caribbean Health Sciences), Google Scholar and reference lists of relevant articles.
Search
We conducted the search using Medical Subject Heading terms ‘mitral disease’ OR ‘mitral stenosis’ OR ‘mitral regurgitation’ OR ‘mitral insufficiency’ OR ‘mitral valve disease’ OR ‘mitral valve prolapsed’ AND ‘mitral valve replacement’ OR ‘mitral valve surgery’ OR ‘heart valve prosthesis implantation’ AND ‘preservation of subvalvular apparatus’ OR ‘preservation of chordae tendineae’ OR ‘chordal-sparing’.
Study selection
The following steps were done: (i) identification of titles of records through database searching; (ii) removal of duplicates; (iii) screening and selection of abstracts; (iv) assessment for eligibility through full-text articles; (v) final inclusion in study.
One reviewer followed steps 1–3. Two independent reviewers followed step 4 and selected studies. Inclusion or exclusion of studies was unanimously decided. When there was disagreement, a third reviewer took the final decision.
Data items
The endpoint was the odds ratio (OR) for 30-day mortality after MVR-P or MVR-NP. Secondary endpoints were the OR for postoperative LCOS or 5-year mortality.
Data collection process
Two independent reviewers extracted the data. When there was disagreement about data, a third reviewer (the first author) checked the data and took the final decision on it. From each study, we extracted patient characteristics, study design and outcomes (number of events and number of total groups). When possible, actual probabilities of mortality after 5-year following MVR-P or MVR-NP were used to calculate ORs. Alternatively, probabilities of mortality were estimated from published Kaplan–Meier survival curves.
Risk of bias in individual studies
Included studies were assessed for the following characteristics: design (prospective or retrospective), presence of randomization (yes or no), multicenter enrolment (yes or no), characteristics of participants (selection bias), characteristics of personnel (performance bias), outcome assessment (detection bias), incomplete outcome data addressed (attrition bias) and match adjustment (yes or no).
Two independent reviewers assessed the risk of bias. Agreement between the two reviewers was assessed using κ statistics for full-text screening, and rating of relevance and risk of bias. When there was disagreement about the risk of bias, a third reviewer (the first author) checked the data and took the final decision on it.
Summary measures
The principal summary measures were ORs with 95% confidence interval and P-values (that will be considered statistically significant when <0.05). The meta-analysis was completed using the software Comprehensive Meta-Analysis version 2 (Biostat Inc., Englewood, NJ, USA).
Synthesis of results
Forest plots were generated for graphical presentations of clinical outcomes and we performed the I2 test and χ2 test for the assessment of heterogeneity across the studies [22]. Each study was summarized by the OR for MVR-P compared with MVR-NP. The ORs were combined across studies using a weighted DerSimonian–Laird random-effects model [23]. The model was weighted by the number of events in each study.
Risk of bias across studies
To assess the publication bias, a funnel plot was generated (for each outcome), which was statistically assessed by the Begg and Mazumdar's test [24] and the Egger's test [25].
RESULTS
Study selection
A total of 798 citations were identified, of which 34 studies were potentially relevant and retrieved as full text. Twenty publications fulfilled our eligibility criteria. Interobserver reliability of study relevance was excellent (κ = 0.84). Agreement for decisions related to study validity was very good (κ = 0.80). The search strategy can be seen in Fig. 1.
Figure 1:
Flow diagram of studies included in data search.
Study characteristics
The characteristics of each study are shown in Table 1. A total of 3835 patients were studied, with 1918 receiving MVR-P and 1917 receiving MVR-NP, including the years 1988–2010. Of the 20 studies, 7 used posterior or bileaflet preservation; 5 used only bileaflet preservation; 3 used only posterior preservation and 5 did not describe the type of preservation. Twelve studies were about mixed mitral pathology; 3 studies involved only rheumatic mitral disease; 2 were about reoperation; 1 study was about ischaemic mitral disease; 1 was only mitral insufficiency; 1 was not described. The overall internal validity was moderate and is illustrated in Table 2.
