Secondary Logo

Journal Logo

A Left Ventricle to Left Atrial Appendage Fistula After Mitral Valve Replacement

Liang, Yafen MD*; Sileshi, Bantayehu MD*; Kaiser, Clayton A. MD; Pretorius, Mias MB, ChB, MSCI*; Shaw, Andrew D. MB, FRCA, FCCM, FFICM*

doi: 10.1213/XAA.0000000000000705
Echo Rounds

From the Departments of *Anesthesiology and Cardiac Surgery, Vanderbilt University Medical Center, Nashville, Tennessee.

Accepted for publication August 24, 2017.

Funding: None.

The authors declare no conflicts of interest.

Supplemental digital content is available for this article. Direct URL citations appear in the printed text and are provided in the HTML and PDF versions of this article on the journal’s website.

Institutional review board: Only echocardiogram images are published in this manuscript. No patient identifiers are shown. The institutional review board deemed this a nonhuman subject research study. The patient gave written consent for publication of these images.

Address correspondence to Yafen Liang, MD, Department of Anesthesiology, Vanderbilt University Medical Center, Nashville, TN 37232. Address e-mail to

A 52-year-old woman with Marfan syndrome who had mitral valve replacement (MVR) 10 years ago presented for redo-MVR due to progressive mitral stenosis and shortness of breath. Intraoperative transesophageal echocardiography (TEE) confirmed the diagnosis of mitral stenosis. The patient underwent uneventful femoral arterial and venous cannulation for cardiopulmonary bypass (CPB). A right anterolateral thoracotomy was made. The pericardium was opened, and a left atriotomy was made. The preexisting bioprosthetic mitral valve (MV) was dissected and explanted. The dissection was notable for extensive adhesion of the prosthesis to the left ventricle (LV) wall and papillary muscles. A 33-mm mitral valve bioprosthesis was reimplanted. After weaning from CPB, a well-seated and functional bioprosthesis was demonstrated with TEE. However, a modified midesophageal 2-chamber view showed a fistulous tract between the LV anterolateral wall and left atrial appendage (LAA) measuring about 8 mm in diameter during systole at the narrowest point (Figure 1A). Color flow Doppler (CFD) revealed an accompanying systolic turbulent jet. This jet originated in the LV, traveled through a fistula into the LAA, and finally into the left atrium (Figure 1B; Supplemental Digital Content 1, Video 1, In addition, a newly developed inferior pericardial effusion was observed. CPB was reinstituted immediately to facilitate repair of the fistula. After primary repair, TEE showed a small residual fistulous tract expanding to 1 mm during systole (Figure 2A), with significantly reduced fistula flow (Figure 2B; Supplemental Digital Content 2, Video 2, Given the complexity of the surgery, tissue fragility, and favorable hemodynamics, a decision was made not to pursue further repair of the fistula. The pericardial effusion (which had expanded circumferentially with LV anterolateral wall impingement during diastole [Supplemental Digital Content 3, Video 3,]) was drained via a left intercostal approach due to adhesions. Patient was taken to the intensive care unit and extubated soon after.

Figure 1.

Figure 1.

Figure 2.

Figure 2.

On postoperative day (POD) 2, a small regurgitation jet was noticed on TEE. The patient decompensated on POD 10 with TEE imaging showing recanalization of fistula. She was taken to the catheterization lab and had percutaneous closure of the fistula. She had a slow recovery and was ultimately discharged on POD 39.

Back to Top | Article Outline


LV to LAA fistula is a rare complication after MVR. TEE has a crucial role in locating the fistula, assessing the hemodynamic consequences of the fistula, assisting in differential diagnosis, and assessing adequacy of surgical repair.

Normally, the LAA is separated from the LV by the MV and the pericardial space. In the event of an LV to LAA fistula, a large pericardial effusion should develop. However, in this case, adhesions allowed the pericardial effusion to be contained. The LV to LAA fistula is a type III LV rupture due to its location between the atrioventricular groove (type I) and posterior papillary muscle (type II).1 The LV rupture is usually created by inappropriate surgical techniques, such as forceful retraction of the LV, extensive resection of papillary muscles, impingement of the LV wall by a large prosthesis strut, etc.1 In this case, the fistula was most likely created during dissection of the existing mitral prosthesis from the LV wall.

