1

Transcatheter Closure of Perimembranous Ventricular Septal Defects in Infants and Children Using the Amplatzer Perimembranous Ventricular Septal Defect Occluder

Basil (Vasilios) D. Thanopoulos,a MD, PhD, Michael L. Rigby,b MD, Evangelos Karanasios,a

MD, Christodoulos Stefanadis,c MD, Nico Blom,d MD, JaapOttenkamp,d MD, Armine

Zarayelyan, e MD, PhD, a Department of Cardiology, “Aghia Sophia” Children’s Hospital, Athens, Greece bDepartment of Pediatric Cardiology, Royal Brompton Hospital, London, UK cDepartment of Cardiology, Hippokration Hospital, University of Athens, Greece dDepartment of Pediatric Cardiology, Leiden University Medical Center, Netherlands e Department of Pediatric Cardiology, Yerevan State Medical University, Armenia

Running title: Device closure of PMVSDs.

Corresponding author:

Basil (Vasilios) D. Thanopoulos

Department of Pediatric Cardiology

“Aghia Sophia” Children’s Hospital

Thivon and Levadias Street

Athens 115 27

Tel: 011-30 210 7794377

Fax: 011-30-210 7797649 -7794377

E-mail: [email protected], [email protected]

There were no sources of financial support 2

ABSTRACT

There are very few published reports of transcatheter closure of perimembranous ventricular septal defect (PMVSD) using the Amplatzer PMVSD occluder with encouraging initial results. In this report we present initial and one-year results from 54 patients with PMVSDs who underwent transcatheter closure at five different Institutions with the Amplatzer

PMVSD occluder. Sixty-five patients with PMVSDs were enrolled in 5 European centers.

Eleven of 65 patients did not fulfil the patient selection criteria at the initial echocardiographic evaluation or at cardiac catheterization. As a result a total of 54 patients underwent an attempt of transcatheter closure using the Amplatzer PMVSD occluder. The mean age of the patients was 5.1 ± 3.6 years (range, 0.3 to 13 years) and the mean weight

18.5 ± 10.3(range, 5 to 45 Kg). The devise was permanently implanted in 49 of 54 patients.

Complete occlusion of the communication at the one-year follow-up was observed in 46 / 49 patients (94%). Main complications included: Early-procedural. a. Device embolization (2 patients), b. severe bradycardia with hemodynamic compromise (2 patients) c. Mobitz II

(2:1) heart block (1 patient). Late. Complete heart block (1 patient). No other complications were observed during follow-up. In conclusion, the Amplatzer PMVSD occluder is promising device that can be used for transcatheter closure in selected patients with

PMVSDs. Further studies and long-term follow-up are required before this technique enters routine clinical practice.

Key words: Perimembranous ventricular septal defects, catheter closure 3

Isolated ventricular septal defect accounts for approximately 30% of all forms of congenital heart disease.1, 2 Eighty percent of these defects are perimembranous involving the membranous septum and the adjacent area of muscular septum.3 Although operative closure of a perimembranous ventricular septal defect (PMVSD) is a straightforward and safe procedure in the current era, it is associated with morbidity, postoperative discomfort and a sternotomy scar.2, 3 As an alternative to surgical closure during the past 15 years a variety of occluding devices have been used for closure of PMVSDs, but none has gained wide acceptance.4-9 The main limitations of these occluders, which were not originally designed for this purpose, include large delivery sheaths, inability for recapture, interference with the aortic and tricuspid valves and a very high rate of residual shunts.4-9 Recently, a new

Amplatzer occluder10 has been used to close PMVSDs in humans with quite satisfactory early results.11-14 In this study, we report initial and one-year results from 54 patients with

PMVSDs who underwent attempted transcatheter closure using the Amplatzer PMVSD occluder.

Methods

Patient population: Between May 2002 and June 2005, a total 65 patients from 5 cardiac centers in Europe were enrolled for attempted transcatheter closure of a PMVSD with the Amplatzer PMVSD occluder. Eleven patients were under one year of age (range 4-11 months). Twenty patients had symptoms of heart failure and/or poor weight gain (body weight at or below 10th percentile for age). All patients were selected according to a previously agreed protocol with full and informed consent. All devices were implanted in the context of a research protocol approved by the Ethical Committees of the participating

Hospitals. 4

Patients were screened by conventional transthoracic two-dimensional and color

Doppler echocardiography. The criteria for inclusion in the study were: a) a distance of at least 2 mm from the PMVSD to the aortic valve b) a left to right shunt across the defect with left ventricular enlargement (end diastolic dimension > 95% for age ) c) body weight  5 Kg and d) calculated pulmonary vascular resistance < 8 Wood units (Rp/Rs < 0.4). Any patient with an important mal-alignment of the outlet septum, overriding of the aorta or pulmonary trunk, aortic valve prolapse or inlet (atrioventricular canal type defects) ventricular septal defect was excluded from the study.

