Novel Prognostic Score for Immediate and Late Success After Percutaneous Mitral Balloon Commissurotomy in Patients With Mitral Stenosis

Abstract: Objectives. Percutaneous mitral balloon commissurotomy (PMBC) remains the preferred treatment for patients with severe symptomatic rheumatic mitral stenosis (MS) and suitable anatomy. The objective of this study was to propose a new score for the prediction of immediate and late success. Methods. This is a single-center, retrospective analysis of all 1582 patients with severe mitral stenosis who underwent PMBC from August 1987 to July 2010. The composite outcome was cardiovascular death, new PMBC, or mitral valve repair surgery up to 24 years of follow-up. Results. Mean patient age was 36.8 ± 12.9 years, most (86.4%) were female, and Wilkins score was between 9-11 in 49.1% of patients. In the multivariate analysis, the predictors of immediate success were age (odds ratio [OR], 0.98; 95% confidence interval [CI], 0.96-0.99; P=.01), left atrium size (OR, 0.96; 95% CI, 0.93-0.99; P=.01), mean preprocedure mitral gradient (OR, 0.93; 95% CI, 0.89-0.96; P<.001), intermediate Wilkins score 9-11 (OR, 0.62; 95% CI, 0.40-0.94; P=.02), and high Wilkins score ≥12 (OR, 0.35; 95% CI, 0.16-0.76; P<.01). For prediction of late events, age (hazard ratio [HR], 0.98; 95% CI, 0.97-0.98; P<.001), New York Heart Association class III-IV (HR, 1.50; 95% CI, 1.18-1.92; P<.001), left atrium size (HR, 1.02; 95% CI, 1.02-0.04; P<.01), and high Wilkins score ≥12 (HR, 2.02; 95% CI, 1.30-3.15; P<.01) were significant. Two nomograms were developed using significant predictors from the model. Conclusions. In this large population, not only the Wilkins score, but also clinical and hemodynamic features, seem to be relevant in predicting immediate and late success for patients with rheumatic MS who underwent PMBC. 

J INVASIVE CARDIOL 2020;32(6):211-217. Epub 2020 April 9.

Key words: mitral valve stenosis, outcomes score, percutaneous mitral balloon commissurotomy

Percutaneous mitral balloon commissurotomy (PMBC) was first described in the 1980s, by Inoue, as an alternative to surgical commissurotomy for the treatment of mitral stenosis (MS).^1,2 The relatively high success rates (comparable to or greater than surgical commissurotomy),^3,4 and low mortality, as well as reductions in hospitalization length of stay and costs, are advantages that have made PMBC the current gold standard for MS treatment, whenever technically feasible.^5,6 Over the years, following the development of new techniques and instruments, the indications for the procedure have expanded to include patients with valve anatomy previously considered less favorable. 

Introduced in 1988, the Wilkins score remains the most widely used tool to select suitable candidates for PMBC. This score was developed based on only four preprocedure echocardiographic morphological features (mobility, calcification, valvular thickening, and subvalvular thickening) of 22 patients with rheumatic MS.³

Lately, alternative scores that include clinical and other echocardiographic variables have been proposed to better predict outcomes after this percutaneous procedure. However, most of these scores were derived from small populations and/or with limited follow-up. The current study therefore sought to introduce a novel score, based on the analysis of echocardiographic characteristics as well as clinical and hemodynamic variables, to predict immediate and late success after PMBC, in patients with rheumatic MS. 

Methods

This is a retrospective, single-center, single-arm registry encompassing all 1915 consecutive patients with rheumatic MS recruited and referred to PMBC between August 3, 1987 and July 19, 2010. All data were previously collected and recorded in a dataset. Patients who had previous mitral valve commissurotomy or replacement, missing immediate result data, or previous balloon valvuloplasty and/or concomitant aortic, tricuspid, or pulmonary valve stenosis were excluded (Figure 1). Clinical status was determined according to the New York Heart Association (NYHA) classification. 

Indications for PMBC were according to the guidelines at the time of the procedure, as follows: symptomatic patients with moderate or severe MS with favorable valve morphology, based on echocardiographic evaluation, including Wilkins score ≤8,³ asymptomatic patients with moderate to severe MS and pulmonary hypertension (systolic pulmonary artery pressure [PAP] >50 mm Hg at rest) with favorable valve morphology, and symptomatic patients with moderate or severe MS with unfavorable valve morphology but not eligible for surgery. Contraindications to PMBC were detailed previously and correspond to guidelines, including patients with moderate or greater mitral regurgitation (MR) or left atrial appendage thrombus.^4-6 

The study protocol and related materials were approved by our center’s institutional review board and ethics committee, and followed the principles of the 1975 Declaration of Helsinki. All subjects provided informed consent.

