- Email: [email protected]
The Advantages of On-line Transesophageal Echocardiography Guide During Percutaneous Balloon Mitral Valvuloplasty
The Advantages of On-line Transesophageal Echocardiography Guide During Percutaneous Balloon Mitral Valvuloplasty Seong-Hoon Park, MD, Myung-A Kim, MD, and Min-Su Hyon, MD, Seoul, Korea
The purpose of this study was to evaluate the potential advantages of on-line transesophageal echocardiography during percutaneous balloon mitral valvuloplasty (PBMV). One hundred thirtyfour consecutive patients who underwent PBMV were included in this study. Group 1 included 64 patients who underwent PBMV under fluoroscopy guide only, and group 2 included 70 patients who underwent PBMV under on-line transesophageal echocardiography guide. Inoue balloons were used in all cases. The mitral valve area after valvuloplasty was comparable between the 2 groups. The procedure time was significantly shorter in group 2 (99 ± 48 min vs 64 ± 22 min, P < .0001 ), and the average
Percutaneous balloon mitral valvuloplasty (PBMV) has become the treatment of choice in patients with pure mitral stenosis.1-6 The role of transesophageal echocardiography (TEE) in patients undergoing PBMV was limited to the evaluation of left atrial thrombus before it began to be used during PBMV procedure as an additional guiding tool, especially for septal puncture, evaluation of early result, and detection of complications.At first TEE was performed with the patient fully sedated or anesthetized and thus made the procedure labor intensive.7,8 Goldstein et al9,10 reported a large experience of on-line TEE monitoring during PBMV under mild sedation of the patients and demonstrated the feasibility and utility of TEE during PBMV. And it has been suggested that on-line TEE may shortFrom the Division of Cardiology, Department of Internal Medicine, Mokdong Hospital, Medical Research Center, Ewha Womans University College of Medicine, Seoul National University College of Medicine, and Soonchunhyang University College of Medicine. Presented in part at the American Society of Echocardiography Seventh Annual Sessions, June 10-12, 1996, Chicago, Ill. Reprint requests: Seong-Hoon Park, MD, Division of Cardiology, Department of Internal Medicine, Mokdong Hospital, Medical Research Center, Ewha Womans University College of Medicine, 911-1, Mokdong, Yangcheongu, Seoul, 158-710, Korea. Copyright © 2000 by the American Society of Echocardiography. 0894-7317/2000 $12.00 + 0 27/1/102063
26
fluoroscopy time was shorter in group 2 without statistical significance (30 ± 17 min vs 19 ± 15 min, P = .25 ). Five (7.8%) patients in group 1 and 2 (2.8%) patients in group 2 underwent surgery because of procedure-related complications. The limitation of this study was the learning curve of the operator, because the 2 groups were treated serially. In conclusion, transesophageal echocardiography can be used effectively during balloon mitral valvuloplasty, and it may help to reduce the rate of complications resulting from trans-septal catheterization and balloon valvuloplasty and to reduce procedure time and fluoroscopy time. (J Am Soc Echocardiogr 2000;13:26-34.)
en procedure time and fluoroscopy time,11 but the advantage of on-line TEE has not been evaluated by comparative study.Therefore we performed a study to evaluate the potential advantages of on-line TEE during PBMV procedure by comparing 2 patient groups of mitral stenosis treated by PBMV with and without on-line TEE.
METHODS Patient Population One hundred thirty-four consecutive patients who underwent PBMV were included in this study. Patients were divided into 2 groups: group 1 included 64 patients (male: 14, average age 40 ± 10 years) who underwent PBMV under fluoroscopy guide alone from October 1991 to April 1995, and group 2 included 70 patients (male: 14, average age 44 ± 13 years) who underwent PBMV under on-line TEE guide in addition to fluoroscopy from May 1995 to May 1996. Four patients in group 2 were pregnant at the time of PBMV. For pregnant women we used lead protector under the patient’s back and wrapped the abdomen with lead protector. All patients underwent transthoracic echocardiography and TEE to obtain baseline data and to rule out thrombi in the left atrium or left atrial appendage before PBMV was performed.We excluded patients with a
Journal of the American Society of Echocardiography Volume 13 Number 1
Park, Kim, Hyon 27
Figure 1 Transesophageal echocardiography is guiding trans-septal puncture at 90°. The tip of the Mullin sheath can be identified by echocardiographic shadowing and the deformity of the limbus and fossa ovalis membrane produced by the tip of the Mullin sheath (small arrow) (A). At the center of the fossa ovalis membrane, the tip of puncture needle is well visualized by the tenting of the fossa ovalis membrane toward left atrial side (arrow) (B).
mitral valve score more than 11 and patients with thrombi in the left atrium. But 5 patients with left atrial thrombi confined to the left atrial appendage even after several months of anticoagulation underwent balloon valvuloplasty under TEE monitoring and were included in this study. Included patients had symptoms with more than New York Heart Association functional class II symptoms. Demographic characteristics of the patients in each group are listed in Table 1.
Table 1 Summary of demographic data
Patient number Age (y) Sex (M:F) Atrial fibrillation (%)
Group 1
Group 2
64 40 ± 10 14:50 23 (36)
70 44 ± 13 14:56 32 (46)
On-line Transesophageal Echocardiography Percutaneous Balloon Mitral Valvuloplasty All patients underwent diagnostic left and right heart catheterization and coronary angiography before undergoing PBMV. Inoue balloons were used for PBMV in all patients. Trans-septal catheterization was performed with an 8F Mullins trans-septal dilator and a modified Brockenbrough needle. After entry into the left atrium, 5000 U heparin were administered. Simultaneous pressure tracing of left atrium and ventricle were recorded. A stainless steel guide wire was advanced in the left atrium and was used to place the Inoue balloon catheter in the left atrium. After entering the left atrium, the balloon was passed through the mitral orifice into the left ventricle with continuous fluoroscopic and transesophageal echocardiographic guidance. Balloon inflation were then performed until hemodynamic and echocardiographic measurements were satisfactory or complications appeared. Mitral regurgitation was evaluated with cine left ventriculography, and its severity was graded by the Seller’s classification. Right side oximetry was performed to evaluate interatrial shunting.
