Routine Coronary Angiography in Adults Undergoing Percutaneous Atrial Septal Defect Closure: Implications for Practice Guidelines

Objectives. Secundum atrial septal defect (ASD) is a common adult congenital heart lesion for which percutaneous closure offers treatment in patients with suitable anatomy. We sought to determine the outcomes of coronary angiography in a population of adults >40 years of age who were undergoing percutaneous ASD closure. Methods. Patients >40 years of age who underwent ASD closure between 2009-2016 were included in this retrospective chart review. Coronary angiograms were reviewed by 2 independent reviewers to evaluate the presence and degree of coronary artery disease (CAD) and the resulting clinical sequelae. Results. A total of 398 patients underwent ASD closure, and 300 had coronary angiography at the time of closure. Mild CAD (10%-39% vessel stenosis) was found in 33 patients (11.0%), moderate CAD (40%-69% vessel stenosis) in 23 patients (7.7%), and severe CAD (≥70% vessel stenosis) in 25 patients (8.3%). Of the 48 patients with moderate to severe CAD, 24 had proximal vessel disease and 21 had multivessel disease. Four patients (8.3%) with moderate to severe CAD subsequently underwent percutaneous interventions, 16 patients (33.3%) had medication changes, 16 patients (33.3%) had perfusion testing followed by medication changes, and 12 patients (25%) had no changes in their medical management. Conclusions. Given the prevalence of CAD in this population, routine coronary angiography at the time of ASD closure should be reserved for patients with an unfavorable cardiovascular risk profile, who have a higher likelihood of CAD. While patients with ASDs suffer from chest pain and dyspnea both before and after percutaneous closure, few have established CAD.

J INVASIVE CARDIOL 2021;33(10):E777-E783. Epub 2021 September 23.

Key words: congenital heart disease, coronary artery disease

An atrial septal defect (ASD) is one of the most common congenital heart lesions, with the most common subtype being a secundum defect, which is generally amenable to percutaneous device closure.¹ In many cases, ASDs go undetected until adulthood.² When an ASD results in right-sided chamber enlargement, closure is indicated. It is well established that percutaneous closure offers results comparable to open surgical repair, is cost effective, less invasive, and associated with fewer complications, and requires a shorter hospital stay.^3-7 As a result, percutaneous closure has become the standard of care in adults with secundum ASDs.⁸

Currently, there are no clinical guidelines to suggest whether this population should also be investigated with coronary angiography at the time of closure. Indeed, little is known about the prevalence of coronary artery disease (CAD) in the adult ASD population and how that diagnosis might impact subsequent management. In 1 long-term follow-up study, nearly 20% of patients experienced chest pain after ASD closure.⁹ In another study, the presence of chest pain and dyspnea in patients with congenital heart disease (CHD) prior to any intervention was reported to be close to 20% and 40%, respectively.¹⁰ Given the overlap between these symptoms and CAD, it is often difficult to distinguish between the 2 conditions clinically. Therefore, defining coronary anatomy at the time of device closure may expedite and simplify the identification of a source of symptoms not otherwise identified in patients already committed to a procedure.

In 2009, after an adverse patient outcome following percutaneous ASD closure where coronary anatomy was unknown, our center adopted a practice of routine angiography at the time of device closure in patients over the age of 40 years. Herein, we report our 8-year experience of performing coronary angiography in adults >40 years old undergoing elective percutaneous ASD closure, assessing prevalence and severity of CAD in this population and its impact on management decisions.

Study design and participants. We established a retrospective registry including 1518 patients who underwent percutaneous ASD closure from 1999 to 2016 at the Peter Munk Cardiac Centre in the Toronto General Hospital. For this study, we selected consecutive patients >40 years old who underwent transcatheter ASD closure between March 2009 and November 2016 and were investigated by coronary angiography at the time of closure. A retrospective chart review was undertaken for all patients in the sample using structured data-extraction forms. Baseline demographics were collected, including age, gender, body mass index, and the presence of cardiovascular risk factors, including hypertension, diabetes, dyslipidemia, family history of early CAD, smoking history, and previous history of coronary revascularization (percutaneous coronary intervention [PCI] or coronary artery bypass graft surgery). Data were also collected on symptoms before and after ASD closure, with a focus on chest pain, dyspnea, palpitations, headache/migraine, and syncope. Symptoms were evaluated during clinical assessments prior to ASD closure and at the first follow-up visit, usually between 2-6 months after closure. Patients and the public were not involved in the design, conduct, reporting, or dissemination plans of our research. The study protocol was approved by the institutional research ethics board of the University Health Network.