To diagnose such a fistulous tract, midesophageal views from commissural to past the standard 2-chamber view (rotation beyond 90°–100°) should be obtained in which both the left ventricle and LAA are adequately visualized. The CFD sector should include the LV anterolateral wall and the LAA because abnormal flow in the LAA can be the first indication of this complication. Spectral Doppler of the LAA in the setting of LV to LAA fistula will reveal high-velocity systolic flow (aliasing of pulsed-wave Doppler), compared with normal LAA flow (biphasic diastolic and early systolic low-velocity flows in patients with normal sinus rhythm, variable low-velocity flows in patients with atrial fibrillation)2 or pulmonary vein flow (atrial reversal followed by biphasic systolic flow and diastolic flow, velocity usually <80 cm/s).3 The mitral annulus calcification or mitral prosthesis can cause acoustic shadowing in the LV wall, which could mimic a defect in the LV wall and make diagnosis difficult. However, in this setting, no regurgitation flow should be visualized in the LAA when CFD is applied. Multiplane mitral valve “clock face” mapping could be used to determine the location and extent of the fistula.4 Due to the eccentric nature, the grading of such flow can be challenging. An orifice width determined by the fistula tract during systole (as fistula expands during systole when the pressure gradient between the LV and LAA is the largest) could be used to estimate the severity. Three-dimensional imaging could add additional value to more precisely locate the fistula in “en face” view of the mitral prosthesis, evaluate the extent of the fistula and effective regurgitation orifice area (since the shape of the fistula tract varies), and thus grade the severity.5

The differential diagnoses of LV to LAA fistula include a paravalvular leak or a coronary artery to left atrium fistula. LV to LAA fistula can reportedly mimic a paravalvular leak.6 Paravalvular leaks are usually located outside the sewing ring but inside the mitral annulus, not distal in the LV wall, as was seen in this case. Unlike the LV to LAA fistula in which regurgitation flow occurs during systole, the flow from coronary artery to left atrium or LAA occurs throughout the cardiac cycle, and peaks during early systole when the pressure gradient between aortic root and left atrium is largest.7 Pulsed-wave Doppler can determine the phase of cardiac cycle when this flow occurs, and may help differentiate between these 2 etiologies. Concomitant pathologies such as LAA to LV outflow tract connection has been reported,6 thus those areas should be interrogated as well. Postsurgical repair, 2-dimensional imaging, and CFD should be applied at the original fistula location and LAA to identify any residual connection and abnormal flow.

In conclusion, LV to LAA fistula is a rare complication after MVR and a variant of contained LV rupture in this case. Intraoperative echocardiogram plays an important role in identifying the pathology and guiding surgical repair.

Back to Top | Article Outline


Name: Yafen Liang, MD.

Contribution: This author helped to participate the case and prepare the manuscript.

Name: Bantayehu Sileshi, MD.

Contribution: This author helped prepare the manuscript.

Name: Clayton A. Kaiser, MD.

Contribution: This author helped prepare the manuscript.

Name: Mias Pretorius, MB, ChB, MSCI.

Contribution: This author helped prepare the manuscript.

Name: Andrew D. Shaw, MB, FRCA, FCCM, FFICM.

Contribution: This author helped prepare the manuscript.

This manuscript was handled by: Nikolaos J. Skubas, MD, DSc, FACC, FASE.

Back to Top | Article Outline


1. Sersar SI, Jamjoom AA. Left ventricular rupture post mitral valve replacement. Clin Med Cardiol. 2009;3:101113.
2. Beigel R, Wunderlich NC, Ho SY, Arsanjani R, Siegel RJ. The left atrial appendage: anatomy, function, and noninvasive evaluation. JACC Cardiovasc Imaging. 2014;7:12511265.
3. Bartzokis T, Lee R, Yeoh TK, Grogin H, Schnittger I. Transesophageal echo-Doppler echocardiographic assessment of pulmonary venous flow patterns. J Am Soc Echocardiogr. 1991;4:457464.
4. Foster GP, Isselbacher EM, Rose GA, Torchiana DF, Akins CW, Picard MH. Accurate localization of mitral regurgitant defects using multiplane transesophageal echocardiography. Ann Thorac Surg. 1998;65:10251031.
5. Azran MS, Romig CB, Locke A, Whitley WS. Echo rounds: application of real-time 3-dimensional transesophageal echocardiography in the percutaneous closure of a mitral paravalvular leak. Anesth Analg. 2010;110:15811583.
6. Valbuena-López S, López-Fernández T, de Torres Alba F, et al. Left atrial appendage fistula mimicking a paravalvular prosthetic regurgitation. J Am Coll Cardiol. 2013;62:e23.
7. Renew JR, Ritter MJ. Coronary artery fistula to left atrium uncovered after mitral valve replacement. Anesth Analg. 2017;124:3032.

Supplemental Digital Content

Back to Top | Article Outline
Copyright © 2017 International Anesthesia Research Society