Patients in group 2 underwent PBMV with on-line TEE guide. TEE was performed immediately after cardiac catheterization and coronary angiography with the patient under mild sedation with 5 mg Valium administered intravenously. In some patients additional sedation with intravenous Demerol was used because of active gag reflex. Transesophageal transducer was introduced with the patient in a recumbent position on the catheterization table. Secretion was drained naturally, but sometimes intermittent suction was done in patients with profuse secretion. In general, patients tolerated the procedure well, and no TEE-related complications occurred. Multiplane 5-MHz transesophageal transducers from Apogee CX200 (Interspec) or Sonos 2500 (Hewlett Packard) were used in all patients. Complete transesophageal echocardiographic examination was performed before trans-septal puncture, with a concentration on the mitral valve structure, subvalvular structure, spontaneous echocardiographic contrast, and thrombi of left atrium and left atrial appendage. Mitral valve area by Doppler pressure half-time, mean
Journal of the American Society of Echocardiography January 2000
28 Park, Kim, Hyon
Figure 2 Transesophageal echocardiography is monitoring balloon position during balloon inflation. The balloon is fully inflated at the subvalvular position (A), and the balloon is reinflated at the correct position with the mitral valve leaflets located at the waist of the balloon (B).
Table 2 Summary of echocardiography data before and after valvuloplasty Group 1
Patient number 64 Baseline LA (mm) 48 ± 6 Baseline MVA (cm2) 0.91 ± 0.24 Post LA (mm) 43 ± 7 Post MVA (cm2) 1.73 ± 0.4 Procedure time (min) 99 ± 48 Fluoroscopy time (min) 30 ± 17
Group 2
52 0.92 49 1.76 64 19
70 ±7 ± 0.17 ±7 ± 0.4 ± 22 ± 15
P value
.547 .016 .830 .611 .0001 .25
LA, Left atrial size; MVA, mitral valve area; post LA, left atrial size after PBMV; post MVA, mitral valve area after PBMV.
transmitral pressure gradient,12 structure and function of aortic and tricuspid valve, and pulmonary vein flow were also evaluated. Multiplane TEE guided trans-septal puncture at 90° and the position of the puncture needle was always double-checked at 0° before puncture was performed. After the Mullin sheath was positioned at the center of fossa ovalis membrane, tenting of the fossa ovalis membrane was identified by pushing of the Brockenbrough needle (Figure 1). After puncture was performed, the position of the fossa ovalis membrane resumed totally or partially to its original position. In the case of huge left atrial dilatation with high left atrial pressure, passing the balloon through the mitral valve into the left ventricle was sometimes impossible at the first septal puncture site. In such a case the second septal puncture site was selected with the reference of the first puncture site visualized by color flow imaging.After the balloon was inserted into the
left ventricle, the distal half of the balloon was inflated, and the proximal half of the balloon was anchored at the stenotic mitral valve orifice by pulling back the balloon shaft. Then the balloon was fully inflated to dilate the stenotic mitral valve. But sometimes the balloon was anchored at the subvalvular space of anterior mitral valve, or the balloon was fully inflated at the subvalvular position (Figure 2). Special attention should be paid to avoid those conditions, because in those cases the chord rupture is likely to happen or the transmitral pressure gradient drops suboptimally. The immediate result of balloon dilatation was assessed with the pressure half-time and transmitral pressure gradient by continuous wave Doppler, which was compared with the left atrial pressure and transmitral pressure gradient measured by the catheter double-tracing method. At the same time maximal diastolic separation of mitral valve tips, the severity of mitral regurgitation, and possible complications such as chord rupture or papillary muscle rupture and pericardial effusion were evaluated. Balloon valvuloplasty was repeated until the transmitral pressure gradient and mitral valve area by pressure half-time were satisfactory, and sometimes the balloon size was upgraded as necessary to further dilate the mitral valve orifice. Iatrogenic atrial septal defect was also evaluated by color flow imaging before the TEE probe was extubated. Statistical Analysis For statistical analysis we used SPSS statistics package. Paired t test or unpaired t test were used for comparison of 2 groups, and P < .05 was regarded as significant.
Journal of the American Society of Echocardiography Volume 13 Number 1
Park, Kim, Hyon 29
Figure 3 The transesophageal echocardiography is monitoring valvuloplasty in a patient with a nonmobile round thrombus (big arrow) in the left atrial appendage (A). The balloon position is well visualized near the posterior mitral valve leaflet (arrow), and the balloon shaft is crossing the orifice of the left atrial appendage (arrow) (B).
RESULTS As summarized in Table 1, no significant difference was seen in the patients’ age and sex distribution between the 2 groups. The incidence of atrial fibrillation was slightly higher in group 2, but the baseline left atrial size did not show significant difference as shown in Table 2. Baseline mitral valve area and postvalvuloplasty mitral valve area did not show significant difference between the 2 groups (Table 2).The average procedure time of group 2 was significantly shorter than that of group 1 (64 ± 22 min vs 99 ± 48 min, P < .0001). The average fluoroscopy time of group 2 was shorter than that of group 1 (19 ± 15 min vs 30 ± 17 min, P = .25) without statistical significance, but the average fluoroscopy time of 4 pregnant women in group 2 was 1.8 minutes (range, 1 to 4 minutes), suggesting that it could be shortened as necessary. All 4 pregnant women of group 2 underwent the valvuloplasty procedure uneventfully. Five patients of group 2 had thrombi in the left atrium that were confined to the left atrial appendage even after the long period (more than 3 months) of anticoagulation. The transesophageal echocardiography was useful in monitoring the position of the balloon to avoid touching the thrombi during the procedure (Figure 3). All 5 patients with left atrial appendage thrombi underwent the procedure uneventfully without major or minor embolic events. The major surgical complications that developed after valvuloplasty are listed in Table 3. All 5
Table 3 Surgical complications developed after valvuloplasty procedure Group 1 Group 2 (n = 64) (n = 70)
Papillary m. rupture Papillary m. rupture + PMVL tear Papillary m. rupture + AMVL tear AMVL tear PMVL tear Pericardial tamponade + AMVL perforation Pericardial tamponade
1 1 1 1 1 1 1
1*
AMVL, Anterior mitral valve leaflet; PMVL, posterior mitral valve leaflet. *Treated by pericardiocentesis and underwent successful balloon valvuloplasty 1 week later.
(7.8%) patients of group 1 who had major complications during valvuloplasty underwent surgery. Two (2.8%) patients of group 2 underwent surgery because of papillary muscle rupture associated with severe mitral regurgitation, which developed after balloon inflation (Figure 4). But in 1 patient of group 2 who had pericardial tamponade during trans-septal puncture, the pericardial tamponade was detected early by TEE, and the transducer was switched to a transthoracic one to guide the pericardial puncture site at the cardiac apex with the patient in the semiFowler’s position (Figure 5).The pericardial tamponade was drained by pigtail catheter, the bleeding stopped 12 hours later, and the patient underwent successful balloon valvuloplasty 1 week later.