Coronary angiography. Available coronary angiograms were reviewed by two experienced interventional cardiologists (AM, HS) using syngo Dynamics imaging software (Siemens Healthineers). Reviewers were blinded to the patients’ clinical diagnoses. Presence and location of CAD were described using the Synergy between Percutaneous Coronary Intervention with Taxus and Coronary Artery Bypass Surgery (SYNTAX) scoring system.¹¹ In patients with evidence of coronary stenosis, lesions were classified by the degree of artery occlusion. Occlusion was defined as luminal irregularities (<10%), mild (10%-39%), moderate (40%-69%), or severe (≥70%). Whenever the 2 observers were in disagreement, a third interventional cardiologist (EH) was asked to review and evaluate the degree of CAD. Moderate and severe CAD was considered proximal in patients who had lesions in the upper right coronary artery (SYNTAX segment 1), the left main artery (SYNTAX segment 5) or either of its proximal branches, ie, the left anterior descending artery (SYNTAX segment 6) or the circumflex artery (SYNTAX segment).¹¹Multivessel disease was considered present if at least 2 of the major epicardial arteries had evidence of moderate to severe CAD.

Statistical analysis. Statistical analyses were performed using SPSS software, version 20 (SPSS). Continuous variables were expressed as mean ± standard deviation and compared using one-way analysis of variance (ANOVA). Categorical variables were reported as counts and percentages and were compared using a Chi-square test. To determine which patients would benefit most from coronary angiography at the time of ASD closure, we conducted an exploratory, multivariable, logistic-regression analysis. P-values <.05 were considered statistically significant. For analysis purposes, we grouped patients with mild CAD or luminal irregularities into 1 group and patients with moderate or severe CAD into another group. The interobserver agreement of CAD severity classification by 2 reviewers was assessed using kappa statistics.^12,13

Of the 398 patients >40 years old who underwent percutaneous ASD closure between March 2009 and November 2016, 300 were investigated by coronary angiography and included in the final analysis (Figure 1). We excluded 33 patients because they did not have coronary angiography performed at the time of ASD closure. Overall, these patients were younger, primarily female, and had a lower prevalence of cardiovascular risk factors compared with the other patients. This was also due to non-specific operator practice variability in regard to investigating patients with coronary angiography at the time of closure. Patients were also excluded from the analysis if their coronary anatomy had been defined within the last 12 months prior to closure (n = 31) or they had pre-existing known CAD (n = 34). These patients were significantly older and had a higher proportion of cardiovascular risk factors; specifically, hypertension, diabetes, and dyslipidemia compared with those who had unknown CAD status and underwent coronary angiography at the time of closure (Table 1). In all patients presenting for ASD closure (n = 398), mild, moderate, and severe CAD was reported in 55 (13.8%), 30 (7.5%), and 46 patients (11.6%), respectively. The prevalence of moderate to severe CAD among all comers was 19.1%.

Interobserver variability was assessed between the 2 independent cardiologists performing coronary angiographic assessments. Cohen’s kappa coefficient showed excellent agreement between their assessments (k = 0.912; P<.001). There were 12 instances (4.0%) where the 2 observers were in disagreement and a third cardiologist was required to evaluate the degree of CAD. The majority of these disagreements (8/12) occurred when 1 evaluator noted the absence of CAD and the other reported mild CAD. The other 4 instances involved disagreements between mild and moderate CAD. Based on the final classification, in patients with unknown CAD status at the time of ASD closure, mild, moderate, and severe CAD was reported in 40 (13.3%), 23 (7.7%), and 25 patients (8.3%), respectively, whereas 212 patients (70.7%) had angiographically normal coronary arteries.

