Patients with isolated ASD are usually asymptomatic through infancy and early childhood and the elective surgery may be deferred to 2-4 years of age.1) However, it has also been well documented that substantial numbers of small children are in need of earlier intervention, especially when the disease is complicated by chronic lung disease in prematurely-born infants or associated with certain chromosomal anomalies.2)3) Nevertheless, the need for early treatment has been recognized in small children who are showing rapidly enlarging defects to prevent rendering a suitable ASD to an unsuitable ASD for device closure.4)5) Some authors advocate that any defect larger than 8 mm with evidence of a significant left to right shunt should be closed even in very young patients, because such a defect will likely never close spontaneously and may even get larger.6) Indeed, in a study on the change in size of ASD, a significant enlargement of the defect, more than 50% of the initial size, has been detected in 30% of the studied population, which suggests that there is a potential for some of the defects to outgrow a possible transcatheter closure.7)

Recent articles on surgical treatment of ASD in young and symptomatic infants emphasize on benefits of early surgery on the grounds of a dramatically improved clinical course.3) There has been a study reporting on increased procedural time and lower procedural success rate for transcatheter ASD closure during their early experience with the Amplatzer septal occluder (ASO, AGA medical corporation, Plymouth, MN, USA) in symptomatic small children.8) However, another recent report states that device closure is a safer and more effective alternative to surgery, with valid advantages in very small children, including infants and those who weigh less than 10 kg.9-12) Cardenas et al.10) reported that there was no additional risk in smaller babies (<10 kg) for transcatheter ASD closure, and also in our study comparing outcomes of device therapy to those of surgical closure in small children with 10 kg of weight or less, the superior safety of the device group was evident, which was proved by a lower complication rate.11)

There are several problems encountered during device closure in small children. The vessels are relatively small compared to the large delivery system, which leaves the possibility for potential damage to vascular access. Moreover, because of the rigid coupling mechanism between the device and delivery cable as well as the relatively stiff delivery system, there is a latent risk for damage to the cardiac tissue. The inherent design of ASO with the relatively-excessive rim width of atrial discs in smaller devices (Fig. 1) may also increase the potential risk for erosion or cessation of the procedure, because the excessive atrial rims may encroach on the adjacent cardiac structures or atrioventricular valves. Because of the small total septal length in small hearts, plans for implantation of large devices may occasionally be thwarted, especially in small children with relative mitral rim deficiency. Therefore, in order to ensure procedural safety, special care must be taken to prevent damage to the vasculature and to cardiac structures, and the procedure should be conducted by an experienced hand in this subset of the patient population. Also, further modifications of the design for smaller sizes of the device to fit a small heart may be needed to increase the applicability of the procedure to smaller children. On the other hand, a "stopgap technique", which is choosing the device size according to the size of the heart so the left atrial disc diameter can fit the maximal septal length has been suggested.13) This technique may be a good solution for small children with a relatively large defect diameter to total septal length ratio, but the interpretation for the usefulness of this technique still requires further investigation.

Elderly patients with hemodynamically significant ASD frequently experience complications with serious long-term adverse consequences such as pulmonary hypertension, atrial arrhythmia, and atrioventricular valve insufficiencies that essentially involve the tricuspid valve as well as right-sided heart failure secondary to chronic volume overload of the right side of the heart.14) This group of patients has a higher prevalence of co-morbid cardiovascular and/or systemic diseases including systemic hypertension, stroke, chronic lung disease, diabetes mellitus, atherosclerosis and coronary heart disease.14-16) The pathophysiologic mechanism is summarized in Fig. 2. Longstanding left to right shunt in the atrial level results in progressive right heart dilatation, significant tricuspid insufficiency and subsequent increase in right atrial pressure. Left heart may also be influenced by chronic volume underload, increased atrial pressure as well as co-morbid diseases such as systemic hypertension or coronary heart disease. Significant numbers of the patients are also affected by more advanced complications inherent to the end stage of this pathophysiologic cascade, including pulmonary hypertension, ventricular dysfunction and atrial arrhythmias which cause significant symptoms. In addition, diastolic compression of left ventricle (LV) by dilated right ventricle reduces LV end-diastolic volume in chronic setting, and this is responsible for so called "masked LV restriction" which may cause serious pulmonary edema secondary to LV dysfunction and left atrium (LA) pressure increase after ASD closure.17) Because of the chronic nature of the disease, patients are usually adapted to the disabilities they have had and invasive interventions are occasionally refused by patients. Nevertheless, scarce evidence for survival benefit of anatomical closure and higher potential risk of definitive treatment in this group of patients may sometimes affect the decision to a more conservative one.