Journal of the American Society of Echocardiography January 2000
30 Park, Kim, Hyon
A
B
C
D
Figure 4 Immediately after valvuloplasty, transesophageal echocardiography visualized ruptured papillary muscle head prolapsing into the left atrium during systole (A) and moving back into the left ventricle during diastole (B). Severe mitral regurgitation was associated with papillary muscle rupture as visualized by color Doppler interrogation in 2-chamber and 4-chamber view (C, D).
DISCUSSION After Kronzon et al13 reported the feasibility and safety of on-line TEE during PBMV with the patient under topical anesthesia to the oropharynx, Goldstein et al10 used TEE during PBMV in 93 patients under mild sedation. The patients tolerated the procedure well without any TEE-related complications. They reported that the trans-septal puncture was
facilitated, the relative position of balloon and mitral valve was easily identified, and the immediate result after valvuloplasty was easily evaluated by Doppler pressure half-time and mean transmitral pressure gradient by concomitant on-line TEE monitoring. And it was also suggested that fluoroscopy time may be reduced with all the benefits of on-line TEE.Vilacosta et al11 and Ramondo et al14 reported that procedure time and fluoroscopy time could be reduced by on-
Journal of the American Society of Echocardiography Volume 13 Number 1
Park, Kim, Hyon 31
Figure 5 After trans-septal puncture, transesophageal echocardiography identified pericardial tamponade in 4-chamber (A) and 2-chamber view (B). Because of the deteriorating hemodynamic status of the patient, the transesophageal transducer was switched to transthoracic transducer to guide the pericardiocentesis. Subcostal view shows pericardial tamponade before pericardiocentesis (C) and after drainage of the pericardial blood by pigtail catheter (D).
line TEE. Our preliminary data also demonstrated shorter fluoroscopy time and procedure time.15 Trans-septal Puncture During trans-septal puncture TEE can visualize the relative position of the Brockenbrough needle and atrial septum (Figure 1, A). In 1 patient of group 2 who had pericardial tamponade after trans-septal puncture, the trans-septal needle may have tracked
through the slitlike space between the septum primum and septum secundum and punctured the atrial wall.The stainless steel wire made a big loop around the atrial wall instead of making small multiple loops within the left atrium.When this sign is observed, the operator should not introduce the dilator to enlarge the puncture site. To avoid this complication, the echocardiographer should double-check the exact position of the tip of the trans-septal needle in both
32 Park, Kim, Hyon
Journal of the American Society of Echocardiography January 2000
Figure 6 In a patient with heavily calcified mitral valve, especially the calcific nodule of the posterior mitral valve leaflet blocked entry of the balloon into the left ventricle (A, B). After passage of the guide wire into the left ventricle (C), the balloon glided easily into the left ventricle along the guide wire and was successfully inflated (D).
the vertical and the horizontal plane to see the tenting of the fossa ovalis membrane (Figure 1, B). Monitoring Balloon Position The relative position of the balloon and cardiac structure is well visualized by TEE, so when the balloon is in the left atrium, the relative direction of the long axis of the balloon and the mitral valve orifice can help the operator redirect the balloon. Once the balloon gets into the left ventricle, the distal half of the balloon is inflated, and the balloon shaft is pulled back to anchor the proximal part of the balloon into the mitral valve orifice before the balloon is fully inflated. During this procedure sometimes the balloon is engaged into the subvalvular space, and sometimes the balloon is fully inflated within the left ventricle under the valve (Figure 2, A). The chord rupture is likely to happen in those cases. As shown in Figure 6, when the balloon cannot be engaged into the mitral orifice because of an anatomic obstacle such as heavily calcified posterior mitral valve leaflet, one can choose to use J-wire or wedge balloon to pass the mitral valve orifice. In patients with left atrial appendage thrombi,TEE visualizes the rela-
tive position of left atrial appendage and the balloon so that the operator can avoid pushing the balloon into the left atrial appendage (Figure 3). Assessment of the Result of Valvuloplasty The immediate result of valvuloplasty is evaluated by Doppler pressure half-time and mean transmitral pressure gradient with continuous wave Doppler. And maximal separation of the mitral valve tip can be measured at the horizontal and vertical planes to assess the adequacy of the valvuloplasty. This information is combined with the possible development of new or worsening mitral regurgitation and the hemodynamic data acquired by catheterization to decide whether to upgrade the balloon size and repeat the procedure or to stop the procedure there. Early Detection of Complications Severe mitral regurgitation caused by papillary muscle rupture or chordae tendinae rupture, pericardial tamponade caused by cardiac perforation, or iatrogenic atrial septal defect may develop during balloon valvuloplasty, and those complications are easily detected by on-line TEE monitoring. Hemodynami-
Journal of the American Society of Echocardiography Volume 13 Number 1
cally insignificant mild mitral regurgitation may develop after valvuloplasty resulting from minor chord rupture or commissural tear. Hemodynamically significant severe mitral regurgitation has been reported to develop in 1% to 6% of the patients who undergo balloon mitral valvuloplasty.5,16,17 We experienced 2 cases of severe mitral regurgitation in group 2. One patient had severe mitral regurgitation caused by papillary muscle rupture immediately after valvuloplasty and underwent mitral valve replacement because of intractable symptoms after 1 month of medical treatment. The other patient had severe mitral regurgitation caused by papillary muscle rupture and cardiogenic shock after valvuloplasty and underwent emergency surgery (Figure 4). Operative findings showed papillary muscle rupture and posterior mitral valve leaflet tear extending to the posterior margin of the mitral annulus. One patient had pericardial tamponade after trans-septal puncture, and it was easily visualized by TEE, so the transducer was switched with a transthoracic one to guide the pericardial puncture site at the cardiac apex with the patient in the