PMBC was performed by an antegrade transseptal approach using the Inoue,⁷ multitrack,⁸ double-balloon,⁷ or metallic commissurotomy⁹ techniques. Balloon size¹⁰ for the double-balloon technique was chosen according to the mitral annulus size or to obtain an effective balloon dilation area/body surface area of approximately 4 cm²/m², and one-step dilation was performed. Maximum balloon size¹² for the Inoue was determined by the following formula: 

(patient height [cm] / 10) + 10

An echo-Doppler study was performed in all patients before and 48 hours after PMBC. The evaluations included mitral valve area using two-dimensional echocardiography (or the pressure half-time method when the planimetry was not viable), as well as the estimation of MR, which was determined by the semiquantitative color-Doppler method¹³ and deemed absent (0), mild (1+), moderate (2+), moderately severe (3+), or severe (4+). 

In addition, the mitral valve morphology study was based on the semiquantitative analysis and ranked from discrete damage (degree 1) to severe damage (degree 4) for leaflet mobility, valve calcification, valve thickening, and subvalvar thickening. The sum of these scores (from 4 to 16 points) resulted in the echocardiographic score according to the criteria proposed by Wilkins and Block et al³ and categorized according to the score obtained: 1 for score ≤8; 2 for score between 9-11; and 3 for score ≥12.

Procedural success was defined as the achievement of postprocedure mitral valve area ≥1.5 cm² and angiographic MR <3+, in the absence of in-hospital major adverse cardiac and cerebrovascular events, including any death, stroke, mitral surgery, and cardiac tamponade. Long-term outcome was a composite of incidence of major adverse cardiac events up to 24 years of clinical follow-up (from 1988 until December 3, 2011), including overall mortality, need for new PMBC, or mitral valve replacement surgery. 

Clinical and echocardiographic evaluations were performed immediately after the procedure and annually after PMBC for up to 24 years. Clinical evaluation was performed by direct interview (history and physical examination) of the patient at outpatient visits. Adverse events were monitored throughout the study period. When the patient was lost to follow-up, the family doctor or personal cardiologist was contacted.

Statistical analysis. Continuous variables were reported as mean ± standard deviation and compared using Student’s t-test or the Wilcoxon test, as appropriate. Categorical variables were reported as percentages and compared using Chi-square or Fisher’s exact test. Spearman’s rank correlation (r) was used to determine the correlation between variables. Significant variables in the univariate analysis and clinically important variables were included in the logistic regression analyses to determine the simultaneous effect of multiple variables and to construct a multivariate model for predictors of immediate success after PMBC. Stepwise-backward method was used in order to access the reduced model. The adjustment of the final regression model was accessed using the Hosmer-Lemeshow goodness-of-fit adjustment test¹⁴ and the discriminatory power of the model was evaluated by the area under the receiver-operating characteristic (ROC) curve and the c-statistic test. ROC curves were also used to identify the point that maximizes overall sensitivity and specificity in predicting suboptimal results after the procedure. Likewise, for the occurrence of long-term events (overall mortality, new PMBC, or mitral valve repair surgery), a multivariate Cox proportional hazards regression model was used to evaluate the independent factors predicting long-term outcomes. Significant variables of known clinical importance were inserted into the Cox multivariable regression model, as well as the significant ones in the univariate analysis. Only events occurring after hospital discharge in patients with successful PMBC were included in the model to predict long-term outcomes. Long-term event-free survival was estimated using the Kaplan-Meier method.

A point-based scoring system was developed from the final multivariable logistic regression model (for immediate and long-term outcomes) in which a number of points were assigned to each predictor in the model (one for immediate success and another for late outcomes). The variables were then graded according to their odds ratio (OR) or hazard ratio (HR) with the creation of a linear predictor, and the final score was the sum of the points corresponding to each variable of the multivariable model. A nomogram¹⁵ was used for the calculation of event scores and clinical success for immediate success and long-term follow-up. 

A P-value <.05 was considered significant for all analyses. The data were analyzed using the Statistical Package for the Social Sciences, version 20.0 (IBM) and R Core Team 2017 (The R Foundation for Statistical Computing).

