Relative Risk Factors for Cardiac Erosion Following Transcatheter Closure of Atrial Septal DefectsCLINICAL PERSPECTIVE
A Case–Control Study
Background—Transcatheter closure of secundum atrial septal defects (ASD) using the Amplatzer septal occluder is generally safe and effective, but erosion into the pericardial space or aorta has been described. Although the absolute risk of this complication is low, there has been no assessment of relative risk factors.
Methods and Results—All erosions reported to St. Jude Medical after ASD closure with an Amplatzer septal occluder (cases) were compared with controls (matched 2:1) who underwent ASD closure but did not develop an erosion. A total of 125 erosions were reported between 2002 and 2014, including 95 with an available echocardiogram. The median duration from implant to erosion was 14 days, but was >1 year in 16 patients. Nine patients (all age ≥17 years) who died were more likely to have an oversized device, and to have erosion into the aorta, than survivors. Aortic or superior vena cava rim deficiencies were more common in cases than in controls. In addition, larger balloon-sized ASD diameter, Amplatzer septal occluder device size, and device size–ASD diameter difference, and smaller weight:device size ratio were associated with erosion. On multivariable analysis, deficiency of any rim, device >5 mm larger than ASD diameter, and weight:device size ratio were associated with erosion.
Conclusions—In addition to aortic rim deficiency, which was almost universal among erosion cases, there were several relative risk factors for erosion after ASD closure with the Amplatzer septal occluder device. To understand the mechanisms of and absolute risk factors for this uncommon but serious complication, an adequately powered prospective study with thorough echocardiographic evaluation will be critical.
- amplatzer septal occluder
- cardiac tamponade
- heart diseases
- vascular closure devices
Transcatheter closure of secundum atrial septal defects (ASDs) using the Amplatzer septal occluder (ASO; St. Jude Medical [SJM], Plymouth, MN) has been demonstrated to be safe and effective relative to surgical repair.1–4 However, erosion of the device through an atrial wall into the aorta or pericardial space (cardiac erosion) is a rare but serious adverse event that can occur early or late after ASD closure.2–17 The absolute risk of erosion after ASO implant has been estimated to range from 0.043% to 0.3%.2,6,8,17–19 The specific mechanisms and risk factors for device erosion remain unclear, and are likely to be multifactorial.20,21 Given the rarity of this event, it is unlikely that there are discrete absolute risk factors that are sufficiently strong to contraindicate device closure of an ASD. However, there may be anatomic, patient-related, and procedural variables that are associated with an increased risk of erosion relative to the general population of patients undergoing ASD closure (ie, relative risk). Although multiple case reports, reviews, and analyses of the Manufacturer And User Facility Device Experience (MAUDE) database have been published,2–17 there are no studies that adequately speak to the question of relative or absolute erosion risk.
Clinical Perspective on p 1746
The ASO device was approved by the US Food and Drug Administration (FDA) for ASD closure in 2001. The following year, the first case of cardiac erosion was published5 and the first field event in the United States was reported to the manufacturer. Soon thereafter, SJM established an independent panel of physicians as an Erosion Board to assist in monitoring and analyzing these events, and ongoing evaluation led to a series of recommendations regarding device sizing and periprocedural imaging, including changes to the instructions for use (IFU) for the ASO.6 Also, to define the incidence of erosion and to understand the association between erosion and defect sizing and other implant-related factors, planning was initiated for a 1000-patient ASO Post-Approval Study (PAS), which ultimately began enrolling in 2008. In January 2012, after discussion with the FDA, SJM made additional modifications to the IFU, the most significant of which was to specify that implantation of an ASO was contraindicated in patients with echocardiographic evidence of absence or deficiency (<5 mm) of the anterior-superior (aortic, retroaortic) ASD rim in multiple and consecutive views by transesophageal or intracardiac echocardiography. That modification had the potential to alter the therapeutic landscape for ASDs dramatically, because 17% to 60% of patients undergoing device closure have been found to have a deficient anterior-superior rim.17,22,23 In addition, the recommendation to use multiple views to evaluate the aortic rim was often confusing to implanters who used intracardiac echocardiography for ASD closure, because this was not necessarily routine practice. Understandably, these IFU modifications raised a number of questions within the interventional cardiology community and led to variable changes in practice.
