Transcatheter Closure of Secundum Atrial Septal Defects with Complex Anatomy (Part II)

Continued from previous page Transcatheter closure was successful in the remaining 35 patients (87.5%). The calculated Qp/Qs was 2.5 ± 0.8. All patients had normal to mildly elevated pulmonary artery pressures. The fluoroscopic and total procedure times ranged from 3–63 minutes (median time, 20 minutes) and 20–210 minutes (median time, 90 minutes), respectively. All patients with a single, large defect (n = 19) received an Amplatzer device, ranging in size from 26–38 mm (median, 32 mm; mean, 32.5 ± 3.5 mm). All patients with 2 close defects (n = 6) received a single device; five patients received ASOs ranging in size from 24–34 mm (median, 28 mm; mean, 28.4 ± 3.8 mm) and 1 patient received a Helex device (35 mm). One patient with 2 distant defects (8 mm apart) received 2 Amplatzer devices (9 and 22 mm) and had a patent ductus arteriosus closed with a Gianturco coil at the same catheterization session. The other patient with 2 distant defects (14 mm apart) received 2 Helex devices (20 and 30 mm). All patients with fenestrated atrial septums (n = 5) received ASOs ranging in size from 14–24 mm (median, 15 mm). This was before the Helex device was introduced for marketing in our country. Amplatzer devices (26 and 28 mm) were used in 2 patients with small to moderate defects associated with an aneurysm of the interatrial septum, and a 35 mm Helex device was used in another patient after a 25 mm device was removed due to a 4–5 mm shunt. In a patient with a 6 mm rim toward the superior vena cava, the right disc of a 28 mm ASO displayed some protrusion toward the superior vena cava, without resulting in local flow disturbance or pressure gradient. Immediate complete closure was achieved in 22/35 patients (63%). Seven patients had trivial to small shunts, five had moderate shunts and 1 had a large shunt. Complications. One patient with a 28 mm ASD (stretched diameter) associated with a deficient posterior rim received a 30 mm Amplatzer device. Soon after its release, TEE showed that the device, although stable, was positioned in an oblique fashion across the atrial septum, with the superior portion of the left atrial disc prolapsing into the right atrium at the superior vena cava level. The centrally located screw of the device was snared via the femoral vein and the device was recaptured into a 14 French sheath with no complications. A 32 mm ASO was subsequently and successfully implanted in the usual fashion. One patient had signs and symptoms of acute arterial occlusion in the left leg despite the use of the right femoral vein for the procedure. Retrospective inspection of the equipment used during the procedure revealed that the Meditech sizing balloon was ruptured and a fragment was missing. Surgical retrieval of the embolized fragment was carried out with no complications. Two patients had transient episodes of supraventricular tachycardia during device deployment, controlled with adenosine administration. One patient (ASD, 20 mm; stretched diameter, 28 mm; ASO device, 32 mm) had transient junctional rhythm with good ventricular response after the procedure. After a 3-day course of corticosteroids, sinus rhythm was restored. Follow-up results. TTE before discharge revealed that 26/35 patients (74%) had complete closure. Interventricular septal motion normalized in 31/35 patients (89%; p p 1,2 As imaging modalities evolved with the advent of high-quality TEE,^19,20 computer software for 3-D reconstruction^21,22 and the recently introduced intracardiac echocardiography (ICE),^23,24 a better understanding of the anatomy of the atrial septum was achieved. In addition, new device designs and sizes became available, which enabled the physicians to extend the eligibility criteria. This study further confirms the feasibility, safety and efficacy of transcatheter closure of secundum ASDs with complex anatomies, especially those that are large, with deficient anterior rims and multiple defects. However, at least in our hands, there are still some anatomical features that limit device placement. This study confirms the good outcomes of transcatheter closure of large secundum ASDs associated with deficient anterior rims using the Amplatzer device. Because of its inherent design, the ASO embraces the posterior wall of the aorta in the setting of no anterior rim. Stable device position is achieved by the stenting mechanism given by the connecting waist in the defect. However, aortic wall erosion can result from this device position.²⁵ We agree with Du et al.²⁶ that as long as the defect has sufficient rims around 75% of its margins, device implantation is feasible, stable position is achieved and elimination of shunting is likely. Although in this report we had a single case of deficient inferoposterior rim, the same concept can be applied to these circumstances, as shown by the same author. Implantation of the ASO in patients with large ASDs associated with a deficient anterior rim can be challenging and technically demanding.^26–28 We believe that opening the left atrial disc at the mouth of the left or right upper pulmonary vein, hand-shaping the delivery cable or using a long sheath with a tight curve can be helpful to overcome the problem of mal-alignment between the left atrial disc and the atrial septum. Also, slight oversizing of the device (2–4 mm larger than the stretched diameter) may be helpful to better anchor the device in the atrial septum and not miss the anterior aortic mould.^26–28 Longer, though not prohibitive, fluoroscopic and procedural times may be required in these challenging cases. On the other hand, this experience suggests that large ASDs associated with a deficient anterior rim and a floppy, thin and hypermobile posterior rim are not good candidates for transcatheter closure using the Amplatzer device. In this scenario, although there is enough rim around 75% of the defect, the floppy posterior rim does not offer enough support for adequate stabilization of the ASO. Because our failures to implant the device in such cases occurred at the beginning of our experience, it is possible that the application of the technical modifications described above at the very start of the procedure may have enabled satisfactory device positioning. This, in turn, should limit device manipulation in the left atrium and reduce the likelihood of tearing the thin posterior septum. However, failure to implant the device in the setting of large defects and a floppy septum has also been described in other series.