semi-Fowler’s position (Figure 5). Left Atrial Appendage Thrombus The presence of left atrial appendage thrombus has been regarded as a contraindication to balloon valvuloplasty. Hung et al16 reported that patients with left atrial thrombi were safely treated with balloon valvuloplasty after thrombi was successfully resolved with several months of anticoagulation. However, several cases of uneventful balloon valvuloplasty have been reported in patients with left atrial appendage thrombi.18,19 Five patients with left atrial appendage thrombi in group 2 underwent balloon valvuloplasty without embolic event (Figure 3). In 3 patients the left atrial appendage thrombi lysed after several months of anticoagulation at the time of follow-up TEE after valvuloplasty. So the presence of left atrial appendage thrombi even after 3 months of anticoagulation should not be regarded as a contraindication to balloon valvuloplasty, which is performed under TEE monitoring. Procedure Time and Fluoroscopy Time The procedure time in group 1 was 99 ± 48 minutes, which was similar to 104 ± 13 minutes that was reported by Bassand et al20 in valvuloplasty with an Inoue balloon without on-line TEE monitoring. The procedure time in group 2 was 64 ± 22 minutes, which is similar to 65 ± 18 minutes reported by Ramondo et al.14 In an effort to reduce radiation for pregnant women, we shielded the patients with lead
Park, Kim, Hyon 33
protectors and tried to minimize the fluoroscopy time so we could reduce fluoroscopy time to an average of 1.8 minutes (range 1 to 4 minutes) in 4 pregnant women. The average fluoroscopy time of group 1 (30 ± 17 minutes) was longer than that of group 2 (19 ± 15 minutes) without statistical significance. Bassand et al20 reported an average fluoroscopy time of 16 ± 8 minutes in patients who underwent valvuloplasty with an Inoue balloon without on-line TEE.TEE may contribute in reducing procedure time and fluoroscopy time in difficult cases with heavy calcification of mitral valve by recommending early use of a guide wire (Figure 6). Limitations This study included 2 patient groups: group 1, who underwent PBMV under fluoroscopy alone from 1991 to 1995, and group 2, who underwent PBMV under on-line TEE in addition to fluoroscopy from May 1995 to May 1996. Because this study was planned just before the beginning of the on-line TEEguided PBMV, the data of group 1 were acquired by retrospective review of the patient’s record. The learning period may be responsible for the higher rate of major complications and longer procedure time in group 1. But the operator was confident during the whole procedure, especially during trans-septal puncture and balloon positioning with the presence of on-line TEE guide. And this may have contributed to the lower major complication rate and shortening of the procedure time in group 2.
REFERENCES 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. Zaibag MA, Kasab SA, Ribeiro PA, Fagih MRA. Percutaneous double balloon mitral valvulotomy for rheumatic mitral valve stenosis. Lancet 1986;1:757-61. 3. Palacios I, Block PC, Brandi S, et al. Percutaneous balloon mitral valvotomy for patients with severe mitral stenosis. Circulation 1987;75:778-84. 4. McKay CR, Kawanishi DT, Rahimtoola SH. Catheter balloon valvuloplasty of the mitral valve in adults using a double-balloon technique. Early hemodynamic results. JAMA 1987; 257:1753-16. 5. Ruiz CE, Allen JW, Lau FYK. Percutaneous double balloon valvotomy for severe rheumatic mitral stenosis. Am J Cardiol 1990;65:473-7. 6. Block PC, Tuzcu EM, Palacios IF. Percutaneous mitral balloon valvotomy. Cardiol Clin 1991;9:271-87. 7. Den Heijer P, Hammer J, Pieper E, van Dijk R, Lie K. Percutaneous mitral balloon valvotomy guided by biplane transesophageal echocardiography [Abstract]. Eur Heart J 1991;12(Suppl):47.
Journal of the American Society of Echocardiography January 2000
34 Park, Kim, Hyon
8. Kültürsay H, Türkoglu C, Akin M, Payzin S, Soydas C, Akilli A. Mitral balloon valvuloplasty with transesophageal echocardiography without using fluoroscopy. Cathet Cardiovasc Diagn 1992;27:317-21. 9. Goldstein SA, Campbell AN. Mitral stenosis. Evaluation and guidance of valvuloplasty by transesophageal echocardiography. Cardiol Clinics 1993;11:409-25. 10. Goldstein SA, Campbell AN, Minz GS, Pichard A, Leon M, Lindsay Jr J. Feasibility of on-line transesophageal echocardiography during balloon mitral valvulotomy: experience with 93 patients. J Heart Valve Dis 1994;3:136-48. 11. Vilacosta I, Iturralde E, San Román JA, Gómez-Recio M, Romero C, Jiménez J, Martínez-Elbal L. Transesophageal echocardiographic monitoring of percutaneous mitral balloon valvulotomy. Am J Cardiol 1992;70:1040-4. 12. Abascal VM, Wilkins GT, Choog CY, Thomas JD, Palacios IF, Block PC, Weyman AE. Echocardiographic evaluation of mitral valve structure and function in patients followed for at least 6 months after percutaneous balloon mitral valvuloplasty. J Am Coll Cardiol 1988;12:606-15. 13. Kronzon I, Tunick PA, Schwinger ME, Slater J, Glassman E. Transesophageal echocardiography during percutaneous mitral valvuloplasty. J Am Soc Echocardiogr 1989;2:380-5. 14. Ramondo A, Chirillo F, Isabella G, Bonato R, Rampazzo C, Razzolini R, Andriolo L, Mazzucco A, Chioin R. Value and limitations of transesophageal echocardiographic monitoring
15.
16.
17.
18.
19.
20.
during percutaneous balloon mitral valvotomy. Int J Cardiol 1991;31:223-34. Park SH, Kim MA, Hyon MS, et al. Advantages of transesophageal echo guide during balloon mitral valvuloplasty [abstract]. J Am Soc Echocardiogr 1996;9:376. Hung JS, Chern MS, Wu JJ, Fu M, Yeh KH, Wu YC, Cherng WJ, Chua S, Lee CB. Short- and long-term results of catheter balloon percutaneous transvenous mitral commissurotomy. Am J Cardiol 1991;67:854-62. Nobuyoshi M, Hamasaki N, Kimura T, et al. Indications, complications and short-term clinical outcome of percutaneous transvenous mitral commissurotomy. Circulation 1989;80:782-92. Thomas MR, Monaghan MJ, Smyth DW, Metcalfe JM, Jewitt DE. Comparative value of transthoracic and transesophageal echocardiography before balloon dilatation of the mitral valve. Br Heart J 1992;68:493-7. Rittoo D, Sutherland GR, Currie P, Starkey IR, Shaw TRD. The comparative value of transesophageal and transthoracic echocardiography before and after percutaneous mitral balloon valvotomy: a prospective study. Am Heart J 1993;125: 1094-105. Bassand JP, Schiele F, Bernard Y, et al. The double balloon and Inoue techniques in percutaneous mitral valvuloplasty: comparative results in a series of 232 cases. J Am Coll Cardiol 1991;18:982-9.