Table 1:
Study characteristics
| Study | MVR-P (n) | Type of mitral valve apparatus preservation | MVR-NP (n) | Mitral pathology |
|---|---|---|---|---|
| Zakai et al. [1] | 90 | Posterior/bileaflet | 32 | Mixed |
| García-Fuster et al. [2] | 410 | Posterior/bileaflet | 156 | Mixed |
| Muthialu et al. [3] | 98 | Posterior/bileaflet | 262 | Mixed |
| Chowdhury et al. [4] | 381 | Posterior/bileaflet | 70 | Mixed |
| Cingoz et al. [5] | 43 | Bileaflet | 51 | Mixed |
| Kayacioglu et al. [6] | 15 | Posterior | 15 | Rheumatic mitral stenosis |
| Borger et al. [7] | 286 | Bileaflet | 227 | Reoperation |
| Wasir et al. [8] | 168 | Bileaflet | 72 | Rheumatic |
| Kirali et al. [9] | 23 | Posterior | 33 | Rheumatic mitral insufficiency |
| Wu et al. [10] | 35 | Bileaflet | 33 | Mixed |
| Glower et al. [11] | 41 | Posterior/bileaflet | 219 | Mixed |
| Popovic et al. [12] | 6 | Bileaflet | 6 | Mitral insufficiency |
| Lee et al. [13] | 68 | NR | 318 | Mixed |
| Rao et al. [14] | 54 | Posterior/bileaflet | 187 | Reoperation |
| Straub et al. [15] | 42 | NR | 19 | Mixed |
| David et al. [16] | 70 | NR | 84 | NR |
| Dubiel et al. [17] | 7 | NR | 5 | Mixed |
| Horskotte et al. [18] | 50 | NR | 50 | Mixed |
| Hennein et al. [19] | 14 | Posterior/bileaflet | 55 | Mixed |
| Goor et al. [20] | 17 | Posterior | 23 | Ischaemic mitral insufficiency |
MVR-P: mitral valve replacement with preservation; MVR-NP: mitral valve replacement without preservation; NR: non-reported.
Table 2:
Analysis of bias risk: internal validity
| Study | Prospective design | Randomization | Multicenter enrolment | Selection bias | Performance bias | Attriction bias | Detection bias | Match adjustment |
|---|---|---|---|---|---|---|---|---|
| Zakai et al. [1] | Yes | No | No | A | B | A | B | Yes |
| García-Fuster et al. [2] | No | No | No | A | B | A | B | Yes |
| Muthialu et al. [3] | No | No | No | B | B | A | B | Yes |
| Chowdhury et al. [4] | Yes | No | No | B | B | A | B | Yes |
| Cingoz et al. [5] | Yes | No | No | A | B | A | B | Yes |
| Kayacioglu et al. [6] | Yes | No | No | B | B | A | B | No |
| Borger et al. [7] | No | No | No | B | B | A | B | No |
| Wasir et al. [8] | No | No | No | B | B | D | B | No |
| Kirali et al. [9] | Yes | Yes | No | B | B | B | B | Yes |
| Wu et al. [10] | Yes | Yes | No | C | B | B | B | Yes |
| Glower et al. [11] | No | No | No | C | B | B | B | No |
| Popovic et al. [12] | Yes | No | No | C | B | A | B | No |
| Lee et al. [13] | No | No | No | C | B | D | B | No |
| Rao et al. [14] | No | No | No | B | B | A | B | Yes |
| Straub et al. [15] | No | No | No | B | B | A | B | No |
| David et al. [16] | No | No | No | B | B | A | B | Yes |
| Dubiel et al. [17] | No | No | No | B | B | A | B | Yes |
| Horskotte et al. [18] | Yes | Yes | No | B | B | A | B | Yes |
| Hennein et al. [19] | Yes | No | No | B | B | A | B | Yes |
| Goor et al. [20] | No | No | No | B | B | A | B | No |
This was performed by two independent reviewers. The overall bias of the combined studies was considered moderate.
A: low bias risk; B: moderate bias risk; C: high bias risk; D: incomplete reporting.
Synthesis of results
The OR of the risk of 30-day mortality in the MVR-P group compared with the MVR-NP group in each study is reported in Fig. 2. There was no evidence of important heterogeneity of treatment effect among the studies of death. The overall OR (95% confidence interval) of 30-day mortality showed a statistically significant difference in favour of MVR-P compared with MVR-NP (random-effect model: OR 0.418, P <0.001).
