Percutaneous closure was performed in 33 patients on 34 consecutive PVLs between July 2009 (first case) and March 2013. All had PVLs that produced serious symptomatic regurgitation (HF and/or hemolytic anemia requiring periodic red blood cell transfusion). Each patient's case was discussed in the medical-surgical session and was considered eligible for percutaneous PVL closure. Patients with single or multiple PVLs that, as a whole, affected more than a third of the prosthetic ring circumference were considered ineligible for percutaneous closure. All patients gave written informed consent for the procedure.

PVL was defined as the presence of a regurgitation jet on Doppler echocardiography that originated between the border of the prosthetic ring and the surrounding native tissue. The American Society of Echocardiography14 recommendations were used to define regurgitation severity. The site of the mitral PVLs was defined according to the classification previously proposed by Cortes et al,15 and the site of aortic PVLs was defined using the classification adopted by Ruiz et al.2

HF was diagnosed according to the classical Framingham criteria, and functional status was assessed at baseline and during follow-up according to the New York Heart Association classification. Hemolytic anemia was defined as plasma hemoglobin ≤ 14g/dL in men or ≤ 12g/dL in women, hemolytic profile (lactate dehydrogenase ≥ 600 U/L, haptoglobin ≤ 10mg/dL), and red blood cell transfusion within the past 6 months.

Technical success was defined as proper device implantation in the PVLs without interference with the prosthetic discs or a need for emergency conversion to conventional surgery. The procedure was considered to be successful if, in addition to the above criteria, the echocardiogram showed a decrease ≥ 1 degree in valvular regurgitation.

Procedure-related events were considered to be complications that occurred during the procedure or within the following 24 h (procedure-related death, cardiovascular death, acute cerebrovascular event, myocardial infarction, cardiac tamponade, vascular complications at the vascular access that required surgery, or red blood cell transfusion and emergency conversion to traditional surgery).

In all patients, the procedure was performed under general anesthesia and guided with transesophageal echocardiography (TEE), 2-dimensional in the first 6 patients and real-time 3-dimensional in all others (81.8%).

In all patients, the percutaneous closure of aortic PVLs was done by a retrograde approach, using previously described techniques.2,9,16,17 Briefly, a catheter with a hydrophilic guidewire is advanced to the aorta and used to cross the PVL; once the leak is crossed with the hydrophilic guidewire, the catheter is advanced and, once in the ventricle, switched to a guidewire with a larger support to allow the device delivery sheath to be advanced through the leak. The device is introduced over this sheath and implanted by sequentially delivering the distal disc in the ventricular aspect and the proximal disc in the aortic aspect of the leak.

Mitral PVL closure was performed using a retrograde or anterograde route, depending on patient anatomy and PVL site, using previously described techniques.2,9,16 Briefly, in the anterograde approach, transseptal puncture is performed and the PVL is crossed with a guidewire from the left atrium to the left ventricle, whereas in the retrograde approach, the guidewire is passed from the left ventricle through the leak to the left atrium. An arteriovenous loop was established in all patients (using a loop to snare the guidewire in the left atrium in the retrograde approach and in the aorta in the anterograde approach) and the delivery sheath was advanced from atrium to ventricle.

In patients with mitral PVLs and a 2-disc aortic prosthesis, once an anterograde access was attempted without success, the retrograde approach was used, crossing the central opening of the aortic prosthesis with a hydrophilic catheter and a straight guidewire, and the same steps were followed as in other retrograde approaches.

The AVP III device was used in all patients. This is an oval nitinol device with a height of 6.5 mm, and sizes that vary from 4 mm to 14 mm along the long axis and 2 mm to 5 mm along the short axis. The delivery sheaths vary from 4 Fr to 7 Fr, depending on the device size (Figure 1).

Figure 1.

Amplatzer Vascular Plug III device (St. Jude Medical).

(0.09MB).

The size of the PVL and, therefore, the size of the device to be implanted, was assessed by TEE. In patients who underwent 3-dimensional TEE, the color images acquired in 3-dimensions were used. The anatomic characteristics (area, width, and length) of each PVL were measured from the image of the complete volume using the multiplanar reconstruction tool, such that the longitudinal and coronal planes crossed in the area of the dehiscence visualizing regurgitant flow, whereas the transverse plane or short axis was used to measure the regurgitant orifice at the point of origin.

After transseptal puncture, all patients were given intravenous heparin to achieve an activated clotting time of 250s and were started oral anticoagulation after the procedure; heparin therapy was maintained until adequate levels were obtained with oral anticoagulants.