In response to the ongoing occurrence of adverse events such as erosion, embolization, residual leak, and wire fracture, the FDA Circulatory System Devices Panel convened an Advisory Committee meeting in May 2012 to review the safety of ASD closure devices from all manufacturers. Statements were solicited from multiple stakeholder organizations, including the American Heart Association, American College of Cardiology, Society for Cardiovascular Angiography and Interventions, and Society of Thoracic Surgeons, almost all of which recommended classifying a deficient aortic rim as a high-risk feature rather than a contraindication to ASD closure.24,25 Not long after the IFU change and the subsequent Advisory Committee meeting, in 2013, the FDA recommended removal of absent/deficient anterior-superior rim as a contraindication, but mandated SJM to perform a prospective study aimed at determining risk factors for erosion after ASO implant. The ADVANCE ASO AMPLATZER Atrial Septal Occluder Post Market Surveillance Study (ClinicalTrials.gov Identifier: NCT02353351) was initiated in June 2014 with a target enrollment of 8000 patients and estimated completion in 2019.
The goal of the current case-control study, in which all erosion cases that have been reported to or entered into the SJM database were analyzed against an age- and sex-matched control cohort, was to describe and evaluate relative risk factors for erosion.
The Erosion Board reviews all potential erosion events reported under a prespecified set of event terms, as identified through product surveillance complaints, the prospective PAS, and monitoring of the MAUDE database and the published literature. On review, the Erosion Board determines whether the event is likely to be attributable to erosion and adjudicates on certain technical, anatomic, and other aspects of the case. In 2011, SJM took the additional step of reviewing all events that had been published but not reported to the MAUDE database or the company, and contacted the authors with requests for supporting data in an effort to establish the most complete database possible before the May 2012 Advisory Committee meeting. For the present study, SJM provided data to the authors for all erosions reported and adjudicated by the Erosion Board, and all available imaging studies supplied by implanters reporting the events, as well. Available details of erosion cases varied depending on the data submitted by the reporting clinician and abstracted from third-party reports (ie, published articles), and complete data were not available for all cases. If the presence or absence of a variable could not be ascertained, it was adjudicated by the Erosion Board to be unknown.
Control subjects were drawn from the PAS cohort, which consisted of 1000 patients enrolled after FDA approval of the device (from 2008 to 2012) for the purpose of ongoing assessment of safety. Criteria for participation in the PAS were essentially identical to the commercial IFU for the device. PAS patients were followed for 2 years after implant as part of the study. Controls were matched 2:1 to cases on the basis of age and sex in an effort to approximate patient size, which was not always available for erosion cases. For the few erosion patients of unknown age (n=8), controls were randomly matched as follows: if the device was >18 mm, the patient was assumed to be an adult and an adult control patient was randomly selected; if the device was <10 mm, it was assumed to be a pediatric patient (<18 years) and a pediatric control patient was randomly matched; if the device was 10 to 18 mm, a control patient was randomly selected without age limitation. Erosion cases for which any pre- or intraprocedure echocardiogram images were available for review were defined as the echocardiography case cohort. Before determining whether to match for sex, the case and PAS cohorts were compared overall and found to include a similar proportion of males and females. Era effect was investigated by dividing erosion cases into tertiles.
Erosion cases were reported to the manufacturer with no identifying information, and the requirement for informed consent was waived. Controls provided written informed consent to participate in the PAS trial, which was approved by the institutional review boards of participating centers.
All available echocardiograms from erosion cases and controls were reviewed by coinvestigators with echocardiographic expertise who were blinded to the study group and clinical outcomes. The echocardiograms submitted to SJM were not catalogued according to their temporal relationship to the ASD implant or the erosion, so all studies were initially reviewed to exclude those performed after device implant or at the time of erosion diagnosis. Only pre- and intraprocedural echocardiograms were subjected to blinded review. The echocardiographic parameters selected for evaluation were derived from previously published studies/reports relating to erosion,6,7,15,26 and are consistent with recently published guidelines,27 including the presence and length of ASD rims and ASD diameter in 3 standard views, specifically, the short-axis aortic (for the aortic and the posterior rims), 4-chamber (for the atrioventricular valve and rims toward the right upper pulmonary vein), and bicaval views (for superior vena cava [SVC] and inferior vena cava rims); the consistency of rims (thin/flailing or adequate); whether balloon sizing was performed; the alignment of the atrial septum (see below); and several measures related to the configuration of the deployed device (see below). All rims were considered deficient if they measured <5 mm in any view. The balloon-sized ASD dimension reflects either the stretched diameter or the diameter determined from stop-flow sizing, depending on the case, although the method of sizing was not always known.