Results

The study population comprised 1582 patients. Mean age was 36.8 ± 12.9 years and the majority (86.4%) were female. Most patients at the time of PMBC had NYHA functional class III (57.7%) and Wilkins score between 9-11 (48.7%). In the analysis of Wilkins score, the majority of the patients presented with grade 2 mobility (61.7%), grade 3 valvular thickening (67.6%), grade 2 subvalvar thickening (69.6%), and grade 1 calcification (54.5%). Most patients (62.8%) had no MR. Tables 1-3 display the baseline clinical, echocardiographic, and procedural data for the overall population as well as comparisons between patients with and without procedural success. 

Immediate success was achieved in 1438 patients (90.9%). Failure was due to MR >2+ in 96 patients, non-achieved target mitral valve area in 29 patients, urgent cardiac surgery in 13 patients, cardiac tamponade in 4 patients, death in 1 patient, and stroke in 1 patient.

Independent predictors of immediate success included: age (OR, 0.98; 95% confidence interval [CI], 0.96-0.99; P=.01), left atrium size (OR, 0.96; 95% CI, 0.93-0.99; P=.01), mean preprocedure mitral gradient (OR, 0.93; 95% CI, 0.89-0.96; P<.001), intermediate Wilkins score 9-11 (OR, 0.62; 95% CI, 0.40-0.94; P=.02), and high Wilkins score ≥12 (OR, 0.35; 95% CI, 0.16-0.76; P<.01) (Table 4). 

To develop a simplified model, variables were then graded according to their ORs, ranging from 0 to 240. This score was used to plot the nomogram, and the higher the overall score, the greater the chance of immediate procedural failure (Figure 2). The accuracy of the model estimated with the ROC curve was 0.674 (P<.001) (Figure 3). Model calibration, estimated with the Hosmer-Lemeshow test, was Chi²=7.77 (P=.46).

Patients were followed annually after PMBC for up to 24 years. Long-term follow-up (median, 8.3 years; mean, 15.6 years; interquartile range, 1-24 years) was obtained in 1138 patients (79.1%). Patients lost to follow-up did not differ from the followed patients in terms of baseline clinical, echocardiographic, and procedural characteristics. During the follow-up period, rates of overall mortality, need for mitral surgery, and repeat PMBC were 0.6% (95% CI, 0.3-1.2), 8.3% (95% CI, 7.0-9.9), and 10.0% (95% CI, 8.5-11.7), respectively. The composite endpoint was 19.1% (95% CI, 17.0-21.1). 

Patients with NYHA I-II showed event-free survival of up to 80% at late follow-up (24 years), while NYHA III-IV showed event-free survival of nearly 60% (P<.001) (Figure 4A). Patients with Wilkins score ≥12 showed worse event-free survival (close to 40%), whereas patients with scores ≤8 and between 9-11 had close to 70% survival free of events at late follow-up (Figure 4A). 

Independent predictors of long-term major adverse cardiovascular events included age (HR, 0.98; 95% CI, 0.97-0.98; P<.001), NYHA III-IV (HR, 1.50; 95% CI, 1.18-1.92; P<.001), left atrium size (HR, 1.02; 95% CI, 1.02-0.04; P<.01), and high Wilkins score ≥12 (HR, 2.02; 95% CI, 1.30-3.15; P<.01) (Table 5).

As in the in-hospital phase, a prediction score of late success was built with attribution of points for each significant variable, obtaining a final score ranging from 0 to 160 points. This score is shown in the prognostic nomogram and a score was plotted toward long-term (15 years) follow-up (Figure 5). The accuracy of the model, measured with the ROC curve, was 0.66 (P<.001). Model calibration, estimated with the Hosmer-Lemeshow test, was Chi²=8.33 (P=.40). 

Discussion

Using data from our local registry, we identified almost six demographic, clinical, and angiographic features associated with in-hospital and long-term (up to 24 years) adverse events after PMBC. They were summarized into a simplified in-hospital and long-term risk score to support outcome comparisons and patient-level preprocedural risk estimation, respectively. 

The stratification of individual risk by using scores is useful in several aspects, including: (1) improved decision-making process; (2) treatment optimization; and (3) standardization of indications. Following the pioneering work of Wilkins and Block, other authors have proposed modified scores to predict PMBC success beyond clinical and hemodynamic variables, including three-dimensional (3D) echocardiographic variables.^16,17 

In 2010, Anwar et al¹⁶ proposed a novel score model based on 3D echocardiography, which enhances the detection of calcification and visualization of the commissural division. In that study, predictors of successful PMBC were valve leaflet mobility and the involvement of the subvalvular apparatus (whereas in the Wilkins score, it was the involvement of the subvalvular apparatus and leaflet calcification). The incidence and severity of MR was associated with higher calcification by the score. An important downside was the lack of clinical and hemodynamic variables.¹⁶ Although preprocedural commissure calcification has been previously associated with adverse effects immediately after PMBC when mitral valve morphology is considered, this parameter was not evaluated in our study.^18,19