Figure 2:
OR and conclusions plot of 30-day mortality associated with MVR-P vs MVR-NP.
The OR of the risk of postoperative LCOS in the MVR-P group compared with the MVR-NP group in each study is reported in Fig. 3. There was no evidence of important heterogeneity of treatment effect among the studies of death. The overall OR (95% confidence interval) of postoperative LCOS showed a statistically significant difference in favour of MVR-P compared with MVR-NP (random-effect model: OR 0.299, P <0.001).
Figure 3:
OR and conclusions plot of postoperative LCOS associated with MVR-P vs MVR-NP.
The OR of the risk of 5-year mortality in the MVR-P group compared with the MVR-NP group in each study is reported in Fig. 4. There was no evidence of important heterogeneity of treatment effect among the studies of death. The overall OR (95% confidence interval) of 5-year mortality showed a statistically significant difference in favour of MVR-P compared with MVR-NP (random-effect model: OR 0.380, P <0.001).
Figure 4:
OR and conclusions plot of 5-year mortality associated with MVR-P vs MVR-NP.
Risk of bias across studies
Funnel plot analysis (Fig. 5) disclosed symmetry around the axis for the treatment effect in all outcomes, which means we probably do not have a publication bias related to these endpoints.
Figure 5:
Publication bias analysis by a funnel plot for the outcomes.
DISCUSSION
Summary of evidence
The results of this meta-analysis demonstrate that there was a statistically significant difference in favour of MVR-P compared with MVR-NP in the OR for 30-day mortality, postoperative LCOS and 5-year mortality, the summary measures being free from the influence of heterogeneity of the effects or publication bias.
Considerations about mitral valve replacement with preservation
Despite the fact that existing evidence advocates the preservation of the mitral apparatus, it is not routinely performed. Surgeons are refractory to use preservation techniques because a possible left ventricular outflow obstruction may occur as preserved tissue impedes with prosthesis function, mainly in patients with septal hypertrophy undergoing anterior leaflet preservation [26]. Despite this aspect, our meta-analysis suggests that surgeons should perform, as much as possible, the preservation of mitral apparatus and use techniques to avoid and/or eliminate the problem of left ventricular outflow obstruction. We demonstrated that patients who underwent MVR-P experienced less risk of postoperative LCOS. In fact, MVR-P may not only impact left ventricular function, but also right ventricular function and this effect may be of importance as many patients undergoing MVR also suffer from right ventricular dysfunction [27].
Our findings associated with these considerations make us consider that the preservation of chordae tendineae is more important than we assume, so surgeons should pay more attention to this issue and as much as possible preserve the annuloventricular connection to achieve a better maintenance of ventricular geometry, myocardial dynamics, left and right ventricular function and, consequently, better results for patients in early- and long-term follow-up.
Risk of bias and limitations
There are inherent limitations with meta-analyses, including the use of cumulative data from summary estimates. Patient data were gathered from published data, not from the individual patient follow-up. Access to individual patient data would have enabled us to conduct further subgroup analysis and propensity analysis to account for differences between the treatment groups.
The studies included were of varying design, ranging from prospective randomized to retrospective studies. Only three studies were randomized and there is an important problem of the definition of ‘randomized’. All randomized studies did not describe the process of randomization, which makes doubt if the samples are truly randomized. The three studies also did not describe how blind they were (single, double, triple), which can generate significant performance and detection biases. There was a variation in selection criteria used by individual surgeons to allocate patients to each group in each study, and the two groups were not always fully matched for important risk factors. This meta-analysis included data from non-randomized observational studies, which reflects the ‘real world’, but they are limited by treatment bias, confounders and a tendency to overestimate treatment effects. Patient selection alters outcome and thus makes non-randomized studies obviously less robust.
Another limitation is the heterogeneity of the strategies across the studies. Among the studies used in this meta-analysis, we identified 10 different techniques of bileaflet and/or posterior leaflet preservation. This aspect may influence the results.
Finally, this meta-analysis does not establish whether there is any difference between bileaflet and posterior preservation during mitral valve replacement. This is an issue to be addressed in future individual studies and meta-analyses.