Patient demographics, baseline cardiovascular risk factors, preprocedural symptoms, and postprocedural symptoms reported during the first follow-up visit are described in Table 2. The CAD group was significantly older, with a higher proportion of men and a higher prevalence of cardiovascular risk factors, specifically, hypertension, dyslipidemia, and smoking (P<.05). Among adults between 40-59 years of age, the prevalence of CAD was 8.6% (16/186); in those 60 years of age and older, the prevalence was 28.1% (32/114). Symptoms such as chest pain, dyspnea, and palpitations reported prior to ASD closure or after closure were not significantly different between the CAD and no-CAD groups. In total, prior to ASD closure, 124 patients (41.3%) complained of chest pain and 162 reported dyspnea (54.0%). After ASD closure, 38 patients (12.7%) complained of chest pain at their first follow-up visit, which occurred at 3.26 ± 1.70 months post closure. Of these patients, 25 (8.3%) had reported chest pain prior to closure and 13 (4.3%) had new-onset pain. Meanwhile, 30 patients (10.0%) reported dyspnea at their first follow-up visit, which occurred at 3.97 ± 2.11 months post closure; 26 of these patients (8.7%) had also reported dyspnea prior to closure.

A further analysis of coronary angiograms was performed in the 48 patients (16.0%) with moderate to severe CAD to describe the lesion as proximal/distal or single/multivessel involvement (Table 3). We found 24 patients with proximal vessel disease and 21 with multivessel disease, with 5 patients noted to have triple-vessel disease. Of the 21 patients with multivessel disease, 2 went on to have PCI, 16 had escalation of pharmacological therapy, and 3 continued with their routine medical management. Similarly, of the 24 patients with proximal vessel disease, 2 went on to have PCI, 17 had medication changes, and 5 were maintained on the same medical regimen (Table 3 and Figure 2). The median time from ASD closure to PCI was 2 months (interquartile range [IQR], 5.5 months); in those who had medication changes, the median time from ASD closure to medication change was 2 months (IQR, 2 months).

Femoral arterial access was used to perform coronary angiography in the majority of patients, while radial arterial access was utilized in 5 patients. In general, suture-mediated closure devices were used as a default strategy for arterial closure when anatomically possible. Complications including intraprocedural, periprocedural, and postprocedural up to the first follow-up were reported in 9 patients (3.0%). These included pulmonary embolism (1/300; 0.3%), deep vein thrombosis (1/300; 0.3%), transient ischemic attack (1/300; 0.3%), and hematomas (6/300; 2.0%). The average procedural radiation dose was 7173 cGy•cm² and an average of 139 mL of iodixanol was utilized per case (n = 112), which included the ASD closure and the coronary angiogram.

An exploratory multivariable logistic-regression analysis was done to determine predictors of CAD in the ASD population. We used a backward stepwise-elimination approach to select the independent predictors of CAD (luminal irregularity and mild/moderate/severe CAD vs no CAD) in this population. Testing was done on age, sex, hypertension, diabetes, dyslipidemia, family history of early CAD, and smoking history; the risk factors that were statistically significant (P<.05) and remained in the final model included age (odds ratio [OR], 1.11; 95% confidence interval [CI] 1.08-1.15), smoking history (OR, 2.88; 95% CI, 1.56-5.35), and male sex (OR, 3.77; 95% CI, 2.02-7.02). The adjusted R² statistic was 35.9% and the C-statistic was 0.824, indicating good discrimination.

This single-center, retrospective study assessed outcomes of routine coronary angiography in adults >40 years old at the time of percutaneous ASD closure. Despite no significant symptomatic differences between groups based on CAD status, nearly 1 in 6 patients were found to have moderate to severe CAD that resulted in further management changes in addition to ASD closure in 36 patients (12.0%).

