INTRODUCTION

Perivalvular leaks (PVL) are common following surgery. If we use a transesophageal echocardiogram (TEE), they are detected in up to 17.6% of all cases after aortic valve replacement and in 22.6% of all cases after mitral valve replacement.1,2 It normally develops when one of the sutures breaks, and is more prevalent in patients with a diseased valve annulus, whether due to infectious endocarditis, calcium, or previous surgery.

These PVLs are normally small and have no significant clinical repercussions. They tend to improve or disappear over time, but they can occasionally give rise to serious complications, such as haemolytic anaemia requiring multiple blood transfusions or congestive heart failure.

Re-intervention due to PVL is uncommon (<2%), but it is the treatment of choice for the vast majority of patients,3 even though mortality rates are high (>10%) and recurrence can be frequent.4,5

Currently, the use of Amplatzer occluders in the percutaneous treatment of certain congenital heart diseases is a very common practice.6-8 Their use in the treatment of PVLs in isolated cases or short series has been described recently.9-11

We would like to describe our initial experience with the use of Amplatzer occluders for the percutaneous closure of PVLs.

METHODS

Patients

Between 2004 and 2006, it was decided to perform percutaneous closure of PVLs in 12 patients, all of them suffering from severe regurgitation. After being evaluated for this procedure and cardiac surgery, percutaneous closure was recommended for 8 patients because they suffered from heart failure (NYHA class III-IV), severe haemolytic anaemia needing multiple blood transfusions, or from both conditions, in addition to having a high surgical risk level or technical difficulties being anticipated. Four patients were excluded due to having one or more of the following: multiple leaks, low surgical risk, or a baseline state with few systems.

A transthoracic echocardiogram and a TEE were performed to confirm the diagnosis and determine the size, location, and number of leaks. Regurgitation was determined as severe according to the parameters established by the American Society of Echocardiography, correlated with the Seller degree from the angiogram.12

The haemolytic anaemia was defined as having a concentration of haemoglobin <10g/dL in the last 2 weeks, with biochemical parameters compatible with haemolysis. Anaemia was considered to be severe when blood transfusions were necessary.

The EuroScore was calculated to predict operation mortality and evaluate the surgical risk.

Patients were monitored during the 24 hours following the procedure. Control echocardiograms were performed before the patients were discharged, and clinical and echocardiographic follow-ups were carried out after 6 and 12 months.

Procedure Technique

Prior to the intervention, an infectious endocarditis prophylactic (cefminox 2g intravenously) and 5000 to 10 000 IU of sodium heparin were administered. All procedures were carried out using local anaesthesia, and only in 1 case was TEE used as a guide measure.

The procedure was carried out according to the technique described previously.13,14 Closure of mitral PVLs was performed with the anterograde approach, using transeptal catheterisation done with a Brockenbrough needle and Mullins introducer sheath. A diagnostic catheter (right Judkins or multi-purpose) is advanced along the Mullins sheath to the left atrium. A Terumo hydrophilic guidewire was passed through the catheter to cross the PVL, the catheter was advanced across the PVL, and the hydrophilic guidewire was replaced with a 260 cm high support guidewire. Next, the Amplatzer delivery system was advanced along; once it was positioned over the opening, the Amplatzer duct occluder (AGA Medical Corporation, Golden Valley, United States) was released, after verifying that it would not interfere in the proper functioning of the prosthesis (Figure 1). The device was at least 2 mm larger than the estimated dimensions obtained by the echocardiogram and the angiogram.

Figure 1. Procedure to close the mitral perivalvular leak guided by TEE due to presence of a mechanical aortic and mitral prosthesis. A: Advancing the delivery sheath through the opening. B: Occluder is released. C and D: The TEE during the procedure shows the decrease in the degree of mitral regurgitation.

The aortic PVLs were closed using the retrograde method through the right humeral artery, since the length of the Amplatzer delivery system (80 cm) makes femoral access impossible, except in very small patients. The hydrophilic guidewire was advanced through an Amplatz-type diagnostic catheter until it crossed the PVL, and the rest of the procedure was similar to that previously described (Figure 2).

Figure 2. Procedure to close the aortic perivalvular leak. A: Initial aortography. B: Passing the hydrophilic guidewire. C: Releasing the occluder. D: Final angiographic check showing a decreased degree of regurgitation.