Device and delivery system: The Amplatzer PMVSD occluder (AGA Medical

Corporation, Golden Valley, MN) has been previously described in detail.10-13 The most important property of this occluder is the asymmetry of its left disk which minimizes the risk of interference with aortic valve movement.

Procedure: The technique of transcatheter closure of PMVSDs using the Amplatzer

PMVSD occluder has been previously described in detail.11-13 All patients underwent right and left cardiac catheterisation and angiography under general anaesthesia and at the same time were studied by transesophageal echocardiography. The device size (waist diameter) was selected to be 1 –2 mm larger than the largest measured diameter of the defect as assessed by transesophageal echocardiography (short and long axis views of left ventricular outflow tract) or by angled angiocardiography (Figure 1A). However, in the presence of a large associated septal aneurysm a device size equal or 1-2 mm larger that the left ventricular diameter of the septal aneurysm was selected. Transesophageal echocardiography and hand injections of contrast medium through the arterial catheter (Pigtail) were utilized for the optimal guidance of the procedure. After release of the device both color Doppler echocardiography and selective angiocardiography were performed to detect aortic and/or atrioventricular valve regurgitation and residual shunts (Figure 1BC). Residual shunts were 5 angiocardiographically and echocardiographically graded as foaming, trivial, small, moderate and large as previously described. 15 Patients were discharged the day after the procedure and taking aspirin 3 to 5 mg / kg daily for six months. Before discharge an electrocardiogram, a biplane chest X-ray, and an echocardiographic study were performed. All patients had a chest

X-ray, an electrocardiogram, a 24-hour Holter electrocardiogram recording, and complete transthoracic two-dimensional and color Doppler echocardiographic studies at 1 and 3 months after the procedure and then serially every 6 months. All echocardiographic measurements were made on line and verified by an experienced echocardiographer. A complete hemodynamic and angiographic study was scheduled for the 12-month follow-up.

Endocarditis prophylaxis was discontinued at the 12-month follow-up if the defect was completely closed and no aortic regurgitation was observed.

Statistical analysis. Results were expressed as mean value  SD, with confidence intervals given were applicable. The student paired t test was used to compare left-ventricular end-diastolic dimension before and after device closure. A p value of < 0.05 was considered statistically significant.

Results

During the study period 65 patients with PMVSD were enrolled for potential device closure. Of this group 8 patients did not fulfill the selection criteria for catheter closure at initial echocardiographic evaluation or cardiac catheterization (4 patients had poor echocardiographic window) and were excluded from device implantation. Also, a device was not permanently implanted in five patients. In 1 patient the right disc of the occluder interfered with the septal leaflet of tricuspid valve causing significant regurgitation. As a result the device was not released and was removed. In a second patient (infant 5 month old weighing 5 kg) the left disc of a 10 mm occluder protruded too far into the LV outflow tract 6 and was retrieved. A third patient (infant 6 month old weighing 6 kg) developed a Mobitz II

(2:1) atrioventricular block (AVB) (Heart rate: 70/min) immediately after the delivery of the right disc of a 10 mm device. Therefore the device was retrieved within the sheath and removed. Full recovery of the atrioventricular conduction was observed 24 hours after the removal of the Amplatzer PMVSD occluder. In two additional patients (infants of 4 and 5 months old weighing 5 and 5.5 Kg, respectively) the procedure was abandoned because they developed severe bradycardia (heart rate 50 min) with hemodynamic compromise during the attempts to place the long sheath in the LV in preparation for device implantation. In all patients with unsuccessful device implantation the defect was later surgically closed. The median age of the patients was 5.1 ± 3.6 years (range 0.3 to 13 years) and the median weight

18.5 ± 10.3 kg (range 5 to 45 kg). The pulmonary systemic flow (Qp/Qs) varied from 1.9 to

2.7 (mean 2.3 ± 0.2). The PMVSD diameter ranged from 5 to 15 mm (mean 7.8 ± 2.5 mm).