More recently, Nunes et al proposed a new score including the echocardiographic variables identified by Anwar et al and adding baseline hemodynamic characteristics.^16,20 In terms of acute success, only echocardiographic variables were shown to predict outcomes, including the involvement of the subvalvular apparatus and the dislodgment of the leaflet ≤12 mm. Long-term outcome predictors included clinical parameters, such as age (OR, 1.03; 95% CI, 1.008-1.053; P<.01), left atrial volume after procedure (OR, 1.009; 95% CI, 1.003-1.015; P<.01), degree of MR (OR, 1.740; 95% CI, 1.092-2.774; P=.02), mean transvalvular mitral gradient after procedure (OR, 1.138; 95% CI, 1.016-1.273; P=.02), and mean pulmonary artery pressure after procedure (OR, 1.035; 95% CI,1.008-1.062; P=.01).^15,19

In the same line as our research, Cruz-Gonzalez et al²¹ presented a multifactorial score based on a relatively large population (n = 1085) with 3 years of follow-up. This score considered clinical, echocardiographic, and hemodynamic characteristics. Six independent predictors of PMBC success were identified: age <55 years, NYHA class I and II, previous mitral valve area ≥1 cm², previous MR <2+, echocardiographic score <8, and male gender. This score demonstrated greater sensitivity and specificity than the echocardiographic score. It should be noted that mean age in the aforementioned study was 54.9 years, as opposed to the younger profile of our sample (36.8 years). This difference may lead to different scenarios, since in Brazil and other developing countries, rheumatic fever is responsible for more than 90% of cases of MS, affecting between the third and fourth decade of life, whereas in more developed countries, cases of degenerative MS associated with mitral ring calcification and other causes are increasing.²¹ 

The American Heart Association and European Society of Cardiology²² and the latest European Association for Cardio-Thoracic Surgery guidelines on valvular heart disease^4,5 delineate the unfavorable characteristics for PMBC, including the presence of several clinical (old age, history of commissurotomy, NYHA class IV, permanent atrial fibrillation, severe pulmonary hypertension) and anatomical characteristics (echocardiographic score >8, Cormier score 3 [calcification of the mitral valve to any extent as assessed by fluoroscopy], very small mitral valve area, and severe tricuspid regurgitation), which corresponds, in part, to our analysis.

Study limitations. This is a single-center report from a tertiary institution with operators who are highly experienced in PMBC. Therefore, the extrapolation of our findings to other sites should be viewed with caution. Given the retrospective nature of the analysis, it is difficult to adjust for variables in the multivariable analysis. The loss to follow-up carries an intrinsic bias, but the analysis indicates that it is randomly assigned. The majority of our sample is represented by patients with MS secondary to rheumatic fever. Furthermore, patients were younger than those included in similar series. Thus, the extrapolation of our findings to other geographies and populations worldwide should be viewed with caution. Finally, the new proposed score requires external validation, in a prospective fashion. Our group is working on this validation; however, the marked decrease in the occurrence of rheumatic MS has slowed the validation process. 

Conclusion

In this large, single-center cohort, it became apparent that the established Wilkins score is not the only relevant factor in predicting immediate and late success for patients with rheumatic MS treated with PMBC. Echocardiographic parameters, as well as clinical and hemodynamic features, also significantly contribute to the acute procedural success, as well as to the long-term maintenance of the results. 

From the ¹Department of Interventional Cardiology, Instituto Dante Pazzanese de Cardiologia, Sao Paulo, Brazil; ²Department of Medicine, Universidade Federal de Sergipe, Lagarto, Brazil; ³Department of Structural Heart Disease, Instituto Dante Pazzanese de Cardiologia, Sao Paulo, Brazil; and ⁴Department of Biostatistics, Instituto Dante Pazzanese de Cardiologia, Sao Paulo, Brazil.

Disclosure: The authors have completed and returned the ICMJE Form for Disclosure of Potential Conflicts of Interest. The authors report no conflicts of interest regarding the content herein.

Manuscript submitted September 7, 2019, final version accepted September 26, 2019.