Based on the previously reported association between aortic rim deficiency and erosion, reviewers were asked to measure the aortic rim length (or indicate its absence) in as many views as possible (eg, short-axis rim at 30, 40, 50 degrees, etc, on transesophageal echocardiography). If closure was guided by intracardiac echocardiography, these instructions did not apply, because the aortic rim was usually not evaluated in multiple views. Other limitations and instructions relating to intracardiac echocardiography were also recognized, including differences in standard or frequently available views.
The reviewers were also instructed to evaluate whether the aortic rim was malaligned, which was assessed in the short-axis view by extending a line from and parallel to the posterior rim to the aorta. If the line intersected the middle of the noncoronary aortic sinus, it was considered centered, otherwise it was malaligned toward the left or right atrium.15 Device orientation relative to the aorta was also categorized as remote, touching/close but not splayed (at any point in the cardiac cycle), or splayed, and inward movement/buckling of the device during the cardiac cycle was recorded as well. Because almost all documented erosions have been through the atrial free wall adjacent to the aortic root or ascending aorta, the appearance of the device in this region was evaluated in greater detail. The reviewers were instructed to evaluate whether there was tenting of the atrial free wall into the transverse sinus (ie, the anatomic space between the aortic wall and the atrial free wall) by the device, and whether the edge of the device was pushing the atrial free wall deep enough to touch the wall of the ascending aorta. Echocardiographic variables that were not overlapping the Erosion Board–adjudicated data (ie, all of the above, except for determinations of ASD rim adequacy and ASD size) are referred to below as part of the supplemental echocardiogram review.
Data were presented descriptively as frequency (%), mean±standard deviation, or median (first–third quartiles). Selected variables within the erosion cohort were compared by using χ2 test, independent samples t test, or Wilcoxon rank sum test. Erosion-associated death was defined as death during the acute episode (on arrival, during acute hospitalization, or early after surgery to treat the erosion) and was treated as a discrete, time-independent outcome and analyzed with the above methods. Variables collected from the database and those ascertained on the supplemental echocardiogram review were compared between cases and their matched controls by using conditional logistic regression analysis. For skewed continuous variables (eg, patient weight:device size ratio), log transformed values were analyzed. Factors significantly associated with erosion on univariable analysis to P<0.05 were considered for inclusion in multivariable models, which were constructed by using forward stepwise conditional logistic regression. Because of the limited data obtained from the supplemental echocardiogram review, these variables were not considered for inclusion in multivariable models. To assess for potential bias introduced by the later implant era of ASD closure in the control cohort, a supplemental conditional logistic regression analysis was also performed with adjustment for year of implant (before 2008 versus 2008 or later). Odds ratios were presented with 95% confidence intervals. The term relative risk is used descriptively in this report, and is not intended to connote the statistical measure relative risk ratio.
A total of 125 events were adjudicated as erosions by the Erosion Board between 2002 and 2014 and were used for this analysis. These events consisted of 111 product surveillance complaints, 3 events from the PAS, and 11 events identified from the published literature. An echocardiogram was available to the study investigators for 95 of these cases (Figure 1). Summary details of erosion cases are presented in Table 1, and a comparison of erosion cases pre-2008 and those diagnosed during or after 2008 (contemporaneous with the control cohort) is presented in Table I in the online-only Data Supplement. As depicted in Figure 2, the distribution of erosion cases was fairly consistent over time. It is noteworthy that devices were significantly more likely to be oversized (as determined by the Erosion Board) in the first tertile of erosion cases than in later patients (Figure 3), although absolute device sizes did not differ.