The device size ranged from 6 to 16 mm (mean 9.3 ± 2.5 mm). Six patients had an associated large ventricular septal aneurysm (largest diameter  10 mm) (Figure 2). Forty- five patients had isolated PMVSD, 2 patients a PMVSD associated with mild subaortic stenosis (Thin subaortic membrane with a peak Doppler pressure gradient 20 mm Hg and 27 mm Hg, respectively), and 2 patients a residual PMVSD following anatomic correction of transposition of the great arteries. According to the old time–honored classification [2] forty- four patients had a PMVSD type II and 5 patients a Gerbode like 3 PMVSD (Figure 3). After release of the device left ventriculography and color Doppler echocardiography revealed no residual shunt in 40 of 49 patients (81.6 % closure rate). In 7 and 2 patients there was a trivial and small residual shunt after the procedure, respectively. Seven patients had a trivial (5 patients) to mild (2 patients) tricuspid regurgitation before the procedure. In two patients trivial regurgitation progressed to mild (+) after placement of the device. In the remaining 5 tricuspid regurgitation remained unchanged. No patient developed aortic or mitral 7 regurgitation. In 1 patient with residual PMVSD after anatomic correction for transposition of the great arteries, a preexisting mild aortic regurgitation remained unchanged. The echocardiographic (M-mode) mean left-ventricular end-diastolic dimension was 45 ± 8 mm before device closure and decreased to 40 ± 7 mm (p< 0.05) the day after the procedure.

Fluoroscopy and total procedural times ranged from 15 to 60 min ( mean 40 min ) and from

116 to 159 min ( mean 145 min ), respectively.

Complications: During the procedure 5 patients (9.4%), including 3 infants less than one year of age, developed transient complete left bundle branch block (normal PR interval) which resolved within 24h. A sixth patient with left anterior hemiblock immediately after device placement and normal for age PR interval went on to develop intermittent complete

AVB 10 days following the procedure. This was a 4 year old female with Down syndrome who weighed 12 Kg at the time of the procedure and who had received a 10 mm occluder.

Although the patient was asymptomatic with a heart rate of 80/min and normal PR interval

(10 msec) a temporary pacemaker was placed due to the presence of pauses longer than 4 seconds in the Holter electrocardiogram. Full recovery of the atrioventricular conduction was observed 5 days later.

Transient supraventricular or ventricular arrhythmias without hemodynamic instability were common during the procedure particularly in small patients. There were also

3 groin hematomas (5.6 %), 5 temporary arterial pulse loses (9.4%), and 4 other minor assorted procedural complications (7.5 %). In 2 patients the device embolized into the descending aorta. In both patients the device was percutaneously retrieved and the defect was successfully closed using another occluder. The 2 patients with an associated membrane-type subaortic stenosis had subvalvular peak pressure gradient ranging from 15- 22 mm Hg and 6-

8 mm Hg before and after the procedure, respectively. No other patient had a significant (> 5 mmHg) device-related gradient across the left or right ventricular outflow tract. 8

Follow-up: Three months following the procedure the patient with the transient incomplete AVB developed permanent AVB (heart rate 60/min) and received an epicardial pacemaker. No other patient had dysrhythmias or conduction abnormalities on Holter monitoring. At 1-month follow-up all patients were asymptomatic with no signs of heart failure. Six-months following device closure the left-ventricular end-diastolic dimension measured by transthoracic echocardiography (M-mode) was within the normal limits for age.

At the same follow-up the echocardiographic (M-mode) stroke fraction was normal in all patients (range 35-63%). Complete occlusion of the communication at the one-year follow- up was observed in 46 / 49 patients (94 %,). (Figure 4, 5). In two and one patients with a trivial and small residual shunt respectively, after the procedure, the shunt was still present at follow-up. There was no evidence of device migration, wire fracture, thromboembolism, aortic regurgitation, endocarditis or hemolysis. No progression of preexisting or device related tricuspid regurgitation was observed during the follow-up.

Discussion

The findings of this study which includes the largest cohort of pediatric patients with 1 year follow-up indicate that transcatheter closure of PMVSDs using the Amplatzer PMVSD occluder can be effectively and safely performed if the patients are properly selected.

Complete closure of the defect was obtained in 46/49 patients (94 %, closure rate) with a low rate of significant complications during the procedure or follow – up. In addition, our experience includes occlusion of PMVSDs with large septal aneurysms, Gerbode type defects, defects associated with mild subaortic stenosis as well as PMVSDs in infants. This to our knowledge is for the first time reported.

