Mitral Stenosis and Percutaneous Mitral Valvuloplasty

Enio E. Guarios, MD, PhD, Ronaldo Bueno, MD, PhD, Deborah Nercolini, MD, Jos© Tarastchuk, MD, Paulo Andrade, MD, Alvaro Pacheco, MD, Alysson Faidiga, MD, Stefan Negrao, MD*, Antonio Barbosa, MD*
Enio E. Guarios, MD, PhD, Ronaldo Bueno, MD, PhD, Deborah Nercolini, MD, Jos© Tarastchuk, MD, Paulo Andrade, MD, Alvaro Pacheco, MD, Alysson Faidiga, MD, Stefan Negrao, MD*, Antonio Barbosa, MD*
The prevailing cause of mitral stenosis (MS) is a progressive reaction to an injury caused by one of the major manifestations of rheumatic fever (RF): rheumatic carditis. The mitral is the valve most commonly prone to suffer a rheumatic process. Therefore, the prevalence of MS reflects the frequency of RF in the community and, consequently, in the region.1,2 In developed countries, with the decline of new RF cases to 0.5 cases per 100,000 inhabitants in the early 1980s,3 rheumatic cardiac diseases began to be seen as geriatric diseases.4 However, from 1987 on, the ratio and distribution of invasive streptococcus serotypes have increased, since RF cases reappeared in those countries.5 In the poorest and densely populated areas of our planet, however, the devastating effects of RF remained constant, since its prevalence in poor countries is at least 10 times higher than in developed countries.6 Additionally, in lower income population areas, the course of MS differs much from the classic one. Due to a more severe initial rheumatic bout or to repeated bouts of rheumatic carditis secondary to recurrent streptococcal infections, the progression of MS in those areas tends to be much faster, leading to functional disability at a much earlier stage.7,8 In underdeveloped countries, MS is a major public health issue. Early intervention and therapeutical advances totally changed the scenario for MS patients in the past 30 years. The excellent results achieved with the closed mitral commissurotomy for treating MS led to the natural development of a percutaneous technique so that the procedure could be performed by a nonsurgical approach. Previous surgical experience in treating MS was the conceptual basis for using a balloon for mechanical dilation of the mitral valve. Important contributors to the technique were the introduction of Brockenbrough’s technique for transseptal catheterization,9 and Rashkind’s and Miller’s balloon atrial septostomy.10 Rashkind and Miller’s work inspired Inoue, a Japanese cardiac surgeon, to design and develop a coaxial double-lumen catheter to create interatrial communication in children with transposition of the great vessels, tricuspid atresia, and other congenital cardiopathies.11 Later, Inoue et al. tested the balloon’s potential for separating fused commissures of the mitral valve, as an adjuvant procedure in open surgery.12 In June 1982, through percutaneous introduction of his balloon, Inoue et al. successfully dilated a stenotic mitral valve in a young patient with disabling symptoms,13 and inaugurated the percutaneous mitral valvuloplasty (PMV) era. At first, the application of Inoue’s technique was restricted to the East, while elsewhere in the world, cylindrical balloons, initially developed for pulmonary valvuloplasty, were adapted for mitral valvuloplasty.14 In 1985, in New Delhi, Lock et al. used those balloons to treat 8 young patients with rheumatic MS.15 Soon after, Al Zaibag et al., from Saudi Arabia, introduced the double-balloon (DB) technique.16 Those publications addressed exclusively the treatment of young patients with slightly distorted mitral anatomy. In 1986, McKay et al.17 and Palacios et al.18 simultaneously reported the first PMVs performed in adult patients with calcified mitral valves. Both used single 25 mm diameter balloons and achieved good results. Several reports of individual and multicentric experiences followed, consolidating the therapeutic role of PMV in MS patient management. At first, the DB technique was predominantly used. Soon after, especially after the Inoue balloon (IB) was approved by the FDA in 1994,19 Inoue’s technique became the prevailing one for percutaneous dilation of the mitral valve worldwide because it was simpler, faster, and yielded similar benefits with a lower complication rate when compared to the DB technique. 