Author + information
- Received April 2, 1999
- Revision received November 2, 1999
- Accepted December 15, 1999
- Published online April 1, 2000.
- Bernard Iung, MDa,* (, )
- Eric Garbarz, MDa,
- Pierre Michaud, MDa,
- Abdallah Mahdhaoui, MDa,
- Steeven Helou, MDa,
- Bruno Farah, MDa,
- Patricia Berdah, MDa,
- Pierre-Louis Michel, MDa,
- Yasuhiro Makita, MDa,
- Bertrand Cormier, MDa,
- Philippe Luxereau, MDa and
- Alec Vahanian, MDa
- ↵*Reprint requests and correspondence: Dr. Bernard Iung, Department of Cardiology, Hopital Bichat, 47 Rue Henri Huchard, 75018 Paris, France
The results of percutaneous mitral commissurotomy were assessed in patients with restenosis after surgical commissurotomy.
Balloon dilation is feasible in patients with restenosis after surgical commissurotomy, but little is known about its late efficacy.
We studied 232 patients who had undergone percutaneous mitral commissurotomy a mean of 16 ± 8 years after surgical commissurotomy. Mean age was 47 ± 14 years; 81 patients (35%) had valve calcification. All patients had restenosis with bilateral commissural fusion as assessed by echocardiography. Technical failure occurred in 9 patients and the procedure used a single balloon in 7 patients, a double balloon in 95, and the Inoue balloon in 121.
Complications were death in 1 patient (0.4%) and mitral regurgitation >2/4 in 10 (4%); 191 patients (82%) had good immediate results (valve area ≥1.5 cm2 without regurgitation >2/4). Predictors of poor immediate results in multivariate analysis were older age (p < 0.001), lower initial valve area (p = 0.01) and the use of the double-balloon technique (p = 0.015). In the 175 patients who underwent follow-up, 8-year survival without operation and in New York Heart Association class I or II was 48 ± 5%, and 58 ± 6% after good immediate results. In this latter group, poor late functional results were predicted by higher cardiothoracic index (p < 0.0001), previous open-heart commissurotomy (p = 0.05) and lower final valve area (p < 0.0001) in a multivariate Cox model.
Percutaneous mitral commissurotomy is safe and provides good immediate results in selected patients with restenosis after surgical commissurotomy. After good immediate results, the conditions of more than half of the patients remained improved at 8 years, enabling reoperation to be deferred.
Mitral restenosis after surgical commissurotomy is a topic of potential interest for percutaneous mitral commissurotomy (PMC) because in western countries, the proportion of recurrent mitral stenosis is increasing due to the decrease of new cases of rheumatic valve disease. Moreover, in the case of mitral restenosis, reoperation often consists of valve replacement with its inherent morbidity and mortality. Since the first report of PMC (1), its safety, immediate and midterm efficacy have been widely demonstrated (2–5). The feasibility of PMC has been documented in patients who had previously undergone surgical commissurotomy, with generally good immediate results (6–13). However, only one series reported midterm results and showed a more rapid deterioration at four years than in patients without prior commissurotomy (11).
The aim of the present study is to analyze immediate and late results of PMC for restenosis in a large single-center series reporting a long follow-up, of up to eight years. This enables the prediction of immediate and late results to be analyzed, in order to improve patient selection.
From March 1986 to June 1996, 232 patients underwent PMC for mitral restenosis after a previous surgical commissurotomy. They had undergone a total of 240 surgical commissurotomies: 184 patients had had one closed-heart commissurotomy (CHC), 40 had had one open-heart commissurotomy (OHC), and 8 had had repeat commissurotomies (repeat CHC in 1 patient, CHC then OHC in 7). Seven patients also had prior aortic valve replacement. All patients had experienced functional improvement following surgical commissurotomy. PMC was performed a mean of 16 ± 8 years (range, 3 to 38 years) after surgical commissurotomy: 17 ± 8 years after CHC and 11 ± 5 years after OHC.
