OUP user menu

Closure of atrial septal defects: is the debate over?

L Swan, M.A Gatzoulis
DOI: http://dx.doi.org/10.1016/S0195-668X(02)00527-4 130-132 First published online: 2 January 2003

See doi:10.1016/S1095-668X(02)00383-4for the article to which this editorial refers.

Debate regarding the optimal management of the patient with an atrial septal defect (ASD) has occupied many pages of the cardiology journals despite it being one of the most common and‘simple’ congenital lesions.1 Three questionscrystallize the debate: (1) Who should have their ASD closed? (2) When should it be closed? and (3) How should it be closed?

Before discussing these issues we should clarify what is meant by the term ASD. In the context of this discussion the majority are isolated secundum defects without significant fixed pulmonary hypertension. Sinus venosus defects can also be included with a few provisos, but primum defects anddefects associated with complex congenital heart disease form a different entity and will not be discussed further.

The atrial septal defect is a progressive lesion. There are few sizeable natural history studies from an era of modern medical therapy, but historical cohorts suggest long-term sequelae of right heartfailure, functional decline, arrhythmia, possible stroke and early death.2 To justify intervention, of any description, therapy must be shown to ameliorate these sequelae with a favourable risk:benefit profile.

In this issue, Roos-Hesselink et al. report on the excellent outcome associated with surgical repair of atrial septal defects during childhood.3 Over 21–33 years of follow-up there were no episodes of heart failure, stroke, pulmonary hypertensionor cardiovascular death. There were minorongoing concerns regarding post-repair arrhythmias and right atrial dilatation, but these were notassociated with significant morbidity.

Young patients referred for surgical closure, in the context of an otherwise structurally normal heart, can expect it to be performed with no mortality and minimal morbidity. There is now a cast-iron case for closing sizeable ASDs in childhood. It should be reiterated that the discussion refers to sizeable ASDs (the mean pulmonary to systemic flow ratio (Qp:Qs) was 2.3:1 in Roos-Hesselink's cohort). The rationale behind a cut-off Qp:Qs is slightly unclear but a shunt associated with right heart dilatation is a useful clinical guide.4

The issue of closure is not so clearly delineated in adults, although recent outcome studies help clarify some of the uncertainties. Attie et al.described a randomized trial of surgical vs medical therapy in patients older than 40 years (mean age 50.8 years) with a Qp:Qs ratio of >1.7 and a pulmonary artery systolic pressure of less than70 mmHg.5 This intention-to-treat study of over 500 subjects demonstrated that surgical closure was associated with a reduction in the composite end-point of major cardiovascular events/overall mortality. However, closer scrutiny shows most of this reduction was due to a fall in the incidence of recurrent pneumonia. There were fewer sudden deaths in the repair group but this was non-significant and had a wide confidence interval(Hazard ratio of medical therapy 4 (0.83–19.3), P=0.08).

Attie's study demonstrates the difficulties in performing quality randomized controlled studies in the congenital field. Even a study of this size, an anchievement for congenital heart disease, suffers from under-powering of individual end-points and in particular mortality. It is tempting to report composite end-points but these can be misleading. However, it is unlikely that a similar study will ever be repeated and decision making will need to be made on the data available. Previous non-randomized studies have suggested a survivalbenefit in older cohorts but again this is notrandomized, controlled data.6,7

The issue of whether ASD closure ameliorates a tendency to arrhythmia deserves specific mention. Conventional wisdom has stated that repair does little to affect the history of atrial dysrhythmia in the adult and this has been cited as one of the ‘failings’ of a surgical approach.6 Roos-Hesselink's group show that with early closure the incidence of atrial arrhythmia is only a fraction of that seen in the unrepaired lesion. After 15 years only 6% had symptomatic supraventricular tachycardias and in the following decade this number increased by only 2%. A Holter positive for multifocal prematureventricular complexes, or non-sustained VT, at15 years did not predict symptomatic ventricular arrhythmia or sudden death over the next 10 years.

When defects are closed in adults the majority of those with pre-operative arrhythmias do not revert to sinus rhythm and new atrial flutter/fibrillation develops in 8% of those over 40 (follow-up3.8 years).8 On an annual basis this is six times more frequent than in childhood repair. The single most important factor predicting post-repair atrial rhythm disturbance was age at time of repair. It is presumed that longstanding adverse haemodynamics with volume overload and atrial and perhaps pulmonary venous wall stretch lead to slow, inhomogeneous atrial activation and refractoriness which in turn causes reentry tachycardias. Although remodelling of the right heart and dilated pulmonary veins occurs post-repair,9 the structure of the right atrial wall does not return to normal—a feature Roos-Hesselink reports even in children.

The final issue of how defects should be closed is unresolved. The majority of isolated secundumdefects are anatomically suitable for deviceclosure. Device closure is a safe and effective method of closing septal defects and is associated with reduced morbidity, shorter hospital stay and reduced costs.10,11 On the downside, long-term outlook is unknown. It is hoped that device closure with diminished atrial scarring will reduce the incidence of post-procedural rhythm disturbance,12 but this is only speculative. At present there are no sizeable outcome studies with follow-up of much more than 24 months. This is especially true of the new generation of devices, including the Amplatzer Septal Occluder13 and we will have to wait20–30 years if we wish to have data comparable with Roos-Hesselink's study! At least until thereis some interim data most clinicians adopt apragmatic approach presenting what is known, and unknown, to the patient and allowing him/her to guide the next step.

Intervention, be that surgical or device closure, should aim towards timely repair of the defect—restoring normal haemodynamics as soon as possible and attempting to reestablishing or maintaining sinus rhythm. Early closure is the key to prevention to late arrhythmia and defects should be closed as early after detection and quantification as possible. For adults with pre-operative rhythm disturbance, consideration should be given to targeted arrhythmia intervention either viatranscatheter techniques, with new mapping and ablative systems or surgical atrial procedures.14,15

Closure of ASD, this ‘simple’ congenital heart lesion, represents an evolving field, but we are hopefully able to lay a few of these fundamental questions to rest. The remaining unresolved issues should not detract from the implementation of what we already know. Yes, close it and yes, do it soon.