2. Patient selection. By consensus, PMV is indicated for symptomatic patients [NYHA functional class (FC) II] who have moderate or severe MS [mitral valve area (MVA) 1.5 cm2] and a valve morphology that favors percutaneous intervention, with no thrombus in the left atrium (LA) or moderate or severe mitral regurgitation (MR).13 Although there are studies suggesting that PMV is safe and effective in asymptomatic patients,20 the procedure is not without risk and therefore should not be routinely indicated in such situations. Exceptions to this rule are patients with severe MS who require other major non-cardiac surgery, young women who wish to get pregnant, and patients at high risk for thromboembolism (former history of thromboembolic phenomena, dense spontaneous contrast in LA, recurring atrial fibrillation).21,22 Likewise, despite the good preliminary results,23 there is no consensus about the indication of such a procedure in patients with mild or moderate MS, in an attempt to delay the progression of the disease, especially because those patients respond well to clinical treatment.22 Because it is widely available and enables dynamic assessment of the mitral valve, allowing for pulmonary pressure estimation and for determining the concurrence of other valve disorders, the Doppler two-dimensional echocardiography is the diagnostic tool of choice for evaluation of a MS patient. The interest in pre-valvuloplasty echocardiography derives from the significant previous surgical experience showing that the success of a surgical mitral commissurotomy is determined by valve morphology.24 Within this line of thought, in an attempt to find predictors for PMV immediate results, Wilkins et al. described a morphological score that graded morphological changes of the mitral valve, including 4 characteristics: leaflet mobility, leaflet thickening, valve calcification, and involvement of the subvalvular apparatus each of them classified in a 0“4 scale. The authors demonstrated with that paper that the only predictor of immediate results after PMV, regardless of any other clinical or hemodynamic variable, was the valve’s total score.25 This score was validated by comparing echocardiographic findings with the anatomopathological exam in a series of autopsies.26 Extending the initial case load to 10027 and 130 patients,28 this group of authors determined that the optimal combination point between sensitivity and specificity (72% and 73%, respectively) to forecast a good immediate PMV result was a score of 8. Because it enables a quantitative assessment of how severely the mitral valve apparatus is affected by rheumatic disease, and because it predicts the immediate result of PMV, the Wilkins score became standard, providing the means to compare populations from different studies. Several limitations, however, were brought up about the score, including the fact that all its components have the same weight in the pathological process. Even the group that proposed the score have established that, among its components, the only one that correlates with an absolute change in MVA after valvuloplasty was thickening of the valve.28 Findings, however, differ among the authors and there are papers defining thickening,29-31 mobility,32 calcification,33 and involvement of the subvalvular apparatus34,35 as the single variable that best defines the score’s predictive value. Furthermore, the score does not take into account factors that are important in predicting results, such as asymmetry of fused commissures36 and their degree of calcification,37 and it fails to predict severe MR. Therefore, other scores were devised, some addressing valve anatomy more thoroughly, others grading specific valve characteristics.21,29,32,33,36,38,39 None of them, however, has shown to be better than others and, to different extents, all of them are limited in terms of reproducibility, underestimation of certain characteristics, and inability to predict MR development.22 3. Technical considerations. Antegrade access to the mitral valve through transseptal puncture is the standard approach for the most widely used PMV techniques (Inoue, DB, metallic commissurotome, single-balloon, Multi-Track system). There are four techniques that approach the mitral valve through retrograde arterial access, all of them with restricted application. The technique described by Babic et al. requires transseptal advancement of a guidewire inserted in the femoral vein, to be exteriorized and recovered and at a site on the femoral artery. Subsequently, the balloon is advanced retrogradely over the guidewire through the mitral valve.40 Bachler et al.41 and Orme et al.42 have described retrograde techniques for mitral dilation without transseptal catheterization, using a single balloon and DB, respectively. Stefanadis et al. developed a different technique using catheters specifically designed to retrogradely pass through the mitral valve. Notwithstanding a multicentric registry encompassing more than 400 patients treated this way,43 this technique is mostly restricted to Greece, where the authors have their practice. The theoretical indication of retrogradely approaching the mitral valve would be treating those patients unsuitable for transseptal access due to anatomical deformities, or as an alternative to complete the procedure after failure to puncture the septum. On the other hand, retrograde atrial catheterization is not always easy, as there may be a greater risk of injuring the chordae tendineae and pulmonary veins, and the technique may not be used in patients who have an aortic valve prosthesis.22 Regardless of the technique employed, the PMV’s mechanism of action is the same and has been described initially by Inoue