The antegrade transvenous approach was used in all cases. There was a failure to perform an effective balloon dilation of the mitral valve in nine cases. Between 1986 and 1990, dilatation was performed using a single-balloon procedure in 7 patients (Trefoil 3 × 12 mm), and then a double-balloon procedure in 95 cases (Trefoil 3 × 10 mm or 3 × 12 mm + 15 or 19 mm balloon) (14). The last 121 patients underwent PMC with the Inoue balloon, which has been systematically used since 1990, following the stepwise technique under echo guidance (15).
Following echocardiography and fluoroscopy, mitral valve anatomy was classified in three anatomic groups as previously described (3,14,16): flexible valves and mild subvalvular disease (length of chordae ≥10 mm) (group 1); flexible valves and extensive subvalvular disease (length of chordae <10 mm) (group 2); and calcified valves as confirmed by fluoroscopy (group 3). In 40 patients of our global series, the mean (± SD) Wilkins’ score was 8.0 ± 0.8 (range, 7 to 9) for anatomic group 1, 9.9 ± 1.3 (range, 8 to 12) for group 2 and 12.5 ± 1.3 (range, 10 to 15) for group 3 (3).
When present, valve calcification was graded from 1 to 4 according to its extent on fluoroscopic examination: 45 patients were graded 1; 25 were graded 2; 9 were graded 3; and 2 were graded 4.
Echocardiographic examination was performed on the day preceding PMC and 24 to 48 h after the procedure to evaluate valve area by planimetry and Doppler pressure half-time, mean Doppler gradient and regurgitation by color Doppler. The reference measurement for valve area was planimetry by two-dimensional echocardiography. In cases of missing data, Doppler was used as substitution measurement (3). Beginning in 1988, transesophageal echocardiography was performed systematically before the procedure to eliminate left atrial thrombosis.
Before and immediately after the procedure, the degree of mitral regurgitation was assessed according to Sellers’ classification on left ventriculography in a 30° right anterior oblique view.
The effective balloon dilating area (EBDA) designs the area encompassed by the circumference of the outer halves of the two fully inflated balloons and the trapezoidal or square area connecting their two diameters for the double-balloon technique and the area at the last inflation for the Inoue balloon (17).
Exclusion criteria for PMC were bicommissural or massive calcification, mitral regurgitation > grade 2, left atrial thrombus and significant aortic valve or coronary artery disease.
In addition, patients were selected according to the status of the commissures as assessed by echocardiography on the parasternal short-axis view. Patients were considered for PMC only in the case of bilateral commissural fusion. When one or both commissures remained opened, restenosis was presumed to be mainly the consequence of a valvular and/or subvalvular rigidity and PMC was not attempted.
End point of good immediate results
Good immediate results of PMC were defined by a composite end point that associates a mitral valve area ≥1.5 cm2 with no regurgitation >2/4 (3,14).
Follow-up was performed prospectively at one-year intervals in patients who were residing in France at the time of their PMC and was assessed by patients’ visits to the department or by a standardized questionnaire sent to the patient’s cardiologist.
It was concluded in June 1997; patients were considered unavailable for follow-up if their last follow-up examination was before June 1996. Follow-up concerned 175 patients and its mean duration was 46 months (range, 1 to 122); 6 patients (3%) were unavailable for follow-up.
Clinical events were combined in the following end points:
• Survival, including all causes of mortality;
• Survival, taking into account only cardiovascular deaths;
• Survival with no need for intervention on the mitral valve (mitral surgery or repeat PMC); and
• “Good functional results,” that is, survival with no need for mitral surgery or repeat PMC, and in New York Heart Association (NYHA) functional class I or II.