Conclusions
We found evidence that argues in favour of the preservation of the mitral valve apparatus during mitral valve replacement (as much as possible).
Conflict of interest: none declared.
REFERENCES
- 1.Zakai SB, Khan SUR, Rabbi F, Tasneem H. Effects of mitral valve replacement with and without chordal preservation on cardiac function: early and mid-term results. J Ayub Med Coll Abbottabad. 2010;22:91–6. [PubMed] [Google Scholar]
- 2.García-Fuster R, Estevez V, Gil O, Cánovas S, Martínez-Leon J. Mitral valve replacement in rheumatic patients: effects of chordal preservation. Ann Thorac Surg. 2008;86:472–81. doi: 10.1016/j.athoracsur.2008.04.046. [DOI] [PubMed] [Google Scholar]
- 3.Muthialu N, Varma SK, Ramanathan S, Padmanabhan C, Rao KM, Srinivasan M. Effect of chordal preservation on left ventricular function. Asian Cardiovasc Thorac Ann. 2005;13:233–7. doi: 10.1177/021849230501300309. [DOI] [PubMed] [Google Scholar]
- 4.Chowdhury UK, Kumar AS, Airan B, Mittal D, Subramaniam KG, Prakash R, et al. Mitral valve replacement with and without chordal preservation in a rheumatic population: serial echocardiographic assessment of left ventricular size and function. Ann Thorac Surg. 2005;79:1926–33. doi: 10.1016/j.athoracsur.2004.10.029. [DOI] [PubMed] [Google Scholar]
- 5.Cingoz F, Gunay C, Kuralay E, Yildirim V, Kilic S, Demirkilic U, et al. Both leaflet preservation during mitral valve replacement: modified anterior leaflet preservation technique. J Card Surg. 2004;19:528–34. doi: 10.1111/j.0886-0440.2004.200306.x. [DOI] [PubMed] [Google Scholar]
- 6.Kayacioglu I, Ates M, Sensoz Y, Gorgulu S, Idiz M, Kanca A. Comparative assessment of chordal preservation versus chordal resection in mitral valve replacement for mitral stenosis (long-term follow-up: 8 years) Tohoku J Exp Med. 2003;200:119–28. doi: 10.1620/tjem.200.119. [DOI] [PubMed] [Google Scholar]
- 7.Borger MA, Yau TM, Rao V, Scully HE, David TE. Reoperative mitral valve replacement: importance of preservation of the sub-valvular apparatus. Ann Thorac Surg. 2002;74:1482–7. doi: 10.1016/s0003-4975(02)03950-4. [DOI] [PubMed] [Google Scholar]
- 8.Wasir H, Choudhary SK, Airan B, Srivastava S, Kumar AS. Mitral valve replacement with chordal preservation in a rheumatic population. J Heart Valve Dis. 2001;10:84–9. [PubMed] [Google Scholar]
- 9.Kirali K, Tuncer A, Uyar I, Mansuroglu D, Daglar B, Ipek G, et al. Is posterior leaflet preservation in the surgical treatment of rheumatic mitral insufficiency without left ventricular dysfunction necessary? Cardiovasc Surg. 2001;9:58–63. doi: 10.1016/s0967-2109(00)00086-7. [DOI] [PubMed] [Google Scholar]
- 10.Wu ZK, Sun PW, Zhang X, Zhong FT, Tong CW, Lu K. Superiority of mitral valve replacement with preservation of sub-valvular structure to conventional replacement in severe rheumatic mitral valve disease: a modified technique and results of 1-year follow up. J Heart Valve Dis. 2000;9:616–22. [PubMed] [Google Scholar]
- 11.Glower DD, Landolfo KP, Davis RD, Cen YY, Harrison JK, Bashore TM, et al. Comparison of open mitral commissurotomy with mitral valve replacement with or without chordal preservation in patients with mitral stenosis. Circulation. 1998;98(19 Suppl):II120–3. [PubMed] [Google Scholar]
- 12.Popovic Z, Barac I, Jwic M, Ristic M, Panic G, Miric M, et al. Chordal preservation improves postoperative ventricular performance following valve replacement for chronic mitral regurgitation. Cardiovasc Surg. 1996;4:628–34. doi: 10.1016/0967-2109(96)00039-7. [DOI] [PubMed] [Google Scholar]