et al., who, during mitral valve surgery, directly viewed the splitting of fused commissures in patients submitted to PMV just before valve replacement.12 Such a finding was later confirmed in autopsies and in echocardiographic and experimental studies.44-47 In calcified valves, Reifart et al. also demonstrated an increase in MVA secondary to fractures of nodular calcium deposits after PMV with a balloon. There was, however, no embolism secondary to the release of tissue fragments. In the same study, the authors observed that valve stretching acts as an immediate dilation mechanism.48 Valves with calcified commissures, rigid leaflets, and subvalvular disease are associated with smaller increases in MVA.47 Despite the inherent limitation of positioning a cylindrical balloon through an elliptic orifice, the Inoue technique became standard in most institutions. Unlike other techniques, PMV with the Inoue balloon does not require floating balloons or the placement of a guidewire in the left ventricle (LV), and is therefore a simpler procedure with a substantially lower risk of perforating the LV. In addition, the balloon has a thinner profile, is shorter in length, easier to guide, and stably sets in the valve orifice during dilation, thanks to its hourglass shape49,50 (Figure 1). The technique’s major limitation is the cost of the balloon. Although it has proven to be efficient, the DB technique (Figure 2) is practically no longer used due to the greater complexity and potential higher morbidity of the procedure. The Multi-Track system, devised by Bonhoeffer et al. in 1995, simplifies the DB technique.51 With this system, one of the balloons is a rapid exchange balloon, while the other has a conventional design, enabling both to be aligned in the mitral valve orifice over a single guidewire. The experience published on the use of this technique, however, is still very restricted.51,52 Also, certain institutions still prefer to perform PMV with a single 25 or 30 mm diameter balloon.53 The main advantage of such techniques is lower cost, not only regarding the balloon, but also because they can be reused after resterilization with ethylene oxide. In 1995, the metallic commissurotome technique (Figure 3) was developed. The principle is basically the same as the Tubbs dilator used for closed mitral commissurotomy (CMC), and its main advantage is that its performance is not affected after being reused several times.54 Additional advantages are that the mitral valve orifice is not totally obstructed during dilation as it is with balloons,55 and that the device’s opening arms are directed along the valve’s commissural line, with no pressure being exerted on the leaflets or on the subvalvular apparatus.56 Because the technique’s attractiveness is its lower cost, it has especially been used in developing countries. A registry exists of more than 14,000 patients submitted to PMV with the metallic commissurotome worldwide. 4. Immediate results and predictors. Depending on the characteristics of the treated population and on the definition used, the success rate of PMV ranges between 80% and 95%. Technical failure rates vary between 0.8% and 2.3%, due to the difficulty of transseptal puncture or of passing through the mitral valve. In successful procedures, the MVA usually doubles, and the transvalvar gradient drops to 50“60% of its initial value just after dilation.3,8 Valve function improvement results in an immediate drop in pressure and pulmonary vascular resistance, and in a slight increase of the cardiac index and of the final LV diastolic volume.22 The pulmonary vascular resistance gradually and continually drops in the first 24 hours, and continues to drop on a long-term basis if there is no restenosis. Function of the right ventricle improves with the increase of its systolic volume index.57 There is also an improvement in the LV ejection fraction (LVEF), especially in patients with a LVEF ? 0.55.58-60 Also, there is an increase in the left atrial appendage peak Doppler velocity related to a decrease or regression of spontaneous echo contrast in the LA. This finding suggests a beneficial effect of the PMV on blood stasis and on the causes of thromboembolism.61,62 Improvement in exercise capacity may be detected in several weeks because of the gradual regression of the chronic effects of subperfusion on the skeletal muscles and of congestion on the pulmonary function.63,64 On the other hand, when the immediate results of the procedure are not satisfactory, the patient’s functional improvement is either transient or nonexistent.21 There are multiple predictors of the immediate results of the procedure. At first, valve anatomy, defined by the echo score, used to be considered the main predictor of good results following PMV.4,25,28,65-68 It seems more likely, however, that the echo score is just a relative predictor. In addition to valve morphology, other independent predictors of immediate results include age of the patient,4,29,49,69,70 functional