Although most (56%) of the erosions occurred in adults (≥21 years of age), nearly 20% were in patients <10 years old. The median duration from implant to erosion diagnosis was 14 days, and was ≤1 day in approximately one-third of patients (n=40); however, there were a number of patients in whom the diagnosis was made much later after ASD closure, >1 year in 16 and >5 years in 6. Patients diagnosed with an erosion within 2 days after implant were significantly younger than those presenting later (P=0.005). A large majority of patients presented with cardiac tamponade or hemopericardium, but 18 (14%) had evidence of a fistula between the atrium and aorta without a pericardial fluid collection (78% of these patients were diagnosed ≥2 weeks after implant, which was significantly more often than patients with a hemopericardium [38%; P<0.001]). Nine of the 125 patients died as a result of complications from the erosion, 7 on or within 1 day of the erosion diagnosis. The likelihood of death related to the erosion was not significantly associated with its acuity after implant; for example, the duration from implant to erosion diagnosis was <4 days in 4 patients who died and >300 days (324–475) in 3. All but 1 of the erosion patients who died were adults, and there were no deaths in patients <17 years of age. Among erosion patients, those who died were significantly more likely to have an oversized device, and to have documented erosion into the aorta (7 of 9 deaths) than those who survived, but there was no difference in the likelihood of death according to implant tertile.
Data from cases and matched controls are compared in Table 2. Deficiency of the aortic (anterior-superior) ASD rim, the SVC rim, or any rim was significantly more common in erosion cases than in controls. In the entire case-control cohort, there were 12 patients with deficiency of both aortic and SVC rims, all of whom were in the erosion group. In addition, the balloon-sized ASD diameter, the difference between sized and static ASD diameters, the absolute ASO device size, patient age:device size and weight:device size ratios, and the device size-static ASD diameter difference were significantly associated with erosion (Table 2, Figure 4). As demonstrated in Figure 5, there were no devices <12 mm in the erosion cohort, and a relative overrepresentation of devices in the 20- to 26-mm range. On multivariable analysis, deficiency of any rim (or of the aortic rim), a device size >5 mm larger than the static ASD diameter, and the patient weight:device size ratio were significantly associated with erosion (Table 3). Results of the era-adjusted analysis are summarized in Table II in the online-only Data Supplement.
As illustrated in Tables 2 and 4, there were substantial missing data for erosion cases, and for controls as well in the supplemental echocardiogram review. The echocardiograms reviewed varied considerably in completeness and imaging of the supplemental details of interest. The data obtained from the supplemental echocardiographic review are summarized in Table 4. Although confounded by the lack of completeness, there were several potentially noteworthy findings. For example, the aortic sinus was more often indented by the device in erosion cases than in controls, but this finding was relatively common in controls as well. Also, the device was remote from the aorta in 13% of controls with available data but in none of the erosion cases, whereas there was no difference on the proportion of devices that were splayed around the aorta. Because detailed rim length and deficiency data were limited in the supplemental echocardiography cohort, the additional measures were determined to be of negligible incremental value in the analysis, and thus they were not analyzed separately but were integrated with data related to rims in the Erosion Board–adjudicated data set.
Given the preponderance of deaths in the subset of patients with erosion into the aorta, subgroup analysis was performed including only the cohort of cases with documented erosion into the aorta. The result of that analysis was similar to the complete cohort, both on univariable and multivariable logistic regression, and did not reveal any unique risk factors.
The issue of cardiac erosion after ASO implant has been a part of the discussion around ASD closure since the first report more than a decade ago, but despite the rarity of this serious event, the lack of data documenting absolute or relative risk factors has created a persistent state of uncertainty. It is not just patients, families, and implanters who have had to contend with this uncertainty, but the FDA and device manufacturers as well, and judging by the short-lived listing of deficient anterior-superior rim as a contraindication to ASO implant and the subsequent recommendation for a long-term prospective trial, the general uncertainty about how to contextualize this problem persists today, 14 years after it was first reported. Thus, the goal of this case-control study was to provide the most complete overview of the experience with ASO erosions to date, and to assess relative risk factors for this rare complication, even as the ADVANCE ASO Post Market Surveillance Study has gotten underway.
