Perimembranous Ventricular Septal Defect Closure with the Amplatzer Device

Rashkind1 performed catheter closure of ventricular septal defects (VSDs) in animal models in the early 1970s; he initially used hooked, single-disc and double-disc Rashkind devices. Subsequently, other investigators2–6 attempted transcatheter closure of VSDs using Rashkind’s double-umbrella or clamshell devices. The clamshell device was also used to close muscular VSDs in complex congenital heart defects as a part of overall patient management.7 Sideris and his associates8,9 utilized the buttoned device, Marshall and Perry10 used the CardioSEAL® and STARFlex® devices (NMT Medical, Inc., Boston, Massachusetts), while Le and his colleagues11 applied the Nit-Occlud® device (pfm - Produkte für die Medizin AG, Köln, Germany) to close VSDs. Thanopoulos, Tofeig and associates12,13 employed the Amplatzer® muscular VSD occluder device (AGA Medical, Golden Valley, Minnesota) to close muscular VSDs. Device closure of post myocardial infarction VSDs using the clamshell, CardioSEAL and STARFlex14 and Amplatzer muscular VSD15 devices has also been undertaken. 

Most of the above-described double-disc devices require septal rims and are useful in closing muscular defects and perimembranous VSDs with a good-sized aortic rim, and may not be useful in the more common perimembranous VSDs because of proximity to the aortic valve. Even the Amplatzer ventricular septal occluder, designed to close muscular VSDs, may not be useful for closing perimembranous VSDs because the left ventricular disc may interfere with aortic valve function. To circumvent this problem, the device was redesigned16 such that the aortic end of the left ventricular disc is short (0.5 mm), while the opposite end is longer (5.5 mm). A platinum marker to indicate the lower pole of the left ventricular disc is built into the system and should be appropriately positioned during device delivery and implantation. This modified device was used in children with smallto medium-sized VSDs17–24 and the results were generally good. Apart from the usual complications associated with complex procedures, complete heart block18,20–27 following perimembranous VSD closure with the Amplatzer has been reported, raising concerns28 regarding the use of this device. 

In this issue of the Journal, Kramoh et al29 systematically address this concern. These authors performed electrophysiological (EP) studies in 19 patients aged 8.9 ± 4.5 years who have undergone transcatheter closure of perimembranous VSDs with the Amplatzer perimembranous VSD device. The EP studies were performed prior to and immediately following VSD closure. They detected prolongation (defined as an increase in conduction interval > 15%) of conduction parameters in 95% patients and a significant (p < 0.007) increase in the His-ventricle interval. Anterograde Wenchebach periodicity appears to predict prolongation of the above-stated intervals. Conduction abnormalities on surface electrocardiography were noted in 31% of patients following VSD closure. Of all the parameters they examined, large-sized Amplatzer device was the only predictive factor for the development of new conduction abnormalities. Surprisingly, however, the authors concluded that the observed conduction abnormalities had no clinical impact. 

The authors ought to be congratulated for embarking on such detailed studies to elucidate the problem of development of heart block following Amplatzer device closure of perimembranous VSDs. However, they studied only 19 patients from the cohort of 67 patients, with selection criteria being the availability of an electrophysiologist, thus making the study group not necessarily representative of the total population intended to be investigated. Furthermore, the His-ventricle interval, which is one of the most critical EP parameters, was recorded in only 5 patients. Despite these limitations, the study is of value in drawing attention to the conduction abnormalities associated with device closure of VSDs. 

With most double-disc devices, the central portion (connecting waist) of the device goes through the VSD and does not necessarily stretch the defect. The occluding mechanism is stop-flow by the discs on either side of the VSD. In contradistinction, the Amplatzer perimembranous VSD occluder actually “stents” the defect. The conduction system traverses at the rims of the VSD, and because of the anatomic proximity, the device is likely to exert pressure on the conduction system. This is presumed to be the cause for the development of heart block. Complete heart block has been noted in 1.1–6.7% patients20–24,28 undergoing Amplatzer perimembranous VSD closure, and pacemaker implantation became necessary21–24,28 in 2–5.4% patients. Consequently, one must question whether routine transcatheter closure of perimembranous VSDs utilizing the Amplatzer is advisable. Redesigning the device to make its edges supple or soft so that less or no pressure is exerted on the conduction system may be worthwhile. If such device modification is feasible, clinical trials to determine whether the redesigned device reduces or abolishes heart block and EP abnormalities should be undertaken. 

The availability of less invasive (than surgery) methods of treating cardiac conditions should not “relax” the indications for treatment.30-34 If defect needs to be closed, the indications for closure should be the same as those used for surgical closure. The standard recommendation for surgical closure of VSDs is a pulmonary-to-systemic flow ratio (Qp:Qs) > 2:1. If one examines Table 2 of this paper,29 only 2 of 19 patients had a Qp:Qs > 2:1. Should the other 17 defects be closed, given the risk for development of complete heart block? My answer is no. Furthermore, the natural history studies have clearly demonstrated that VSDs undergo spontaneous closure, and such spontaneous closure continues to take place during childhood, adolescence and adulthood. 

In conclusion, while the study by Kramoh and associates is a useful addition to the literature, bringing attention to the development of conduction abnormalities following Amplatzer device closure of perimembranous VSDs, the fundamental question, whether small defects should be closed, given the incidence of heart block, should be carefully addressed by the pediatric cardiology community.