Quantitative variables were expressed as mean ± SD. For the log-rank test, logistic and Cox models, they were divided into subgroups with clinically chosen cutoff points. Comparison before and after PMC used the paired Student t test. The predictive factors of immediate results were analyzed by univariate analysis using the unpaired Student t test for quantitative variables and the chi-square test for qualitative variables. Univariate analysis included the 11 preprocedure variables listed in Tables 1 and 2, ⇓⇓the type of balloon and the EBDA. Variables with p < 0.25 were entered in a multivariate logistic model and were selected by a backward procedure with a threshold of p = 0.05. The fit of the model was assessed by comparing the predicted and observed numbers of immediate results according to the method described by Hosmer and Lemeshow (17a). This logistic model was then used to predict the probability of good immediate results for any given patient, according to his or her characteristics and procedure-related variables.
Cumulative survival curves were determined according to the Kaplan-Meier method. Univariate analysis of the predictive factors of late results was performed using a log-rank test. The dependent variable was the previously described composite end point “good functional results.” Independent variables with p < 0.25 were entered in a multivariate Cox proportional hazards model and were selected by a backward procedure with a threshold of p = 0.05. A predictive model of continuing good functional results at five years was then established using the Cox model, in which the baseline survival function So(t) was estimated from the study population.
All analyses was performed with statistical software (SAS Institute Inc., release 6.11).
The characteristics of the 232 patients are detailed in Table 1. These patients represent 13% of the 1,741 procedures of PMC that were performed in our department during the same period.
The nine cases of failed procedures (4%) were related to tamponade in one case, inability to perform transseptal catheterization in two and to stabilize the balloon across the mitral orifice in six.
There was one in-hospital death (0.4%) in a 69-year-old woman who had previously undergone a transatrial CHC 21 years before and who had a massive tamponade immediately after the second balloon inflation using a double-balloon technique, suggesting a ventricular perforation. The other severe adverse events were severe mitral regurgitation ≥ Sellers’ grade 3 in 10 cases (4%). Six patients experienced transient ST-segment elevation after balloon inflation, suggesting gas embolism. There was no other embolic event, in particular no transient ischemic attack, and therefore no embolism leaving sequelae. No patient required surgery within the first 24 h following PMC.
Good immediate results as defined by the composite end point were obtained in 191 patients (82%). Poor immediate results were related to insufficient valve opening (valve area <1.5 cm2) in 22 cases (10%) and to mitral regurgitation ≥ Sellers’ grade 3 in 10 (4%). The improvement of valvular function is shown in Table 2.
The predictors of the immediate results in univariate analysis are shown in Table 3.
The results of multivariate analysis are detailed in Table 4: the significant predictors of poor immediate results were an older age (p < 0.0001), a lower valve area (p = 0.01) and the use of the single- or double-balloon procedure as compared with the Inoue balloon (p = 0.015). The good fit of the multivariate model to the observed data was demonstrated by the nonsignificant results of the Hosmer and Lemeshow goodness-of-fit test (chi-square = 6.81, df = 9, p = 0.66). The prediction of immediate results according to the value of the three predictors for any given patient is shown on Table 5.
In the 175 patients who underwent follow-up, the eight-year actuarial rates of global survival, survival considering only cardiovascular deaths, survival with no need for intervention and the composite end point of good functional results were, respectively, 78 ± 4%, 90 ± 3%, 55 ± 6%, and 48 ± 5% (Fig. 1).
In the 148 patients who had had good immediate results of PMC, the 8-year rates of global survival, survival considering only cardiovascular deaths, survival with no need for intervention, and the composite end point of good functional results were, respectively, 86 ± 4%, 97 ± 2%, 65 ± 6%, and 58 ± 6% (Fig. 2). Of these 148 patients, 13 died during follow-up. Three deaths were of cardiovascular origin, related to congestive heart failure, while the other 10 were noncardiac: respiratory insufficiency in 4, neoplasia in 3, cirrhosis in 1, AIDS in 1 and suicide in 1. In 24 patients, a subsequent intervention was required on the mitral valve. The type of intervention was an isolated mitral valve replacement in 13 cases, a mitral and aortic valve replacement in 3, an OHC in 3 and a repeat PMC in 5. The mean duration between PMC and the intervention was 46 months (range, 3 to 95). All the patients who underwent a new procedure on the mitral valve had mitral restenosis with a mean valve area of 1.1 ± 0.2 cm2 (range, 0.8 to 1.5); two patients also had moderate mitral regurgitation and three showed progression of an aortic valve disease. Finally, 11 patients were in NYHA class III or IV; they were waiting for an intervention or had contraindications to surgery.