In 1999, the American College of Cardiology/American Heart Association (ACC/AHA) developed guidelines outlining the use of coronary angiography in adults with CHD prior to surgical treatment,¹⁴ particularly if symptomatic or in those with non-invasive evidence of CAD (class I, level of evidence C). In 2008, the ACC/AHA developed more specific guidelines for managing Ebstein’s anomaly, recommending coronary angiography when surgical repair was planned in adults >35 years of age who had coronary risk factors (class IIa, level of evidence B).¹⁵ Currently, there are no guidelines for coronary angiography at the time of percutaneous ASD closure in adults.

Adult patients with CHD are also at risk of developing CAD based upon their cardiovascular risk profile.^16,17 Since many patients are diagnosed with an ASD later in life, they may have coexisting CAD that can influence management strategies.¹⁸ This is particularly relevant since a small proportion of patients report new-onset symptoms after ASD closure, and defining their coronary arteries before closure may help assist in subsequent management. The prevalence of CAD increases with age; for patients between 40-59 years old, the prevalence is reported to be 6.3% for men and 5.6% for women, which increases to 19.9% in men and 9.7% in women 60-79 years old.¹⁹ We found comparable rates of CAD among patients undergoing percutaneous ASD closure.

Given the high prevalence of symptoms pre closure, performing coronary angiography as an adjunct to percutaneous ASD closure was thought to have merit. Examining the findings of coronary angiography from this cohort provides data to determine whether angiography should be performed during ASD closure. Coronary angiography in selected patients at higher likelihood of having CAD at the time of device closure facilitates symptom interpretation and identifies those who might benefit from revascularization or optimization of medical management. In addition, the absence of CAD in a symptomatic adult with an ASD provides information to direct further symptom investigations, if not thought to be due to the atrial-level shunt. In patients with intermediate pretest probability of having CAD, non-invasive computed tomography angiography could be a suitable alternative to coronary angiography in detecting patients with CAD prior to ASD closure.²⁰

Routine coronary angiography posed a limited additional risk during the procedure. We noted a complication rate of 3.0% in our cohort, with thromboembolic events and neurologic complications occurring in 1.0% of the cohort. It is speculated that the right-sided complications were related to the ASD closure itself and not the arterial puncture. The Manufacturer and User Facility Device Experience database reports thromboembolic and neurological complication rates of 0.05% and 0.03%, respectively, among patients undergoing percutaneous ASD closure using an Amplatzer atrial septal occluder (ASO) device (AGA Medical Corporation). These rates include both adults and children undergoing ASD closure with the Amplatzer ASO device and therefore likely underestimate the complication rate in adult-specific populations.²¹ More recently, coronary angiography has been routinely performed through radial arterial access, which can potentially further reduce the associated access-related complications.²²

Many of the earlier studies evaluating the prevalence of CAD in patients with adult CHD were based on autopsy findings.^23,24 These studies produced conflicting results that must be interpreted with caution. The invasive nature of coronary angiography makes it unattractive as a screening tool to evaluate CAD in the general adult CHD population. Giannakoulas et al examined adult patients who underwent catheterization to evaluate their CHD with selective coronary angiography from May 1999 to February 2006 at the Royal Brompton Hospital in London.¹⁰ They excluded patients referred for angiography for suspicion of CAD to avoid overestimating the general prevalence of CAD in the adult CHD population. Of the 250 patients included in the study (mean age, 51 ± 15 years; 53.0% men), 35 patients (14.0%) had angiographic evidence of atherosclerosis on visual assessment, which was clinically significant in 23 patients (9.2%). Of these patients, 14 had single-vessel disease, 5 had double-vessel disease, and 4 had triple-vessel disease. Our study is the first to evaluate the prevalence of CAD in adult ASD patients >40 years old with coronary angiography performed at the time of percutaneous closure. Prior to this study, the only other study to specifically look at the prevalence of CAD in patients undergoing ASD closure was by Cay et al in Turkey.²⁵ They assessed a group of 138 patients (mean age, 54 ± 10 years; 29% men) and found angiographically significant CAD in 12 patients (8.7%). The prevalence of moderate to severe CAD reported in our study was higher than what was found in previously published studies,^10,25 in part due to older age at the time of closure in our population.