class,4,65,66,69 previous surgical commissurotomy,4,43,49,70 calcification at fluoroscopy,4,69,71 initial MVA,21,29,70,72 previous MR,22,43 symmetry of commissural fusion,36 and procedure-related variables such as balloon size or extent of opening of the device used.32,49,72,73 Better than individually, all those factors together play an important role in postintervention results. Predictive models derived from such data, however, are still not capable of accurately identifying which patients will develop severe MR after PMV.22 5. Complications. The National Heart, Lung and Blood Institute (NHLBI) PMV registry (1992) classified PMV-related complications as major (death, shock, severe MR, systemic embolism, cardiac tamponade, emergency surgery, and acute myocardial infarction) and minor (vasovagal reaction, prolonged hypotension, arrhythmia requiring treatment, significant atrial septal defect -ASD, pericardiocentesis, blood transfusion), and the global incidence of major complications was about 12%.74 The incidence rate as a whole is higher in multicentric studies or in early experience reports than in single-center studies with a large number of cases. This reflects the importance of receiving specific training to perform mitral valve intervention.22,74,75 In most of the series, mortality ranges between 0“3%, with the lower rates having been reported more recently. The major causes of death are cardiac tamponade, severe MR, and deterioration of the patient’s general condition.22,76,77 In a subanalysis of the NHLBI’s registry, PMV-related mortality clearly correlated with patient selection for the procedure: comparing patients at high and low surgical risk, the immediate and 30-day mortality rate was 2% versus 0.5%, and 8% versus 0.6%, respectively.78 The occurrence of hemopericardium, at average rates of 1-3%, is related to transseptal puncture or perforation of the LV’s apex by the guidewire or by the balloon itself.79 Cardiac perforation was directly related to patient age, and inversely related to the surgeon’s experience.80 Right atrium perforations are usually not that severe, and most of them can be managed by pericardiocentesis and protamine administration. Left ventricle lacerations, however, produce immediate hemodynamic deterioration and, in general, require emergency corrective surgery.81 Surgery within the first 24 hours postprocedure is seldom indicated, but may be necessary in case of massive hemopericardium unresponsive to pericardiocentesis and, less frequently, by severe MR leading to acute refractory pulmonary edema.21 Elective indications for surgery after PMV are failure due to inadequate valve dilation, which usually occurs with tunnel-shaped deformed valves;37 correction of hemodynamically significant ASD, and elective correction of severe MR. The rate of systemic embolism varies between 0.5% and 3%, very seldom causing permanent disability, and more seldom leading to death.82 The origins of embolism were found to be dislodgement of atrial thrombi or thrombi formed on the catheter during the procedure, which were found to be gas embolism after rupture of the balloon and dislodgement of valve debris.14 Most of the symptomatic emboli affect the brain; others affect the lower limbs, kidneys and coronary arteries, the right one being the most affected.1,22 The more extensive use of the Inoue technique, which requires less LA manipulation, and the introduction of transesophageal echocardiography to detect atrial thrombi, have significantly reduced the incidence of such complication. Nonetheless, Rocha et al. studied 27 patients submitted to brain magnetic ressonance pre- and post-PMV, and found new hyperintensive focuses compatible with thromboembolism in 11 of them after the procedure. All of the focuses were smaller than 1 cm and none were followed by neurological symptoms, but those findings suggest that the real incidence of thromboembolism is higher than it seems to be.83 The detection of ASD after PMV is related to clinical variables (age, low CO, valve calcification, high echo score, previous surgical commissurotomy), to post-procedure MVA, and to technical factors (septal puncture site, type and size of balloon).84-86 Depending on the chosen technique, ASD has been diagnosed in up to 87% of PMV patients, but only 1-2% of them had a significant ASD (relation between systemic and pulmonary flow > 1.5:1.0).2,3,87 By and large, the defects are small, undetectable by oxymetry, and most of them close spontaneously with time, except if the shunt is significant or if dilation failed.88 With valve restenosis, defects may reappear, creating a variation of the Lutembacher Syndrome.22,89 Among the minor complications, Iung et al. reported 1.1% of peripheral vascular complications that required surgery after 1,024 PMVs,90 a rate similar to the 0.9% of patients in the NHLBI’s PMV registry. In this same registry, bleeding requiring blood transfusion occurred in 8.5% of the cases, and arrhythmia requiring treatment in 10%.74 Very