Six patients in our study had a large associated septal aneurysm. According to our experience, in these patients a device size equal or 1-2 larger than the LV diameter of the aneurysm should be used in order to keep the left disc of the device in stable position. 9

Successful closure of the defect with no significant residual pressure gradient (10 mm Hg) across the left ventricular outflow tract was observed in 2 of the patients of this study with a PMVSD associated with mild subaortic obstruction. Disruption and/or trapping of the subaortic membrane by the left disc of the Amplatzer PMVSD occluder are the most possible explanations of the gradient reduction in these patients.

Potential major complications of transcatheter closure of PMVSDS include device embolization and interference with the aortic and atrioventricular valves leading to regurgitation. An incidence of device-related complete heart block of approximately 6 % has been recently reported. 16 It should be noted, that complete AVB following surgical repair of isolated PMVSDs occurs in  2% of patients.2 We had one instance of complete heart block in this patient population. This patient had a perimembranous VSD with posterior extension to the inlet of the right ventricle which might explain the development of advanced degree

AVB. These patients like the ones with a typical inlet defect, may be at increased risk for development of conduction abnormalities because the ventricular conduction tissue axis passes in close proximity to the anterosuperior rim of the defect and thus will be overlapped by the right ventricular retention disk. 1 As in our study this is more likely to occur in small infants who have received a relatively large device. However, longer follow – up in a larger patient population is necessary in order to establish the incidence of this serious complication.

Based on this experience we believe that all patients need a 24-hour Holter ECG recording at regular intervals. In addition, retrieval of the device is advocated in case advanced conduction problems occur immediately before or after the release of the device. This will prevent the development of permanent heart block necessitating placement of an artificial cardiac pacemaker. Complete recovery of the atrioventricular conduction system occurred following the removal of the Amplatzer PMVSD occluder in 1 of our patients who developed a 2:1

AVB immediately after the deployment of the device. Nevertheless, anatomical and device 10 related risk factors predisposing to serious conduction abnormalities following device closure of PMVSDs need further investigation.

Device-related aortic insufficiency is an infrequent complication of percutaneous closure of PMVSDs using the Amplatzer PMVSD occluder and has been reported to occur in cases with insufficient subaortic rim of tissue.10, 11 However, mild damage of the aortic valve resulting to regurgitation may be caused by repeated crossing of the aortic valve by catheters and mainly guidewires during the efforts to create an arteriovenous loop in preparation for proper device implantation.16 Tricuspid regurgitation progressed from trivial to mild in two of our patients after closure, but no device-related aortic or mitral regurgitation was observed.

Two instances of device embolization that occurred in this study group were successfully managed with catheter retrieval and occlusion of the defect with another device.

There are limitations to our study. In contrast to the reported surgical series the number of small infants in this study is limited. It should be noted, that about 25% of patients with PMVSDs present with heart failure in early infancy and will need interventional treatment during the first few months of life. Closure of PMVSDs employing the Amplatzer

PMVSD occluder seems to carry an increased risk when it is applied to small infants with suitable PMVSDs. This is mainly related to the large size of the currently available delivery system. It should be noted that most of the procedural complications in our study were observed in small infants and particularly those less 6 months of age (weight  6 Kg).

Catheter closure of PMVSDs using the Amplatzer PMVSD occluder is a new and elaborate technique and, therefore, our results might have been affected by the learning curve of each operator in the participating centers. With increased experience, major procedural complications, such as device embolization, may potentially be prevented in the future.

11

1. Anderson RH, Becker AE, Tynan M. Description of ventricular septal defects- or

how long is a piece of string? Int J Cardiol 1986; 13:267-268.

2. Mavroudis C, Baker CL, Idriss FS. Ventricular septal defect. In: Mavroudis C,

Baker CL, eds. Pediatric Cardiac Surgery. 3nd ed. Philadelphia: Mosby, 2003: 298-

320.

3. Kirklin WJ, Barrat-Boys BG. Ventricular septal defect. In: Kirklin WJ, Barrat-Boys

BG eds. Cardiac Surgery. 2nd ed. New York: Churchill Livingston, 1993:749-824.

4. Lock JE, Block PC, McKay RG, Baim DS, Keane JE. Transcatheter closure of

ventricular septal defects. Circulation 1988; 78:361-368.

5. Bridges ND, Perry SB, Keane JF, Goldstein SAN, Mandell V, Mayer JE, Jr.

Preoperative transcatheter closure of congenital muscular ventricular septal defects.

N Engl J Med 1991; 324:1312-1317.

6. Rigby ML, Redington N. Primary transcatheter closure of perimembranous

ventricular septal defect. Br Heart J 1994; 72:368-371.