The purpose of this study was to evaluate the potential advantages of on-line transesophageal echocardiography during percutaneous balloon mitral valvuloplasty (PBMV). One hundred thirtyfour consecutive patients who underwent PBMV were included in this study. Group 1 included 64 patients who underwent PBMV under fluoroscopy guide only, and group 2 included 70 patients who underwent PBMV under on-line transesophageal echocardiography guide. Inoue balloons were used in all cases. The mitral valve area after valvuloplasty was comparable between the 2 groups. The procedure time was significantly shorter in group 2 (99 ± 48 min vs 64 ± 22 min, P < .0001 ), and the average
Percutaneous balloon mitral valvuloplasty (PBMV) has become the treatment of choice in patients with pure mitral stenosis.1-6 The role of transesophageal echocardiography (TEE) in patients undergoing PBMV was limited to the evaluation of left atrial thrombus before it began to be used during PBMV procedure as an additional guiding tool, especially for septal puncture, evaluation of early result, and detection of complications.At first TEE was performed with the patient fully sedated or anesthetized and thus made the procedure labor intensive.7,8 Goldstein et al9,10 reported a large experience of on-line TEE monitoring during PBMV under mild sedation of the patients and demonstrated the feasibility and utility of TEE during PBMV. And it has been suggested that on-line TEE may shortFrom the Division of Cardiology, Department of Internal Medicine, Mokdong Hospital, Medical Research Center, Ewha Womans University College of Medicine, Seoul National University College of Medicine, and Soonchunhyang University College of Medicine. Presented in part at the American Society of Echocardiography Seventh Annual Sessions, June 10-12, 1996, Chicago, Ill. Reprint requests: Seong-Hoon Park, MD, Division of Cardiology, Department of Internal Medicine, Mokdong Hospital, Medical Research Center, Ewha Womans University College of Medicine, 911-1, Mokdong, Yangcheongu, Seoul, 158-710, Korea. Copyright © 2000 by the American Society of Echocardiography. 0894-7317/2000 $12.00 + 0 27/1/102063
26
fluoroscopy time was shorter in group 2 without statistical significance (30 ± 17 min vs 19 ± 15 min, P = .25 ). Five (7.8%) patients in group 1 and 2 (2.8%) patients in group 2 underwent surgery because of procedure-related complications. The limitation of this study was the learning curve of the operator, because the 2 groups were treated serially. In conclusion, transesophageal echocardiography can be used effectively during balloon mitral valvuloplasty, and it may help to reduce the rate of complications resulting from trans-septal catheterization and balloon valvuloplasty and to reduce procedure time and fluoroscopy time. (J Am Soc Echocardiogr 2000;13:26-34.)
en procedure time and fluoroscopy time,11 but the advantage of on-line TEE has not been evaluated by comparative study.Therefore we performed a study to evaluate the potential advantages of on-line TEE during PBMV procedure by comparing 2 patient groups of mitral stenosis treated by PBMV with and without on-line TEE.
METHODS Patient Population One hundred thirty-four consecutive patients who underwent PBMV were included in this study. Patients were divided into 2 groups: group 1 included 64 patients (male: 14, average age 40 ± 10 years) who underwent PBMV under fluoroscopy guide alone from October 1991 to April 1995, and group 2 included 70 patients (male: 14, average age 44 ± 13 years) who underwent PBMV under on-line TEE guide in addition to fluoroscopy from May 1995 to May 1996. Four patients in group 2 were pregnant at the time of PBMV. For pregnant women we used lead protector under the patient’s back and wrapped the abdomen with lead protector. All patients underwent transthoracic echocardiography and TEE to obtain baseline data and to rule out thrombi in the left atrium or left atrial appendage before PBMV was performed.We excluded patients with a
Journal of the American Society of Echocardiography Volume 13 Number 1
Park, Kim, Hyon 27
Figure 1 Transesophageal echocardiography is guiding trans-septal puncture at 90°. The tip of the Mullin sheath can be identified by echocardiographic shadowing and the deformity of the limbus and fossa ovalis membrane produced by the tip of the Mullin sheath (small arrow) (A). At the center of the fossa ovalis membrane, the tip of puncture needle is well visualized by the tenting of the fossa ovalis membrane toward left atrial side (arrow) (B).
mitral valve score more than 11 and patients with thrombi in the left atrium. But 5 patients with left atrial thrombi confined to the left atrial appendage even after several months of anticoagulation underwent balloon valvuloplasty under TEE monitoring and were included in this study. Included patients had symptoms with more than New York Heart Association functional class II symptoms. Demographic characteristics of the patients in each group are listed in Table 1.
Table 1 Summary of demographic data
Patient number Age (y) Sex (M:F) Atrial fibrillation (%)
Group 1
Group 2
64 40 ± 10 14:50 23 (36)
70 44 ± 13 14:56 32 (46)
On-line Transesophageal Echocardiography Percutaneous Balloon Mitral Valvuloplasty All patients underwent diagnostic left and right heart catheterization and coronary angiography before undergoing PBMV. Inoue balloons were used for PBMV in all patients. Trans-septal catheterization was performed with an 8F Mullins trans-septal dilator and a modified Brockenbrough needle. After entry into the left atrium, 5000 U heparin were administered. Simultaneous pressure tracing of left atrium and ventricle were recorded. A stainless steel guide wire was advanced in the left atrium and was used to place the Inoue balloon catheter in the left atrium. After entering the left atrium, the balloon was passed through the mitral orifice into the left ventricle with continuous fluoroscopic and transesophageal echocardiographic guidance. Balloon inflation were then performed until hemodynamic and echocardiographic measurements were satisfactory or complications appeared. Mitral regurgitation was evaluated with cine left ventriculography, and its severity was graded by the Seller’s classification. Right side oximetry was performed to evaluate interatrial shunting.