RESULTS

Four of the PVLs were aortic and the other 4 mitral. The baseline clinical characteristics are summarised in Table.

The average mortality rate based on a pre-operation estimate using the EuroSCORE was 21.5% (interval, 5.48%-54.82%). The patient with a EuroSCORE of 5.48% had undergone a second operation for PVL and had a history of infectious endocarditis.

Implementation, Complications, and Immediate Results

The percutaneous closure of the PVL was undertaken after a median of 79 months (7-264) since the most recent valve replacement. In 1 case of mitral PVL, a tricuspid valvuloplasty was performed simultaneously.

The implantation of the device was possible in 7 patients (87.5%), although in 1 case of aortic PVL a second procedure was necessary (the device used in the first attempt was not the right size, due to the PVL being larger than the estimate given by the echocardiogram). In 1 patient with aortic PVL, the implantation of the device was not feasible due to the difficulty of passing the delivery system through the PVL. There were no complications during the procedure.

In 2 of the 3 patients with aortic PVL and an implanted device, the decrease in the degree of regurgitation was observable by angiogram and echocardiogram. In 2 of the 4 patients whose mitral PVL was treated, the degree of regurgitation lessened; severe perivalvular regurgitation continued in the others.

During the medium-term follow-up (Table), those patients whose degree of regurgitation had been reduced (4 of 7) showed a clinical improvement. One patient with severe ventricular dysfunction needed to have a defibrillator implanted due to sustained ventricular tachycardia.

Of the 3 patients whose degree of regurgitation did not improve, 1 underwent a procedure a month later with no complications and 2 died of non-cardiac causes (pneumonia 30 days later, and a digestive haemorrhage 14 months later).

DISCUSSION

Symptomatic PVL, for which re-intervention is indicated, has a low incidence rate. Re-operation is the treatment of choice for the vast majority of patients, and many published surgical series refer to a mortality rate <10%, which is a better result than that given by medical treatment.15 However, these results may be underestimated because some patients with a very high surgical risk are never referred for re-intervention when percutaneous treatment with Amplatzer occluders could be a therapeutic possibility.

In our study, the patients were very symptomatic and the operation risk was high with an average EuroSCORE of 21.5%. We successfully implanted the device in 87.5% of the patients and in 57% of the cases, clinical improvement was observed which was associated with reducing periprosthetic regurgitation. This effect (GF I or II) continued after 12 months of follow-up. There were no cases of mortality or serious complications during the procedure (for example, embolisation of the occluder or interference with the prosthesis). This suggests that the technique may be safe. During follow-up, no cases of haemolytic anaemia were detected. The regurgitation that persisted in some patients was possibly due to incomplete obstruction of the opening, which in turn may be due to choosing the wrong size of occluder, multiple openings, and most of all, use of the circular Amplatzer occluder which is designed for closing certain congenital heart diseases and not PVLs, which may have a half-moon shape.

The clinical and angiographic improvements to the mitral and aortic valves were similar despite an aortic implantation error. Although technical differences exist in the treatment, which is more complex for mitral PVLs and may affect the result, the result does also depend on the type of PVL.

The literature describes isolated cases and series with reduced numbers, mostly involving mitral PVLs that are always treated under general anaesthesia and guided with TEE, which have results similar to our own. Shapira et al,9 in a study of 11 cases (8 mitral) reached an implantation rate of 90.9% and a reduction of 60% in the degree of regurgitation. Pate et al,10 in a study of 10 patients (9 mitral) reached an implantation rate of 70% and saw angiographic success in 60%, but 1 patient did require emergency surgery.

Limitations

The principal limitations are that the series is small and studied in one centre, and that the results depend on the operator, given that the technique is complex.

In addition, the Amplatzer occluder is not specifically designed for treating PVLs, which could explain certain failures and difficulties with the technique.

In conclusion, the results seem to indicate that percutaneous treatment of PVLs, whether mitral or aortic, is technically feasible and safe, and can be used as an alternative treatment for patients with a high surgical risk level.

Correspondence: Dr. A.J. Muñoz-García.
Campus de Teatinos, s/n. 29010 Málaga. España.
E-mail: ajmunozgarcia@secardiologia.es

Received February 19, 2008.
Accepted for publication July 15, 2008.