- 13.Lee EM, Shapiro LM, Wells FC. Importance of sub-valvular preservation and early operation in mitral valve surgery. Circulation. 1996;94:2117–23. doi: 10.1161/01.cir.94.9.2117. [DOI] [PubMed] [Google Scholar]
- 14.Rao V, Komeda M, Weisel RD, Ivanov J, Ikonomidis JS, Shirai T, et al. Results of represervation of the chordae tendineae during redo mitral valve replacement. Ann Thorac Surg. 1996;62:179–83. doi: 10.1016/0003-4975(96)00264-0. [DOI] [PubMed] [Google Scholar]
- 15.Straub U, Huwer H, Petzold T, Kalweit G, Volkmer I, Gams E. Preservation of mitral sub-valvular apparatus in multiple valve procedures. J Heart Valve Dis. 1995;4:477–82. [PubMed] [Google Scholar]
- 16.David TE, Armstrong S, Sun Z. Left ventricular function after mitral valve surgery. J Heart Valve Dis. 1995;4(Suppl II):S175–80. [PubMed] [Google Scholar]
- 17.Dubiel TW, Borowiec JW, Mannting F, Landelius J, Hansson HE, Nystrom SO, et al. Mitral valve prosthetic implantation with preservation of native mitral valve apparatus. Scand J Thorac Cardiovasc Surg. 1994;28:115–21. doi: 10.3109/14017439409099115. [DOI] [PubMed] [Google Scholar]
- 18.Horskotte D, Schulte HD, Bircks W, Strauer BE. The effect of chordal preservation on late outcome after mitral valve replacement: a randomized study. J Heart Valve Dis. 1993;2:150–8. [PubMed] [Google Scholar]
- 19.Hennein HA, Swain JA, McIntosh CL, Bonow RO, Stone CD, Clark RE. Comparative assessment of chordal preservation versus chordal resection during mitral valve replacement. J Thorac Cardiovasc Surg. 1990;99:828–36. [PubMed] [Google Scholar]
- 20.Goor DA, Mohr R, Lavee J, Serraf A, Smolinsky A. Preservation of the posterior leaflet during mechanical valve replacement for ischaemic mitral regurgitation and complete myocardial revascularization. J Thorac Cardiovasc Surg. 1988;96:253–60. [PubMed] [Google Scholar]
- 21.Moher D, Liberati A, Tetzlaff J, Altman DG. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. Ann Intern Med. 2009;151:264–9. doi: 10.7326/0003-4819-151-4-200908180-00135. [DOI] [PubMed] [Google Scholar]
- 22.Higgins JP, Thompson SG, Deeks JJ, Altman DG. Measuring inconsistency in meta-analyses. BMJ. 2003;327:557–60. doi: 10.1136/bmj.327.7414.557. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.DerSimonian R, Laird N. Meta-analysis in clinical trials. Control Clin Trials. 1986;7:177–88. doi: 10.1016/0197-2456(86)90046-2. [DOI] [PubMed] [Google Scholar]
- 24.Begg CB, Mazumdar M. Operating characteristics of a rank correlation test for publication bias. Biometrics. 1994;50:1088–101. [PubMed] [Google Scholar]
- 25.Egger M, Davey Smith G, Schneider M, Minder C. Bias in meta-analysis detected by a simple, graphical test. BMJ. 1997;315:629–34. doi: 10.1136/bmj.315.7109.629. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.DeCanniere D, Jansens J, Unger P, Le Clerc J. Left ventricular outflow obstruction after mitral valve replacement. Ann Thorac Surg. 1997;64:1805–6. doi: 10.1016/s0003-4975(97)01115-6. [DOI] [PubMed] [Google Scholar]
- 27.Le Tourneau T, Grandmougin D, Foucher C, McFadden EP, de Groote P, Prat A, et al. Anterior chordal transection impairs not only regional left ventricular function but also regional right ventricular function in mitral regurgitation. Circulation. 2001;104(12 Suppl 1):I41–6. doi: 10.1161/hc37t1.094602. [DOI] [PubMed] [Google Scholar]