The univariate analysis of the predictors of late functional results in the 148 patients who had had good immediate results of PMC is shown in Table 6. The multivariate Cox analysis identified three predictors of poor late functional results after good immediate results of PMC (Table 7). Two predictors were preprocedure variables, that is, a high cardiothoracic index (p < 0.0001) and a previous OHC (p = 0.05). The third predictor was a lower valve area after PMC (p < 0.0001). The prediction of good functional results 5 years after good immediate results of PMC is shown on Table 8.
Among the 27 patients who had had poor immediate results of PMC, there were 8 deaths, 7 from cardiovascular causes and 1 from extracardiac (neoplasia) causes. All the other 19 patients underwent surgery because of insufficient valve opening in 10 cases and mitral regurgitation ≥ Sellers’ grade 3 in 9. Surgery consisted of mitral valve replacement in 18 patients and OHC in 1 patient who had poor results from this surgery. The mean duration between PMC and the intervention was five months (range, 1 to 26), with 13 of the 19 patients (68%) being operated on within the first month following PMC.
This series further confirms the safety and immediate efficacy of PMC for restenosis after surgical commissurotomy, with 82% showing good immediate results. As concerns the late results, 48% of all patients were alive with no new intervention on the mitral valve, and with few or no symptoms at 8 years, with the corresponding figure being 58% after good immediate results. The prediction of the results is multifactorial: it depends on the patient characteristics, the technique used and the quality of the immediate results.
Restenosis after surgical commissurotomy
Progressive clinical deterioration has been shown to occur with a linear trend in different series reporting long-term follow-up after surgical commissurotomy (18–20). Mitral restenosis is one of the mechanisms of deterioration after surgical commissurotomy, the other causes being the consequence of unsatisfactory results of commissurotomy, with a residual stenosis or a severe mitral regurgitation (18,19). The incidence of restenosis can be estimated accurately only in studies comprising serial assessments of mitral valve area, both early after the procedure and during follow-up. The few studies available reported restenosis rates of 11% for a mean follow-up of 6.5 years (21) and 28% for a mean follow-up of 11 years (22).
In the present study as in the literature, the sustained functional improvement following surgical commissurotomy and the duration between surgical commissurotomy and PMC (mean of 11 to 17 years) suggested restenosis after an initially successful procedure, rather than a residual stenosis after surgery (6,7,10,12).
Safety of PMC
The low mortality of 0.4% in our series compares favorably with surgery. However, it should be stressed that the only death in our series was a tamponade, which can occur in patients who have undergone a previous surgical procedure despite frequent pericardial adherence (11). Particular attention must be given to patients who have undergone transventricular CHC, because of the possibility of a residual apical aneurysm (10). There was no embolism leaving sequelae; this can be related to the systematic use of transesophageal echocardiography since 1988. Despite the frequent exclusion of left atrial appendage after CHC, embolic risk is still present in these patients (23). The frequency of severe traumatic mitral regurgitation was 4% in the present series, which is close to the 3.4% observed in our global experience of PMC (3). Series of PMC generally do not report higher rates of severe mitral regurgitation in patients with prior commissurotomy as compared with first procedure (7–13).
Immediate results of PMC
The composite end point of good immediate results used in this study associates a mitral valve area ≥1.5 cm2 and no regurgitation >2/4, such conditions generally providing normal hemodynamics (3,14,24).