seldom, there may be conduction abnormalities after valve dilation, including total atrial-ventricular block, most of them transient.91 In the past few years, the increased experience in performing PMV associated with the use of transesophageal echocardiography, and the increased use of the Inoue technique have contributed to a significant drop in procedure-related complication rates, favorably comparing to those of surgical commissurotomy.87 6. Mitral regurgitation. In some cases, the degree of a pre-existing MR may decrease, probably resulting from improved leaflet mobility after valve dilation.22 Discreet increases in MR magnitude may be due to stretching of the valve annulus, small excesses in commissural splitting, edema of the papillary muscle, or abnormal cusp coaptation, occurring in up to 40% of the patients after PMV; patients with some degree of previous MR are at a three-fold higher risk. The regurgitation jets are usually located at the area in which the commissures were opened, suggesting complete commissurotomy.14,22,92,93 These discreet increases of MR degree are well tolerated by the patients, and one study even suggested that patients who have commissural MR after PMV have a better outcome. In that study, Kang et al. have shown that, after PMV, commissural MR patients had lower MVA late loss when compared to patients who did not have such regurgitation, probably due to a lower commissural refusion rate.93 Sequential echocardiographies performed in patients who develop discreet or moderate MR after PMV have shown that the regurgitation degree remains stable or decreases with time.21 Certain mechanisms have been deemed to account for such decrease: elastic recoil of the valve annulus, fibrosis of commissure extremities after excessive commissural splitting, improved function of the papillary muscles after balloon trauma, and commissural refusion.27,94 When investigating severe MR after PMV, Iung et al., after analyzing the results of 1,514 procedures, reported MR ? grade 3 in 3.4% of the patients. The mechanisms responsible for that, either isolated or combined, were paracommissural lacerations in 36% of the cases, median laceration of a leaflet in 40%, excessive opening of one or both commissures injuring the valve annulus in 11%, partial or total rupture of the papillary muscle in 8%, and rupture of the chordae tendineae in 34%. In 42.5% of the cases, severe MR was associated with a lack of commissural opening.95 Although an emergency surgery to correct severe MR after PMV is seldom required, most of these patients will need surgery eventually.30,96 In a report of 3,650 patients submitted to PMV in India, Kaul et al. described severe MR in 3.3% of the procedures. All the patients whose MR was due to rupture of valve leaflets (usually the posterior one) required urgent valve replacement; however, when regurgitation was due to excessive commissural splitting, the severity of MR tended to decrease with time.97 Among 566 patients submitted to PMV, Padial et al. performed pathological analysis of the valves of 37 patients who developed severe MR. In those valves, 3 characteristics were constant: heterogeneous valve thickening with thick areas concomitant with normal ones, involvement of the subvalvular apparatus, and calcium in one or both commissures.92 These data corroborate the findings of Reifart et al. who, in in vitro dilations, observed that in valves with heavily diseased commissures, the leaflets ruptured at the least affected portions. The authors thus suggested that the occurrence of severe MR depends more on the location of the morphological changes of the valve than on their severity.48 It may also be the reason why no clinical, anatomical or technical predictor was consistent enough to predict severe MR after PMV:22,94,96,98 changes at specific portions of the valve can neither be characterized by a general morphologic description, nor be overcome by a specific technique.99 7. Late outcome and its predictors. The clinical utility of PMV can be assessed exclusively by information regarding patients’ late follow-up. However, such analysis is subject to several limitations. Most of the data available refers to procedures carried out in the late 1980s when dilation techniques and learning curves were still incipient. Furthermore, most follow-up studies emanate from industrialized countries. In those countries, late results are clearly worse that in developing countries, where patients are treated at a younger age, with less comorbidities and by and large less deformed valves.22,65,100,101 In addition, late follow-up period for PMV patients is still short when compared to surgical series, especially considering that peak of valve function deterioration after CMC is after 12 years.102,103 Palacios et al. recently published data concerning the 15-year follow-up (mean 4.2 ± 3.7 years) of 879 patients submitted to 939 PMVs, and showed that although the rate of major events (death, surgery, new PMV) may be low in the first 