7. Vogel M, Rigby ML, Shore D. Perforation of the aortic valve cusp: complication of

ventricular septal defect closure with a modified Rashkind umbrella. Pediatr

Cardiol 1996; 17:416-418.

8. Sideris EB, Walsh KP, Haddad JL, Chen CR, Ren SG, Kulkar H. Occlusion of

congenital ventricular septal defects by the buttoned device. Heart 1997; 77:276-

279.

9. Kalra GS, Verma PK, Dhall A, Singh S, Arora R. Transcatheter device closure of

perimembranous septal defects: Immediate results and intermediate-term follow-

up. Am Heart J 1999; 138:339-344. 12

10. Gu X, Han Y-M, Titus JL, Amin Z, Berry JM, Kong H, Rickers C, Umess M,Bass

JL. Transcatheter closure of membranous ventricular septal defects with Nitinol

prosthesis in a natural swine model. Cathet Cardiovasc Interv 2000; 50:502-509.

11. Hijazi Z, Hakim F, Havaleh A, Madani A, Tarawna W, Hi A, Cao QL. Catheter

closure of perimembranous ventricular septal defects using the new Amplatzer

membranous VSD occluder: Initial clinical experience. Cathet Cardiovasc Interv

2002; 56:508-515.

12. Bass JL, Karla GS, Arora R, Masura J, Gavora P, Thanopoulos BD, Torres W,

Sievert H, Carminati M, Fisher G, Ewert P. Transcatheter closure of

perimembranous VSDs with an eccentric Amplatzer device : initial human

experience. Cathet Cardiovasc Interv 2003; 58:238-245.

13. Thanopoulos BD, Tsaousis G, Karanassios E, Eleftherakis NG, Paphiytis C.

Transcatheter closure of perimembranous ventricular septal defects using the

Amplatzer asymmetric ventricular septal defect occluder. Preliminary experience in

children. Heart 2003; 89: 918-922.

14. Pedra CAC, Pedra SRF, Esteves CA, Pontes SC, Braga SLN, Arrieta SR, Santana

MVT, Fontes F, Masoura J. Percutaneous closure of perimembranous ventricular

septal defects with the Amplatzer device. Cathet Cardiovasc Interv 2004; 61: 403-

410.

15. Thanopoulos B (V) D, Laskari CV, Tsaousis GS, Vekiou A, Papadopoulos GS.

Closure of atrial septal defects with the Amplatzer occlusion device: preliminary

results. J. Am Coll Cardiol 1998; 31: 1110-1116.

16. Fu Y-C, Bass J, Amin Z, Radke W, Cheatham J, Hellebrand WE, Baltzer D, Cao Q-

L, Hijazi ZM. Transcatheter closure of perimembranous ventricular septal defects 13 using the new Amplatzer membranous VSD occluder. J Am Coll Cardiol 2006; 47:

319-325 14

LEGENDS TO FIGURES

Figure 1. (A) Long axial oblique left ventriculogram in LAO projection demonstrating a type

II PMVSD (arrow). (B) Long axial oblique left ventriculogram after release of the Amplatzer

PMVSD occluder revealing complete closure of the defect with good device position. (C)

Ascending aortogram in the long axial oblique projection after release of the device showing absence of aortic regurgitation.

Figure 2. Long axial oblique left ventriculogram (A) obtained demonstrating a PMVSD associated with a large ventricular septal aneurysm (arrow). Long Axial oblique left ventriculogram (B) after release of the Amplatzer PMVSD occluder revealing complete closure of the defect with good device position. (C) Ascending aortogram in the long axial after release of the occluder showing absence of aortic regurgitation.

Figure 3. (A) (B) Oblique left ventriculograms demonstrating a Gerbode-like PMVSD

(arrow). (C) (D) Oblique left ventriculograms after release of the Amplatzer PMVSD occluder revealing complete closure of the defect with good device position.

Figure 4. Transthoracic two-dimensional and color Doppler images obtained from five- chamber view in systole (A) and diastole (B) 3 months after implantation of the Amplatzer prosthesis. Note complete closure and good device position (arrows) with no evidence of aortic regurgitation. LV = left ventricle; RV = right ventricle; AO= ascending aorta; RA = right atrium; LA = left atrium;

Figure 5. Ascending aortogram (A) and long axial oblique left ventriculogram (B) in the

LAO projection obtained one year after implantation of the Amplatzer PMVSD occluder.

Note complete closure of the defect, good device position, and no evidence of aortic regurgitation. The platinum marker of the left inferior rim of the device is located at the 6:30 o’ clock position (arrow) indicating appropriate orientation of the left ventricular disk.