Patients in group 2 underwent PBMV with on-line TEE guide. TEE was performed immediately after cardiac catheterization and coronary angiography with the patient under mild sedation with 5 mg Valium administered intravenously. In some patients additional sedation with intravenous Demerol was used because of active gag reflex. Transesophageal transducer was introduced with the patient in a recumbent position on the catheterization table. Secretion was drained naturally, but sometimes intermittent suction was done in patients with profuse secretion. In general, patients tolerated the procedure well, and no TEE-related complications occurred. Multiplane 5-MHz transesophageal transducers from Apogee CX200 (Interspec) or Sonos 2500 (Hewlett Packard) were used in all patients. Complete transesophageal echocardiographic examination was performed before trans-septal puncture, with a concentration on the mitral valve structure, subvalvular structure, spontaneous echocardiographic contrast, and thrombi of left atrium and left atrial appendage. Mitral valve area by Doppler pressure half-time, mean
Journal of the American Society of Echocardiography January 2000
28 Park, Kim, Hyon
Figure 2 Transesophageal echocardiography is monitoring balloon position during balloon inflation. The balloon is fully inflated at the subvalvular position (A), and the balloon is reinflated at the correct position with the mitral valve leaflets located at the waist of the balloon (B).
Table 2 Summary of echocardiography data before and after valvuloplasty Group 1
Patient number 64 Baseline LA (mm) 48 ± 6 Baseline MVA (cm2) 0.91 ± 0.24 Post LA (mm) 43 ± 7 Post MVA (cm2) 1.73 ± 0.4 Procedure time (min) 99 ± 48 Fluoroscopy time (min) 30 ± 17
Group 2
52 0.92 49 1.76 64 19
70 ±7 ± 0.17 ±7 ± 0.4 ± 22 ± 15
P value
.547 .016 .830 .611 .0001 .25
LA, Left atrial size; MVA, mitral valve area; post LA, left atrial size after PBMV; post MVA, mitral valve area after PBMV.
transmitral pressure gradient,12 structure and function of aortic and tricuspid valve, and pulmonary vein flow were also evaluated. Multiplane TEE guided trans-septal puncture at 90° and the position of the puncture needle was always double-checked at 0° before puncture was performed. After the Mullin sheath was positioned at the center of fossa ovalis membrane, tenting of the fossa ovalis membrane was identified by pushing of the Brockenbrough needle (Figure 1). After puncture was performed, the position of the fossa ovalis membrane resumed totally or partially to its original position. In the case of huge left atrial dilatation with high left atrial pressure, passing the balloon through the mitral valve into the left ventricle was sometimes impossible at the first septal puncture site. In such a case the second septal puncture site was selected with the reference of the first puncture site visualized by color flow imaging.After the balloon was inserted into the
left ventricle, the distal half of the balloon was inflated, and the proximal half of the balloon was anchored at the stenotic mitral valve orifice by pulling back the balloon shaft. Then the balloon was fully inflated to dilate the stenotic mitral valve. But sometimes the balloon was anchored at the subvalvular space of anterior mitral valve, or the balloon was fully inflated at the subvalvular position (Figure 2). Special attention should be paid to avoid those conditions, because in those cases the chord rupture is likely to happen or the transmitral pressure gradient drops suboptimally. The immediate result of balloon dilatation was assessed with the pressure half-time and transmitral pressure gradient by continuous wave Doppler, which was compared with the left atrial pressure and transmitral pressure gradient measured by the catheter double-tracing method. At the same time maximal diastolic separation of mitral valve tips, the severity of mitral regurgitation, and possible complications such as chord rupture or papillary muscle rupture and pericardial effusion were evaluated. Balloon valvuloplasty was repeated until the transmitral pressure gradient and mitral valve area by pressure half-time were satisfactory, and sometimes the balloon size was upgraded as necessary to further dilate the mitral valve orifice. Iatrogenic atrial septal defect was also evaluated by color flow imaging before the TEE probe was extubated. Statistical Analysis For statistical analysis we used SPSS statistics package. Paired t test or unpaired t test were used for comparison of 2 groups, and P < .05 was regarded as significant.
Journal of the American Society of Echocardiography Volume 13 Number 1
Park, Kim, Hyon 29
Figure 3 The transesophageal echocardiography is monitoring valvuloplasty in a patient with a nonmobile round thrombus (big arrow) in the left atrial appendage (A). The balloon position is well visualized near the posterior mitral valve leaflet (arrow), and the balloon shaft is crossing the orifice of the left atrial appendage (arrow) (B).
RESULTS As summarized in Table 1, no significant difference was seen in the patients’ age and sex distribution between the 2 groups. The incidence of atrial fibrillation was slightly higher in group 2, but the baseline left atrial size did not show significant difference as shown in Table 2. Baseline mitral valve area and postvalvuloplasty mitral valve area did not show significant difference between the 2 groups (Table 2).The average procedure time of group 2 was significantly shorter than that of group 1 (64 ± 22 min vs 99 ± 48 min, P < .0001). The average fluoroscopy time of group 2 was shorter than that of group 1 (19 ± 15 min vs 30 ± 17 min, P = .25) without statistical significance, but the average fluoroscopy time of 4 pregnant women in group 2 was 1.8 minutes (range, 1 to 4 minutes), suggesting that it could be shortened as necessary. All 4 pregnant women of group 2 underwent the valvuloplasty procedure uneventfully. Five patients of group 2 had thrombi in the left atrium that were confined to the left atrial appendage even after the long period (more than 3 months) of anticoagulation. The transesophageal echocardiography was useful in monitoring the position of the balloon to avoid touching the thrombi during the procedure (Figure 3). All 5 patients with left atrial appendage thrombi underwent the procedure uneventfully without major or minor embolic events. The major surgical complications that developed after valvuloplasty are listed in Table 3. All 5
Table 3 Surgical complications developed after valvuloplasty procedure Group 1 Group 2 (n = 64) (n = 70)
Papillary m. rupture Papillary m. rupture + PMVL tear Papillary m. rupture + AMVL tear AMVL tear PMVL tear Pericardial tamponade + AMVL perforation Pericardial tamponade
1 1 1 1 1 1 1
1*
AMVL, Anterior mitral valve leaflet; PMVL, posterior mitral valve leaflet. *Treated by pericardiocentesis and underwent successful balloon valvuloplasty 1 week later.
(7.8%) patients of group 1 who had major complications during valvuloplasty underwent surgery. Two (2.8%) patients of group 2 underwent surgery because of papillary muscle rupture associated with severe mitral regurgitation, which developed after balloon inflation (Figure 4). But in 1 patient of group 2 who had pericardial tamponade during trans-septal puncture, the pericardial tamponade was detected early by TEE, and the transducer was switched to a transthoracic one to guide the pericardial puncture site at the cardiac apex with the patient in the semiFowler’s position (Figure 5).The pericardial tamponade was drained by pigtail catheter, the bleeding stopped 12 hours later, and the patient underwent successful balloon valvuloplasty 1 week later.