Based on results from our exploratory analysis, we found that age, male sex, and smoking history were useful predictors in determining whether a patient had CAD. In the model, with every 1-year increase in age, starting from the age of 40 as per the inclusion criteria of this study, the risk of CAD increased by 11%. Furthermore, in the model, men had almost 4-times higher odds of having CAD than women, and smokers had almost 3-times higher odds. Additionally, coronary angiography should be considered at the time of ASD closure in patients with an established diagnosis of CAD to ensure that bypass surgery with concomitant surgical ASD closure is not more suitable.

Study limitations. The main limitations of this study are its retrospective design and the use of medical records as source documents. To minimize errors typical to medical chart abstraction, we used standardized data collection by data abstractors and conducted audits and data cleaning. Another potential limitation is that data were collected from a single high-volume institution, which may limit the generalizability of our findings. Furthermore, this study relied heavily on coronary angiographic visual assessment. This technique is inherently subjective and difficult to standardize. We attempted to achieve consistent data collection from both evaluators by creating a standardized CAD data-collection form and by having up to 3 reviewers examine angiograms where there was a difference of opinion. We wanted our analysis to mirror real-world practice and did not carry out quantitative coronary angiographic analysis.

While a definitive recommendation cannot be made based on this data set, given our experience, we feel there is little value in coronary angiography as an adjunct to ASD closure in adults without significant suggestive symptoms or risk factors. There is no evidence to suggest that patients with ASDs develop premature CAD. The prevalence of CAD in adult patients undergoing percutaneous ASD closure is similar to the general population and more common in patients with cardiovascular risk factors. Furthermore, assessing for CAD through routine coronary angiography at the time of ASD closure has potential risks. It may be reasonable to reserve routine coronary angiography as a procedural adjunct to percutaneous ASD closure to patients with unfavorable preclosure cardiovascular risk profiles, concerning symptoms, and in those patients at intermediate risk who had positive functional tests. There should be some consistency between existing guidelines for coronary angiography around elective cardiac surgery and interventional procedures that may require surgical intervention. Coronary risk factors should be assessed in patients to establish those with the greatest clinical benefit of undergoing coronary angiography at the time of percutaneous ASD closure.

From the ¹Faculty of Medicine, University of Toronto, Ontario, Canada; ²Faculty of Medicine, University of Toronto General Internal Medicine, Ontario, Canada; ³Department of Cardiology, Toronto General Hospital, Ontario, Canada; ⁴Department of Interventional Cardiology, Peter Munk Cardiac Centre, Ontario, Canada; ⁵Institute of Health Policy, Management and Evaluation (IHPME), University of Toronto, Toronto, Ontario, Canada; ⁶Toronto Health Economics and Technology Assessment (THETA) Collaborative, Toronto General Hospital Research Institute, University Health Network, Toronto, Ontario, Canada; ⁷St Boniface Hospital, University of Manitoba, Winnipeg, Manitoba, Canada; and ⁸Toronto Congenital Cardiac Centre for Adults, University Health Network, Ontario, Canada.

Funding: This work is supported by the Peter Munk Chair in Structural Heart Disease Intervention.

Disclosure: The authors have completed and returned the ICMJE Form for Disclosure of Potential

Conflicts of Interest. Dr Horlick and Dr Osten are consultants for Abbott Structural Heart. The Peter Munk Cardiac Centre receives support from Abbott for its educational mission and its administration. Dr Horlick reports research grant support from Abbott for other projects; consultant for Medtronic and Edwards Lifesciences; medical advisory board of Realview Imaging. Dr Benson reports grant support to his institution from Abbott. The remaining authors report no conflicts of interest regarding the content herein.

Manuscript accepted December 7, 2020.

Address for correspondence: Dr Eric M. Horlick, Toronto General Hospital, 200 Elizabeth St, 6E – Room 249, Toronto, Ontario, M5G 2C4. Email: Eric.Horlick@uhn.ca