Patients were checked by the cardiology outpatient clinic and TEE was performed 90 days after the procedure. The following variables were collected: New York Heart Association functional class, need for red blood cell transfusion, all-cause death, cardiovascular death, acute cerebrovascular events, myocardial infarction, need for cardiac surgery, vascular complications requiring red blood cell transfusion or surgery and degree of valvular regurgitation.18

Clinical success in follow-up was considered to be proven clinical improvement ≥ 1 degree in New York Heart Association functional class in the 90 days after the closure procedure.

A retrospective descriptive study was performed. Continuous variables are presented as mean (standard deviation) and qualitative variables, as absolute number and percentage. The statistical analysis was performed using SPSS 16.0 (IBM SPSS Statistics).

In aortic PVLs, the retrograde approach was used in all patients. Among mitral PVLs, the retrograde approach was used in 57.7% of patients, the anterograde approach in 26.9%, and the retrograde approach crossing the mechanical aortic prosthesis to close the leak in 15.4% (Figure 2). In 1 patient, 2 contiguous devices were implanted using the anterograde approach to close 1 PVL. The procedure characteristics are summarized in Table 2.

Figure 2.

Device (arrow) implanted in the paravalvular leak of a patient with a mitral and aortic prosthesis. A, 3-dimensional echocardiographic image. B, angiographic image.

(0.14MB).

In 4 of 33 patients, the implant was initially unsuccessful. In 1 patient, the device moved after delivery and interfered with the mitral prosthetic discs, it could not be snared again and the patient underwent emergency surgery. In 1 patient, it was impossible to cross the PVLs (anterior PVL in biological prosthesis with no radiopaque structure) and the patient was scheduled for surgery. In 1 patient, transseptal puncture (atrial septal aneurysmal) was initially impossible and, therefore, the procedure was repeated using a steerable catheter (Agilis, St. Jude Medical) without complications (Figure 3). In 1 patient, the device interfered with the mitral discs by the angle of access to the PVLs and was not delivered; the procedure was rescheduled and repeated using a higher transseptal puncture with the aid of a steerable catheter, which allowed the device to be delivered without complications. In summary, technical success was achieved in 93.5% of patients, although 2 of them required a second planned procedure. In all patients except 1, implant success was accompanied by a decrease ≥ 1 1 regurgitation grade (procedure success, 90.9%).

Figure 3.

Angiographic images of the steerable catheter (Agilis, St. Jude Medical) (arrow) used for transseptal puncture and approach for paravalvular leaks.

(0.17MB).

Procedure-related complications occurred in 4 patients. As mentioned, 1 patient required surgery and 3 required red blood cell transfusion in the first 24 h after the procedure due to hematoma in the puncture region; all had hemolytic anemia and low hemoglobin levels before the procedure, and transfusions were given to aid recovery. There were no reports of death, acute myocardial infarction, stroke, or need for vascular surgery.

In 28 of 31 (90.3%) patients who received the device, an improvement ≥ 1 degree in functional class (clinical success at 90 days) was achieved. Only 3 patients experienced no improvement in functional class after the procedure (Figure 4). In addition, the need for periodic transfusions was significantly lowered with the closure procedure, from 51.5% to 9.1% (Table 3).

Figure 4.

Functional class before the procedure and 3 months after paravalvular leak closure. NYHA, New York Heart Association.

(0.11MB).

Complications occurred during follow-up in 4 (12.9%) patients, all of vascular etiology (femoral pseudoaneurysm). Two of them were patients who underwent closure for aortic PVL with an 8-Fr delivery sheath, which could have favored the appearance of pseudoaneurysm. In the other 2 patients (mitral PVL closures), the left femoral access was used to perform the arteriovenous loop with 6-Fr introducers; in 1 patient, closure was performed with a device and in the other, manual compression. Both patients presented obesity as a risk factor for vascular complications. Traditional surgery was performed in 3 patients, and percutaneous treatment with thrombin was chosen in the other, all with good outcome. No deaths, myocardial infarctions, strokes, or need for cardiac surgery were recorded during follow-up (Table 3).

In the follow-up TEE, valvular regurgitation was undetectable or trivial in 67.7% of patients and mild in 19.3%. Only 4 (12.9%) patients showed moderate valvular regurgitation (Table 3).

This is a single-center descriptive study and, therefore, the conclusions may not be applicable to other settings. The study was not randomized and, therefore, the efficacy of the AVP III was not compared with that of other devices. The small number of patients made it difficult to draw conclusions on the factors that influenced the appearance of vascular complications. The follow-up presented is short term; hence, the medium- and long-term clinical and echocardiographic repercussions of this procedure with the AVP III device are unknown.