The immediate results observed in our series were good, with 82% of patients having a valve area ≥1.5 cm2 and no mitral regurgitation greater than Sellers’ grade 2. Series of PMC for restenosis reported excellent immediate results in young populations, with no difference when compared with patients who did not have prior surgical intervention (7,10,12,13). Age is a predictor of immediate results in large series of PMC (2,3), but its predictive value is even more marked in patients with a history of commissurotomy, as illustrated by the value of odds-ratio in this study. This is in accordance with our global series of PMC, in which we found an interaction between age and previous commissurotomy, that is, an increase in the risk of poor immediate results in the case of prior commissurotomy only in patients aged >50 years (3). This explains that results were less satisfying in series including patients with a mean age >50 years (11). As in other series of PMC, we found that valve area before PMC was a predictor of immediate results.
The last predictor of immediate results in our series was the technique used, with the Inoue balloon coming out better than the double-balloon technique. The published series of PMC for restenosis included only a few cases performed with the Inoue balloon (10,12,13). Patients who have a history of prior commissurotomy have more severe impairment of valve anatomy. Our findings are consistent with the work of Feldman et al. (25) who showed that the impairment of valve anatomy was not a predictor of poor results of PMC when the Inoue balloon was used. The possibility of controlling valve opening by stepwise inflation under echo guidance might be an explanation for the apparent benefit of the Inoue balloon in case of impaired valve anatomy.
Valve calcification and the duration between surgery and PMC have been identified as predictors in other series (7–9), but not in ours. It should be noted that valve calcification, when present, was generally mild or moderate in the patients of our series.
Late results of PMC
In the present series, half of the patients who had undergone PMC for restenosis remained alive, with no intervention on the mitral valve, and in NYHA class I or II, 8 years after the procedure. To our knowledge, only the article of Jang et al. (11) has so far reported midterm results of PMC in 68 patients who had recurrent mitral stenosis after a previous commissurotomy. They reported a rate of survival without intervention and in NYHA class I or II of 41% at four years, with a marked difference according to valve anatomy, the corresponding rates being 72% for patients with a Wilkins score ≤8 and 17% for a score >8. The comparison with our series must take into account the difference in the patient characteristics: patients were older, had more frequent valve calcification and less good immediate results than in the present series.
The quality of the immediate results is an important determinant of late results. The mechanism of poor late results differs according to whether they are related to the continuation of poor immediate results or to a secondary deterioration following an initially successful PMC. After a successful PMC, the major anatomic finding in the patients who had undergone a new intervention on the mitral valve was restenosis, either isolated or, less frequently, associated with mitral regurgitation or with the progression of aortic valve disease. Most patients underwent valve replacement, but the possibility of repeating PMC in selected patients is of interest in further deferring a surgical procedure.
In the present series, we identified three predictors of late deterioration after good immediate results of PMC. Of the two preprocedure related predictors, the most significant was cardiothoracic index. Cardiomegaly is a marker of the advanced stage of heart disease and it was a predictor of late outcome in a large series of CHC (20).
Another predictor of late deterioration after PMC for restenosis was the type of the initial commissurotomy. The risk of late deterioration was higher after OHC, as compared with CHC. A possible explanation is that the patients in our series who had initially undergone OHC generally had more severe valve disease than in the case of CHC. Other series of PMC for restenosis included only a few patients who had had OHC and did not specifically study those patients (7,10,12).
Mitral valve area after PMC has been shown to be predictive of late results after PMC in series including all procedures, regardless of the immediate results (4,5). Even when considering patients who had had good immediate results from PMC, valve area after the procedure remains a predictor of late outcome after PMC for restenosis.
Valve anatomy was not a predictor of late outcome in the present series, whereas it was singled out as the major determinant of late outcome in the only other series reporting midterm follow-up after PMC for recurrent mitral stenosis (11). In the particular field of recurrent mitral stenosis, we chose to select patients according to the echocardiographic analysis of commissural status in addition to the global evaluation of valve anatomy. Anatomic studies have shown that mitral restenosis after surgical commissurotomy may be associated with recurrent fusion of the commissures, but also to rigidity of the leaflets and/or the subvalvular apparatus, with commissural opening persisting from the first commissurotomy (26,27). The latter cases are obviously not suitable for an iterative commissurotomy (27). Besides the aforementioned differences between the populations involved, the scoring system used to assess valve anatomy differs between different series. Apart from the usual exclusion criteria, the selection of patients who had fusion of both commissures in the present series may have led to the exclusion of the patients with the most severe impairment of valve anatomy. However, it should be noted that no patient had ideal anatomic conditions (anatomic group 1). Therefore, the lack of patients with ideal anatomy, together with the small number of patients with severe valve impairment, probably narrowed the spectrum of anatomic presentations encountered in the present series and therefore limited the importance of valve anatomy in the predictive analysis.