Journal of the American Society of Echocardiography January 2000
30 Park, Kim, Hyon
A
B
C
D
Figure 4 Immediately after valvuloplasty, transesophageal echocardiography visualized ruptured papillary muscle head prolapsing into the left atrium during systole (A) and moving back into the left ventricle during diastole (B). Severe mitral regurgitation was associated with papillary muscle rupture as visualized by color Doppler interrogation in 2-chamber and 4-chamber view (C, D).
DISCUSSION After Kronzon et al13 reported the feasibility and safety of on-line TEE during PBMV with the patient under topical anesthesia to the oropharynx, Goldstein et al10 used TEE during PBMV in 93 patients under mild sedation. The patients tolerated the procedure well without any TEE-related complications. They reported that the trans-septal puncture was
facilitated, the relative position of balloon and mitral valve was easily identified, and the immediate result after valvuloplasty was easily evaluated by Doppler pressure half-time and mean transmitral pressure gradient by concomitant on-line TEE monitoring. And it was also suggested that fluoroscopy time may be reduced with all the benefits of on-line TEE.Vilacosta et al11 and Ramondo et al14 reported that procedure time and fluoroscopy time could be reduced by on-
Journal of the American Society of Echocardiography Volume 13 Number 1
Park, Kim, Hyon 31
Figure 5 After trans-septal puncture, transesophageal echocardiography identified pericardial tamponade in 4-chamber (A) and 2-chamber view (B). Because of the deteriorating hemodynamic status of the patient, the transesophageal transducer was switched to transthoracic transducer to guide the pericardiocentesis. Subcostal view shows pericardial tamponade before pericardiocentesis (C) and after drainage of the pericardial blood by pigtail catheter (D).
line TEE. Our preliminary data also demonstrated shorter fluoroscopy time and procedure time.15 Trans-septal Puncture During trans-septal puncture TEE can visualize the relative position of the Brockenbrough needle and atrial septum (Figure 1, A). In 1 patient of group 2 who had pericardial tamponade after trans-septal puncture, the trans-septal needle may have tracked
through the slitlike space between the septum primum and septum secundum and punctured the atrial wall.The stainless steel wire made a big loop around the atrial wall instead of making small multiple loops within the left atrium.When this sign is observed, the operator should not introduce the dilator to enlarge the puncture site. To avoid this complication, the echocardiographer should double-check the exact position of the tip of the trans-septal needle in both
32 Park, Kim, Hyon
Journal of the American Society of Echocardiography January 2000
Figure 6 In a patient with heavily calcified mitral valve, especially the calcific nodule of the posterior mitral valve leaflet blocked entry of the balloon into the left ventricle (A, B). After passage of the guide wire into the left ventricle (C), the balloon glided easily into the left ventricle along the guide wire and was successfully inflated (D).
the vertical and the horizontal plane to see the tenting of the fossa ovalis membrane (Figure 1, B). Monitoring Balloon Position The relative position of the balloon and cardiac structure is well visualized by TEE, so when the balloon is in the left atrium, the relative direction of the long axis of the balloon and the mitral valve orifice can help the operator redirect the balloon. Once the balloon gets into the left ventricle, the distal half of the balloon is inflated, and the balloon shaft is pulled back to anchor the proximal part of the balloon into the mitral valve orifice before the balloon is fully inflated. During this procedure sometimes the balloon is engaged into the subvalvular space, and sometimes the balloon is fully inflated within the left ventricle under the valve (Figure 2, A). The chord rupture is likely to happen in those cases. As shown in Figure 6, when the balloon cannot be engaged into the mitral orifice because of an anatomic obstacle such as heavily calcified posterior mitral valve leaflet, one can choose to use J-wire or wedge balloon to pass the mitral valve orifice. In patients with left atrial appendage thrombi,TEE visualizes the rela-
tive position of left atrial appendage and the balloon so that the operator can avoid pushing the balloon into the left atrial appendage (Figure 3). Assessment of the Result of Valvuloplasty The immediate result of valvuloplasty is evaluated by Doppler pressure half-time and mean transmitral pressure gradient with continuous wave Doppler. And maximal separation of the mitral valve tip can be measured at the horizontal and vertical planes to assess the adequacy of the valvuloplasty. This information is combined with the possible development of new or worsening mitral regurgitation and the hemodynamic data acquired by catheterization to decide whether to upgrade the balloon size and repeat the procedure or to stop the procedure there. Early Detection of Complications Severe mitral regurgitation caused by papillary muscle rupture or chordae tendinae rupture, pericardial tamponade caused by cardiac perforation, or iatrogenic atrial septal defect may develop during balloon valvuloplasty, and those complications are easily detected by on-line TEE monitoring. Hemodynami-
Journal of the American Society of Echocardiography Volume 13 Number 1
cally insignificant mild mitral regurgitation may develop after valvuloplasty resulting from minor chord rupture or commissural tear. Hemodynamically significant severe mitral regurgitation has been reported to develop in 1% to 6% of the patients who undergo balloon mitral valvuloplasty.5,16,17 We experienced 2 cases of severe mitral regurgitation in group 2. One patient had severe mitral regurgitation caused by papillary muscle rupture immediately after valvuloplasty and underwent mitral valve replacement because of intractable symptoms after 1 month of medical treatment. The other patient had severe mitral regurgitation caused by papillary muscle rupture and cardiogenic shock after valvuloplasty and underwent emergency surgery (Figure 4). Operative findings showed papillary muscle rupture and posterior mitral valve leaflet tear extending to the posterior margin of the mitral annulus. One patient had pericardial tamponade after trans-septal puncture, and it was easily visualized by TEE, so the transducer was switched with a transthoracic one to guide the pericardial puncture site at the cardiac apex with the patient in the