Ventricular septal defect following mitral valve replacement

Abstract

Mitral valve surgery (MVS), with repair preferred to replacement, is a common procedure for the treatment of severe primary mitral regurgitation related to leaflet prolapse. Structural complications after MVS include left ventricular outflow obstruction, paravalvular leak and atrial septal defect. Intraoperative transoesophageal echocardiography and predischarge transthoracic echocardiography (TTE) specifically screen for these complications. Ventricular septal defect (VSD), a known complication after aortic valve surgery, is rarely reported after MVS. Recently, unsuccessful valvuloplasty prior to replacement was suggested as a risk factor. We present such a case and explore mechanisms with advanced cardiac imaging. In this case, the patient was found to have an elongated membranous septum that likely predisposed her to septal injury. Finally, we provide guidance on specific transoesophageal/transthoracic echocardiography views to avoid a missed diagnosis.

INTRODUCTION

Mitral valve surgery (MVS), with repair preferred to replacement, is a common procedure for the treatment of severe primary mitral regurgitation (MR) related to leaflet prolapse [1]. Structural complications after MVS include left ventricular outflow obstruction, paravalvular leak and atrial septal defect [2]. Intraoperative transoesophageal echocardiography (TOE) and predischarge transthoracic echocardiography (TTE) are specifically used to screen for these complications. Ventricular septal defect (VSD), a known complication after aortic valve surgery, is rarely reported after MVS. Recently, unsuccessful valvuloplasty prior to replacement was suggested as a risk factor. We present such a case and explore mechanisms with advanced cardiac imaging. In this case, the patient was found to have an elongated membranous septum that likely predisposed her to septal injury. Finally, we provide guidance on specific TOE/TTE views to avoid a missed diagnosis.

CASE DESCRIPTION

A 73-year-old female with long-standing moderate MR was referred for MVS after developing symptomatic severe MR. TOE showed a myxomatous mitral valve with P2 scallop prolapse causing severe MR. Concomitant coronary artery bypass grafting was planned because a coronary angiogram showed severe proximal left anterior descending artery stenosis. The patient underwent a median sternotomy with standard cardiopulmonary bypass. The left internal mammary artery was anastomosed to the left anterior descending artery. Next, a left atriotomy was made via a transseptal approach. Mitral valve repair using an Edwards Physio II 32-mm annuloplasty ring and P2 triangular resection was attempted. This attempt was unsuccessful with moderate residual MR related to calcified anterior chordae. Thereafter, a mitral valve replacement was performed in a standard fashion using a St Jude Epic 29-mm bioprosthetic valve. Intraoperative TOE showed a well-functioning prosthesis without any complications, and the patient was successfully weaned off bypass. The patient had an uneventful postoperative course with quick extubation and haemodynamic stability. Pre-discharge limited TTE was unremarkable with no VSD reported.

At her routine post-discharge clinic visit, she was noted to have a loud holosystolic murmur. She was otherwise doing well and was asymptomatic. A new TTE was obtained and surprisingly showed a new restrictive basal VSD with a left–right shunt, along with a well-functioning mitral prosthesis and normal cardiac function (Fig. 1, Video 1). Cardiac magnetic resonance (CMR) imaging was recommended and confirmed the small basal inferoseptal VSD that was haemodynamically insignificant with a pulmonary-to-systemic shunt ratio (Qp/Qs) of 1.1. Ventricular volumes and contractility values were normal. There was no infarction or fibrosis near the VSD on late gadolinium enhancement imaging (Fig. 2), which excluded an ischaemic aetiology. A detailed review of her prior preoperative echocardiograms confirmed the absence of a congenital VSD (Fig. 1, Video 2) though suggested a long membranous septum (MS). A cardiac computed tomography scan was obtained to further understand the mechanism of this iatrogenic VSD and it showed a very long MS with a small VSD (Fig. 2). Supplementary figure 1 shows an illustration of the surgical findings and proposed mechanism of septal injury. Given the small size of the VSD and the asymptomatic status, non-surgical conservative management was recommended. She is doing well 1.5 years after the operation and will be followed with an annual TTE in the cardiology clinic.

DISCUSSION

This case report describes an extremely rare and unexpected complication of VSD after MVS. Literature on this phenomenon describing the mechanism and risk factors is extremely scarce. Prior case reports mention unsuccessful valvuloplasty before valve replacement, suboptimal suture placement or imperfect surgical closure of a paravalvular leak as risk factors [2]. MVS is performed via a direct left atriotomy through the interatrial groove or a transseptal approach. The transseptal approach is associated with a higher incidence of postoperative conduction issues. It is not known whether a particular approach is associated with a higher risk of VSD, though there is extensive literature regarding overall similar safety profiles. The primary mechanism that we wanted to exclude with advanced cardiac imaging was ischaemic, possibly related to failure of the bypass graft or another coronary injury. The CMR was very helpful in this regard because it excluded any infarction in the region of the VSD. The cardiac computed tomography scan was then obtained to understand the 3-dimensional relation of this VSD to her MVS. The scan confirmed the elongated MS measuring 2.2 cm and its close relation to the VSD, allowing us to hypothesize about the role of the elongated MS in facilitating septal injury. In our case, we believe it was the combination of the failed valvuloplasty weakening the septum in the medial trigone area, along with the elongated MS that was injured during the surgery, possibly from the rigid cardiotomy sucker. The MS is a thin fibrous membrane that is about 1 cm long, though an elongated MS has been reported [3]. The length of the MS is recognized as an important predictor of complications after transcatheter aortic valve replacement, with a shorter length associated with more pacemaker implants, and a longer length associated with a risk of VSD [4]. As such, elongated MS may be a risk factor for VSD after MVS. Preoperative knowledge of elongated MS should be sought, and when present, appropriate measures should be taken to reduce the risk of injury. The potential for injury to the surrounding structures during MVS was discussed recently in an editorial in this journal [5].

This case also highlights the fact that small VSDs may be missed on TOE/TTE if not specifically looked for. In our case, the VSD was located inferiorly in the membranous septum, an area that is often not well seen on TTE. Sonographers should specifically check the ventricular septum, along with inferior angulation from the apical 4-chamber view, in all patients after MVS.

ACKNOWLEDGEMENT

We would like to thank Cynthia Faraday for her assistance with the figures.

Funding

This study received no specific grant from any funding agency in the public, commercial or not-for-profit sectors.

Conflict of interest: No authors have any conflicts of interest to report.

Data availability

The data underlying this article cannot be shared publicly due to patient privacy.

REFERENCES