5 years, it progressively increases after this period, reaching a total of 47.2% combined events (especially mitral valve replacements) by the end of the study. Of the 446 event-free patients, 94% were at FC I or II. The authors identified the following predictors of late events: age, functional class IV, previous surgical commissurotomy, initial MR 2, echo score > 8, severe MR postdilation and higher pulmonary artery pressure post-dilation.104 Those results add to others from large series published earlier: Hernandez et al. have shown, in a 561-patient follow-up, a 69% major event-free survival rate after 7 years (88% for patients with good final outcome and low echo-score), and found that the only independent predictor of good late results is the final procedure result (MVA and MR degree post-dilation).30 Likewise, Iung et al. analyzed the late outcome of 528 PMV patients, and described, after 5 years, an event-free actuarial survival in 84 ± 6%, with 76 ± 6% of the patients having no further intervention and in FC I or II.101 The same group of authors, following 1,024 patients for a mean period of 49 months, determined that after 10 years, 56 ± 4% of the patients and 61 ± 5% of patients with a good immediate outcome would not need re-intervention, and would be in FC I or II after 10 years. Older age, unfavorable anatomy, worse FC, atrial fibrillation, higher residual gradient, and smaller final MVA were identified as multivariate predictors of unfavorable outcomes.90 Moreover, Orrange et al. determined an event-free actuarial survival of 65 ± 6% 7 years after PMV, having found independent predictors of a final MVA ? 1.5 cm2 and a final mean pulmonary capillary pressure ? 18 mmHg.38 After studying a series of 350 younger patients, Pan et al. reported that after 38 ± 15 months, 84% of those patients were event-free and FC I or II; and sinus rhythm and lack of valve calcification at fluoroscopy were found to be predictors of good outcome.105 Aiming at studying the late outcome of PMV in developing countries’ population, Ben Farhat et al. described, in Tunisia, the evolution of 654 patients with a mean age of 33 ± 13 years, and showed an event-free actuarial survival after 5, 7 and 10 years of 85%, 81% and 72%, respectively. For that population, the multivariate predictors of good outcome after 10 years were lower echo score, sinus rhythm and, after the procedure, lower LA pressure, lower gradient and MR 100 On the whole, and considering the above mentioned limitations, these results confirm the long-term benefits of PMV. Predictors of good late outcome is multifactorial, and vary according the population studied. By and large, however, elderly patients with more severe valve disease and poorer immediate outcome are those who have the worst outcome in the long term.65,90 8. Restenosis Post-PMV valve restenosis can only be diagnosed for sure based on three sequential echocardiographic and/or hemodynamic exams: the first, before intervention; the second immediately after intervention, demonstrating a successful valve dilation; and the third, evidencing late reduction of the valvar orifice. Based on symptom recurrence alone, the restenosis rate varies between 2% and 60%, and it may be caused, in addition to real restenosis, by a residual stenosis due to immediate unsatisfactory outcomes, development of MR, or diseases unrelated to the mitral valve such as coronary artery disease or aortic valve disease.106,107 Most of the patients who had successful intervention and get worse in the late follow-up have valve restenosis.108 On the other hand, about 40% of the patients who satisfy anatomic criteria for restenosis are asymptomatic.109 Unsatisfactory early results, commissure refusion, progressive leaflet thickening, and calcification and progressive subvalvular disease were found to be the mechanisms of post-PMV restenosis.107,110,111 The relation between restenosis and recurrence of RF remains unknown, although this mechanism has been considered, especially in patients who develop restenosis within a short period after valve dilation.112,113 Restenosis rates reported in literature vary considerably, depending on the follow-up time and definition adopted, if based on absolute MVA, percentage loss of MVA and/or loss of initial gain. Nevertheless, the most common definition used is the determination of a MVA smaller than 1.5 cm2, with a loss of 50% of the gain achieved with valve dilation.114 The available data allow one to conclude that restenosis is uncommon up to three years after successful PMV, but the rate increases with time, reaching a peak 50 months, on average, after the intervention. Depending on the population treated, this rate varies between 1.7% and 26%, and selected patients can undergo re-dilation.22,30,50,100,115,116 More often, restenosis predictors in different populations have been found to be age, higher echo score and smaller final MVA, but the short follow-up time prevents definitive conclusions on this matter.22,67,100,115,116