The benefits from ASD closure in the elderly population has been documented in fewer reports compared to the younger population. In the 1990s, there was a nonrandomized study debating the clinical benefits between medical treatment only and anatomical closure.18) However, a relatively recent randomized study in a large population confirmed that anatomical closure is superior to medical treatment in preventing major events.19) Moreover, the realistic benefits of ASD closure include symptomatic relief, improvements of functional status as well as the quality of life.20)21) These favorable clinical changes are supported by the immediate and substantial reverse remodeling of the heart after closure from many studies.14)16)17)22) Excellent safety and efficacy in terms of improving functional class and favorable reverse remodeling of the heart have been consistently reported, and the solutions for major complications such as restrictive LV, chronic atrial arrhythmia and pulmonary hypertension have been discussed as major issues.

The most important problem in elderly patients is that they have more complex presentations of the disease, mainly because they frequently have co-morbid systemic and heart diseases as previously mentioned. This necessitates precise evaluation of the risk factors as an essential part of successful treatment. Co-morbid diseases in elderly patients in recent publications are shown on Table 1.14-16) More than one third of the patients showed systemic hypertension and systemic diseases such as diabetes mellitus, and a considerable extent of pulmonary and neurological diseases were also present.

Cardiac co-morbidities include pulmonary hypertension in nearly half of the patients, chronic atrial arrhythmia in more than 20% and ischemic heart disease in about 15% of the patients. Post-closure pulmonary edema caused by "masked LV restriction" may appear in 2 to 4% of the elderly patients and should be ruled out before ASD closure through a balloon occlusion test. It was previously described that most of the patients with evidence of LV restriction during temporary balloon occlusion of the defect may show favorable response to several days of anti-congestive therapy, so that they could go through with the device closure without any serious events.

Even though LV dysfunction after ASD closure is not a frequent finding, it may lead to a serious consequence once it occurs. Thus, the detection of the risk for serious LV restriction before the procedure is very important. There are 2 different protocols for detection and management of high risk patients, which have been independently developed by 2 separate groups (Fig. 3).5)14)15)17)23) Both protocols involve a temporary balloon occlusion test and measurement of LA pressure, but the definition of significant LV restriction is quite different for each protocol. Both protocols recommend several days of medical therapy including diuretics, inotropic agents and afterload reducing therapy once LV restriction is diagnosed, and if the patient proves to be refractory to medical therapy on a subsequent balloon occlusion test, they are recommended a fenestrated device (Fig. 4). These protocols were believed to be useful in the therapeutic approach to high-risk patients, but both protocols are based on empirical experiences rather than cumulated data which raises doubt on the reliability of each numerical criterion that they have suggested.

Previous studies identified several risk factors for persistent atrial arrhythmias, such as pulmonary hypertension, valvular dysfunction with atrial enlargement, presence of atrial arrhythmia before closure of ASD, and not surprisingly, all these factors are commonly linked to the most important risk factor, the age of the patient.24-26) Therefore, it is very important to close a hemodynamically significant defect early before cascade of deleterious events begin.

In elderly ASD patients with permanent atrial fibrillation, there are 2 therapeutic options which include rhythm control and rate control. Rhythm control involves surgical correction of ASD with Maze procedure at the time of operation, or device closure with catheter ablation. Rate control is achieved by medical treatment including anticoagulation and occasional antiarrhythmic drugs according to the baseline heart rate or arrhythmic event after closure of the defect. A rhythm control policy may be the preferable option in patients with symptomatic tachyarrhythmia.

A recent study on elderly patients with ASD and chronic atrial fibrillation, reported that the outcome of device closure was as good as the outcome from patients without atrial fibrillation in terms of safety, improvement of functional status and reverse remodeling of the heart.27) Their strategy was based on the previous study comparing rhythm control and rate control, which revealed that there was no difference in mortality and quality of life between these strategies (AFFIRM study).28) However, this subject may be controversial since for the time being, the predominant opinion is that rhythm control appears to be superior to rate control in the treatment of atrial fibrillation.29)