semi-Fowler’s position (Figure 5). Left Atrial Appendage Thrombus The presence of left atrial appendage thrombus has been regarded as a contraindication to balloon valvuloplasty. Hung et al16 reported that patients with left atrial thrombi were safely treated with balloon valvuloplasty after thrombi was successfully resolved with several months of anticoagulation. However, several cases of uneventful balloon valvuloplasty have been reported in patients with left atrial appendage thrombi.18,19 Five patients with left atrial appendage thrombi in group 2 underwent balloon valvuloplasty without embolic event (Figure 3). In 3 patients the left atrial appendage thrombi lysed after several months of anticoagulation at the time of follow-up TEE after valvuloplasty. So the presence of left atrial appendage thrombi even after 3 months of anticoagulation should not be regarded as a contraindication to balloon valvuloplasty, which is performed under TEE monitoring. Procedure Time and Fluoroscopy Time The procedure time in group 1 was 99 ± 48 minutes, which was similar to 104 ± 13 minutes that was reported by Bassand et al20 in valvuloplasty with an Inoue balloon without on-line TEE monitoring. The procedure time in group 2 was 64 ± 22 minutes, which is similar to 65 ± 18 minutes reported by Ramondo et al.14 In an effort to reduce radiation for pregnant women, we shielded the patients with lead
Park, Kim, Hyon 33
protectors and tried to minimize the fluoroscopy time so we could reduce fluoroscopy time to an average of 1.8 minutes (range 1 to 4 minutes) in 4 pregnant women. The average fluoroscopy time of group 1 (30 ± 17 minutes) was longer than that of group 2 (19 ± 15 minutes) without statistical significance. Bassand et al20 reported an average fluoroscopy time of 16 ± 8 minutes in patients who underwent valvuloplasty with an Inoue balloon without on-line TEE.TEE may contribute in reducing procedure time and fluoroscopy time in difficult cases with heavy calcification of mitral valve by recommending early use of a guide wire (Figure 6). Limitations This study included 2 patient groups: group 1, who underwent PBMV under fluoroscopy alone from 1991 to 1995, and group 2, who underwent PBMV under on-line TEE in addition to fluoroscopy from May 1995 to May 1996. Because this study was planned just before the beginning of the on-line TEEguided PBMV, the data of group 1 were acquired by retrospective review of the patient’s record. The learning period may be responsible for the higher rate of major complications and longer procedure time in group 1. But the operator was confident during the whole procedure, especially during trans-septal puncture and balloon positioning with the presence of on-line TEE guide. And this may have contributed to the lower major complication rate and shortening of the procedure time in group 2.
REFERENCES 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. Zaibag MA, Kasab SA, Ribeiro PA, Fagih MRA. Percutaneous double balloon mitral valvulotomy for rheumatic mitral valve stenosis. Lancet 1986;1:757-61. 3. Palacios I, Block PC, Brandi S, et al. Percutaneous balloon mitral valvotomy for patients with severe mitral stenosis. Circulation 1987;75:778-84. 4. McKay CR, Kawanishi DT, Rahimtoola SH. Catheter balloon valvuloplasty of the mitral valve in adults using a double-balloon technique. Early hemodynamic results. JAMA 1987; 257:1753-16. 5. Ruiz CE, Allen JW, Lau FYK. Percutaneous double balloon valvotomy for severe rheumatic mitral stenosis. Am J Cardiol 1990;65:473-7. 6. Block PC, Tuzcu EM, Palacios IF. Percutaneous mitral balloon valvotomy. Cardiol Clin 1991;9:271-87. 7. Den Heijer P, Hammer J, Pieper E, van Dijk R, Lie K. Percutaneous mitral balloon valvotomy guided by biplane transesophageal echocardiography [Abstract]. Eur Heart J 1991;12(Suppl):47.
Journal of the American Society of Echocardiography January 2000
34 Park, Kim, Hyon
8. Kültürsay H, Türkoglu C, Akin M, Payzin S, Soydas C, Akilli A. Mitral balloon valvuloplasty with transesophageal echocardiography without using fluoroscopy. Cathet Cardiovasc Diagn 1992;27:317-21. 9. Goldstein SA, Campbell AN. Mitral stenosis. Evaluation and guidance of valvuloplasty by transesophageal echocardiography. Cardiol Clinics 1993;11:409-25. 10. Goldstein SA, Campbell AN, Minz GS, Pichard A, Leon M, Lindsay Jr J. Feasibility of on-line transesophageal echocardiography during balloon mitral valvulotomy: experience with 93 patients. J Heart Valve Dis 1994;3:136-48. 11. Vilacosta I, Iturralde E, San Román JA, Gómez-Recio M, Romero C, Jiménez J, Martínez-Elbal L. Transesophageal echocardiographic monitoring of percutaneous mitral balloon valvulotomy. Am J Cardiol 1992;70:1040-4. 12. Abascal VM, Wilkins GT, Choog CY, Thomas JD, Palacios IF, Block PC, Weyman AE. Echocardiographic evaluation of mitral valve structure and function in patients followed for at least 6 months after percutaneous balloon mitral valvuloplasty. J Am Coll Cardiol 1988;12:606-15. 13. Kronzon I, Tunick PA, Schwinger ME, Slater J, Glassman E. Transesophageal echocardiography during percutaneous mitral valvuloplasty. J Am Soc Echocardiogr 1989;2:380-5. 14. Ramondo A, Chirillo F, Isabella G, Bonato R, Rampazzo C, Razzolini R, Andriolo L, Mazzucco A, Chioin R. Value and limitations of transesophageal echocardiographic monitoring
15.
16.
17.
18.
19.
20.
during percutaneous balloon mitral valvotomy. Int J Cardiol 1991;31:223-34. Park SH, Kim MA, Hyon MS, et al. Advantages of transesophageal echo guide during balloon mitral valvuloplasty [abstract]. J Am Soc Echocardiogr 1996;9:376. Hung JS, Chern MS, Wu JJ, Fu M, Yeh KH, Wu YC, Cherng WJ, Chua S, Lee CB. Short- and long-term results of catheter balloon percutaneous transvenous mitral commissurotomy. Am J Cardiol 1991;67:854-62. Nobuyoshi M, Hamasaki N, Kimura T, et al. Indications, complications and short-term clinical outcome of percutaneous transvenous mitral commissurotomy. Circulation 1989;80:782-92. Thomas MR, Monaghan MJ, Smyth DW, Metcalfe JM, Jewitt DE. Comparative value of transthoracic and transesophageal echocardiography before balloon dilatation of the mitral valve. Br Heart J 1992;68:493-7. Rittoo D, Sutherland GR, Currie P, Starkey IR, Shaw TRD. The comparative value of transesophageal and transthoracic echocardiography before and after percutaneous mitral balloon valvotomy: a prospective study. Am Heart J 1993;125: 1094-105. Bassand JP, Schiele F, Bernard Y, et al. The double balloon and Inoue techniques in percutaneous mitral valvuloplasty: comparative results in a series of 232 cases. J Am Coll Cardiol 1991;18:982-9.