^27,28 Some have suggested that other anatomic features may also limit transcatheter closure of secundum ASDs, including those near (29 In this regard, we believe that a borderline (6–7 mm) superior rim near the superior vena cava is not per se a contraindication for transcatheter closure. In this setting, the device should be implanted using the right upper pulmonary vein approach and as long as there is no major flow abnormality in the superior vena cava and in the right upper pulmonary vein, device release can be performed despite slight protrusion toward the superior vena cava. Reduction of the device profile with time will result in a satisfactory cosmetic result, as seen in one of our patients. This study has also shown that transcatheter closure of fenestrated septums and multiple ASDs is feasible using some technical variations. We were able to close ASDs associated with a fenestrated septum using a single standard ASO device implanted in the most centrally located hole. The stenting mechanism within the central defect helps to “smash” the adjacent holes, providing effective occlusion. Some may argue that if the central defect is too small, it may limit the size of the ASO to be used. Consequently, the adjacent holes may not be effectively covered. Although we believe that this scenario is extremely uncommon, performing a Rashkind septostomy to implant a larger device can be an option in such cases.³⁰ Using a “patch” type of device, such as the CardioSEAL or Helex device, or the new modified Amplatzer ASD device (Cribiform), may be a more reasonable alternative in this setting; however, the standard ASO worked well and closed all such defects in this series. When there are 2 separate ASDs, the distance between them will dictate the management strategy. Because the standard ASO has a self-centering mechanism and the left atrial disc is 12–16 mm larger than the connecting waist, close defects ( 8 mm, the use of 2 separate ASO devices becomes mandatory. In contrast, the use of a single device without a self-centering mechanism, such as Helex or CardioSEAL, may provide coverage for a distant defect in some cases. However, too distant defects do require 2 different devices, as seen in 1 patient in this series. It is important to note that multiple defects may be diagnosed only after balloon occlusion of the larger hole. Leaving a too distant defect of 1–2 mm of diameter uncovered may also be acceptable from the clinical efficacy viewpoint in some cases. Although we did not achieve complete closure in both patients with 2 distant defects in this experience, better rates have been seen in larger series of such patient.³¹ Although the standard ASO has been used to close ASDs associated with aneurysmal and hypermobile septum (as seen in 7 cases in this series), such anatomy is probably more suitable for devices that do not rely on a stenting mechanism within the defect to achieve stabilization in the septum. This may be particularly true in small to moderate defects. The relatively larger area of each disc on each side of the septum provides septal stability in “patch” type devices, such as Helex or CardioSEAL, which are also lighter and have a lower profile than the standard ASO. This was the rationale to use a Helex device in 2 subsequent patients with small ASDs associated with an aneurysmal septum in this series. However, large defects associated with an aneurysm of the interatrial septum (as seen in 2 patients here) may well require a large ASO device or may not even be amenable to transcatheter closure. Failure to implant a Helex device in a patient in this series might have been related to our initial learning curve or even to a technical problem with the locking mechanism of the device rather than the underlying anatomy. Further developments by the company have been made to correct this problem in the latest version of the Helex device. Unfortunately, device removal resulted in a septal tear, which precluded implantation of a new device. In this regard, the finding of tears in the floppy portion of the atrial septum in 3 patients in this series is worrisome and suggests that extreme care should be taken during manipulation of devices and delivery systems in patients with this anatomical feature. The rate of complete closure in this series was 89%, which is slightly lower than the classically reported 94–96% overall occlusion rate observed with the use of the ASO and Helex devices. Lower overall closure rates have also been reported for patients with large defects.²⁷ In addition, patients with deficient rims may have lower rates of complete closure compared to those with sufficient rims, although this was not statistically significant in a previous published report.²⁶ More importantly, the procedure corrects the hemodynamic burden to the right ventricle, reflected by the significant and progressive reduction of the right ventricular end-diastolic diameter associated with normalization of the septal motion in the whole population in this series. Moreover, right ventricular dimensions returned to normal in all patients with residual shunting, providing clinical cure even in a patient with a 4 mm residual shunt. The risk of endocarditis and paradoxical embolization in patients with residual shunts after device closure remains unknown. Interestingly, the presence of a residual shunt in the patient with impaired left ventricular function, although not deliberate, may have been helpful to decompress the left ventricle. It has even been suggested that fenestrated devices should be used in this clinical scenario.³² In conclusion, this report shows the feasibility, safety and efficacy of transcatheter closure of large secundum ASDs associated with a deficient anterior rim and a firm posterior rim, multiple ASDs, ASDs with fenestrated septum, and ASDs associated with an aneurysmal septum. In our hands, patients with large ASDs associated with a deficient anterior rim (Disclosure. This paper was presented, in part, at the XIV Brazilian Congress of Echocardiography and received an award for Best Abstract in Congenital Heart Disease. Addendum. After this paper was accepted for publication, we performed transcatheter closure of complex ASDs in 8 additional patients. One had 2 distant holes, requiring 2 ASOs, with immediate occlusion and no complications. Five had large defects, four associated with a deficient anterior rim and 1 with a deficient posterior rim. All had ASOs implanted with immediate occlusion and no complications. One patient had 2 defects close to each other associated with an aneurysmal septum. A single ASO was implanted; however, a 2 mm immediate residual shunt was smaller hole was located. The remaining patient had a moderate ASD associated with an aneurysm of the interatrial septum, completely closed with a Helex device. This further experience corroborates our previous observations.