Address for correspondence: J. Ribamar Costa Jr, MD, PhD, Av Dante Pazzanese, n. 500, 04012-909 Sao Paulo, Brazil. Email: rmvcosta@uol.com.br

1. Inoue K, Owaki T, Nakamura T, Kitamura F, Miyamoto N. Clinical application of transvenous mitral commissurotomy by a new balloon catheter. J Thorac Cardiovasc Surg. 1984;87:394-402.
2. Nobuyoshi M, Arita T, Shirai S, et al. Percutaneous balloon mitral valvuloplasty: a review. Circulation. 2009;119:e211-e219.
3. Wilkins GT, Weyman AE, Abascal VM, Block PC, Palacios IF. Percutaneous balloon dilatation of the mitral valve: an analysis of echocardiographic variables related to outcome and the mechanism of dilatation. Br Heart J. 1988;60:299-308.
4. Nishimura RA, Otto CM, Bonow RO, et al. 2014 AHA/ACC guideline for the management of patients with valvular heart disease: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines. J Am Coll Cardiol. 2014;63:e57-185.
5. Baumgartner H, Falk V, Bax JJ, et al. 2017 ESC/EACTS guidelines for the management of valvular heart disease. Eur Heart J. 2017;38:2739-9127.
6. Iung B, Cormier B, Ducimetiere P, et al. Immediate results of percutaneous mitral commissurotomy. A predictive model on a series of 1514 patients. Circulation. 1996;94:2124-2130.
7. Feldman T, Herrmann HC, Inoue K. Technique of percutaneous transvenous mitral commissurotomy using the Inoue balloon catheter. Cathet Cardiovasc Diagn. 1994;(Suppl 2):26-34.
8. Bonhoeffer P, Piechaud JF, Sidi D, et al. Mitral dilatation with the multi-track system: an alternative approach. Cathet Cardiovasc Diagn. 1995;36:189-193.
9. Cribier A, Eltchaninoff H, Carlot R, et al. Percutaneous mechanical mitral commissurotomy with the metallic valvulotome: detailed technical aspects and overview of the results of the multicenter registry on 882 patients. J Intervent Cardiol. 2000;13:255-262.
10. Kang DH, Park SW, Song JK, et al. Long-term clinical and echocardiographic outcome of percutaneous mitral valvuloplasty: randomized comparison of Inoue and double-balloon techniques. J Am Coll Cardiol. 2000;35:169-175.
11. Kirchhof P, Benussi S, Kotecha D, et al. 2016 ESC guidelines for the management of atrial fibrillation developed in collaboration with EACTS. Europace. 2016;18:1609-1678.
12. Palacios IF. Percutaneous mitral balloon valvuloplasty. In: Sievert H, ed. Percutaneous Interventions for Congenital Heart Disease. London: Informa Healthcare, 2007:177-184.
13. Helmcke F, Nanda NC, Hsiung MC, et al. Color Doppler assessment of mitral regurgitation with orthogonal planes. Circulation. 1987;75:175-183.
14. Hosmer DW, Hosmer T, Le Cessie S, Lemeshow S. A comparison of goodness-of-fit tests for the logistic regression model. Stat Med. 1997;16:965-980.
15. Nomograms. In: Encylopedia of Statistical Sciences. Kotz S, Johnson NL, Read CB, eds. New York: Wiley, 1985.
16. Anwar AM, Attia WM, Nosir YF, et al. Validation of a new score for the assessment of mitral stenosis using real-time three-dimensional echocardiography. J Am Soc Echocardiogr. 2010;23:13-22.
17. Soliman OI, Anwar AM, Metawei AK, McGhie JS, Geleijnse ML, Ten Cate FJ. New scores for the assessment of mitral stenosis using real-time three-dimensional echocardiography. Curr Cardiovasc Imaging Rep. 2011;4:370-377.
18. Wei T, Zeng C, Chen F, et al. Influence of commissural calcification on the immediate outcomes of percutaneous balloon mitral valvuloplasty. Acta Cardiol. 2003;58:411-415.
19. Sutaria N, Northridge DB, Shaw TR. Significance of commissural calcification on outcome of mitral balloon valvotomy. Heart. 2000;84:398-402.
20. Nunes MC, Tan TC, Elmariah S, et al. The echo score revisited: impact of incorporating commissural morphology and leaflet displacement to the prediction of outcome for patients undergoing percutaneous mitral valvuloplasty. Circulation. 2014;129:886-895.
21. Cruz-Gonzalez I, Sanchez-Ledesma M, Sanchez PL, et al. Predicting success and long-term outcomes of percutaneous mitral valvuloplasty: a multifactorial score. Am J Med. 2009;122:581.e11-e19.
22. Hernandez R, Banuelos C, Alfonso F, et al. Long-term clinical and echocardiographic follow-up after percutaneous mitral valvuloplasty with the Inoue balloon. Circulation. 1999;99:1580-1586.