Surgery for mitral restenosis
The severity of the valve disease observed in many patients with mitral restenosis and the desire to avoid a third intervention explains why mitral valve replacement is frequently performed in those patients (28,29). This reoperation carries an inherent risk of operative mortality and also prosthetic-related complications.
The feasibility of a repeat commissurotomy in case of restenosis was demonstrated when the only treatment of mitral stenosis was CHC. Morbidity and mortality were higher than for the first operation, but repeated commissurotomy could lead to further clinical improvement, in particular in young patients with noncalcified valves (30). Repeated CHC is still considered as a treatment of choice for restenosis in developing countries, in particular for economic reasons (23).
Good late functional results are obtained with OHC, but no series has specifically studied its results in case of restenosis, and its superiority over CHC is controversial (29,31). When performed for restenosis, OHC has the disadvantage of potentially leading to a third intervention for most patients.
The present findings demonstrate the safety and immediate efficacy of PMC in a wide range of patients with restenosis after surgical commissurotomy who were selected, in particular, on the basis of bilateral commissural fusion as assessed by echocardiography. After poor immediate results, valve replacement is generally required at short-term. In contrast, after good immediate results, PMC enables reintervention to be deferred for at least 8 years in more than half of the patients.
In addition, the analysis of the results shows that their prediction is intrinsically multifactorial and it therefore provides useful information in optimizing patient selection. The possibility of deferring surgery for many years is of particular importance in young patients without advanced heart disease, for whom no surgical alternative is totally satisfying. However, in older patients whose heart disease is at an advanced stage and in whom other characteristics predict a probable very poor outcome, surgical therapy can be considered as the first alternative in the absence of contraindications.
- closed-heart commissurotomy
- effective balloon dilating area
- New York Heart Association
- open-heart commissurotomy
- percutaneous mitral commissurotomy
- Received April 2, 1999.
- Revision received November 2, 1999.
- Accepted December 15, 1999.
- American College of Cardiology
- NHLBI Balloon Valvuloplasty Registry Participants
- Iung B,
- Cormier B,
- Ducimetière P,
- et al.
- Dean L.S,
- Mickel M,
- Bonan R,
- et al.
- Orrange S.E,
- Kawanishi D.T,
- Lopez B.M,
- et al.
- NHLBI balloon valvuloplasty registry participants,
- Davidson C.J,
- Bashore T.M,
- Mickel M,
- Davis K
- Lau K.W,
- Ding Z.P,
- Gao W,
- et al.
- Gupta S,
- Vora A,
- Lokhandwalla Y,
- et al.
- Vahanian A,
- Cormier B,
- Iung B
- Block P.C,
- Palacios I.F
- Hosmer D.W,
- Lemeshow S
- Ellis L.B,
- Singh J.B,
- Morales D.D,
- Harken D.E
- John S,
- Bashi V.V,
- Jairaj P.S,
- et al.
- Rihal C.S,
- Schaff H.V,
- Frye R.L,
- et al.
- Heger J.J,
- Wann L.S,
- Weyman A.E,
- et al.
- Iung B,
- Cormier B,
- Ducimetière P,
- et al.
- Feldman T,
- Carroll J.D,
- Isner J.M,
- et al.
- Kay P.H,
- Belcher P,
- Dawkins K,
- Lennox S.C
- Hickey M.S.J,
- Blackstone E.H,
- Kirklin J.W,
- Dean L.S