The findings of this study were confirmatory of previous interpretations of erosion data in identifying an association between deficiency of the anterior-superior (aortic) ASD rim and erosion. This feature was noted for almost all of the erosion cases with available data but only 24% of controls. Although the frequency of a deficient rim in the control cohort was lower than in some other studies,24,25 it was similar to that observed by Butera et al17 in the largest study to report on that measure. Even if a higher frequency of deficient aortic rim, such as the 60% reported by O’Byrne et al,24 had been used for the control cohort in this analysis, the difference between cases and controls still would have been significant.
Although most of the deficient rims were in the retroaortic location, the deficiency of any ASD rim provided better explanatory power on multivariable analysis than did aortic rim deficiency. The difference was modest, but reflects the fact that 14 of 16 patients with a deficient SVC rim and 3 of 4 with a deficient inferior vena cava rim were in the erosion cohort. Thus, even though deficient SVC and inferior vena cava rims are less common in patients undergoing transcatheter ASD closure than a deficient aortic rim, these features should be considered similarly as relative risk factors for erosion. As with deficiency of the aortic rim, it has long been suspected that patients with deficient SVC rim may be at higher risk than those with an adequate rim, because a high secundum ASD was identified in the 2004 review by Amin at al6 as a possible risk factor.
Other than the near universality of a deficient aortic rim and higher proportions of SVC and inferior vena cava rim deficiency, erosion cases differed from controls in several respects. Specifically, balloon-sized ASD diameter, the difference between static and balloon-sized diameters, and ASO size were all larger in erosion cases than in controls. Although the same assessment of oversized device made by the Erosion Board was not available for the control cohort, these factors are all related to and consistent with a relationship between an oversized ASO device and erosion. In an effort to normalize device size to patient size, given limited data (eg, total septal length and body surface area were not available), various relationships between age or weight and device size were explored. Simple age:device size and weight:device size ratios provided the clearest differentiation between erosion cases and controls, with smaller ratios in cases overall, and a smaller proportion of erosion cases than controls above the respective medians. Although a potentially interesting finding, suggesting either a higher risk of erosion in patients with large device size for age/weight or, alternatively, a lower risk in patients with small device size for age/weight, these measures have not been validated and should be considered exploratory.
Given the strong association between a deficient aortic rim and erosion, and the paucity of other documented relative risk factors, it is important to consider whether there are differentiating features within the subset of patients with a deficient rim that may modify risk. To this end, future studies should attempt to probe more deeply into the deficient rim issue by characterizing variations within the larger group of patients with an aortic rim <5 mm, such as absent versus small rim, deficiency in multiple views (ie, longer segments of deficiency), or other discernible anatomic patterns. Notably, the current study, like most previous analyses of this topic, did not differentiate between an absent aortic rim and one that was present but deficient (ie, <5 mm). The relevance of this distinction is unknown, although a recent case series found that absence of the aortic rim was common among patients who developed erosion.15
In undertaking this study, with supplemental echocardiogram review, one of the objectives was to explore the relationship of erosion to atrial septal and device orientations, and to the interaction between the device and the transverse sinus and aortic root. Unfortunately, the supplemental echocardiogram review yielded limited data in most areas, including measures of atrial septal alignment, device orientation, indentation of the transverse sinus or aortic sinus of Valsalva by the device, and dynamic inward movement/buckling of the device. The intrinsic limitations imposed by the extensive missing data notwithstanding, there were several notable differences between erosion and control cohorts, including the orientation of the device relative to the aorta, the frequency of aortic sinus indentation by the device, and the prevalence of a thin aortic rim. In another recent case series, which included a small cohort of erosion cases that were eventually included in the Erosion Board–adjudicated data set, a thin posterior ASD rim, tenting of the atrial wall into the transverse sinus, and malalignment of the atrial septum were relatively common.15 We did not find that to be the case in this series, although, because of the limitations in the data available for these evaluations, we cannot be sure that such relationships did not exist.
The supplemental echocardiogram parameters interrogated in this study were drawn from a handful of reports and studies published over the past 14 years. The first report published by the Erosion Board in 2004 highlighted deficient aortic rim and device oversizing as likely risk factors for erosion.6 Subsequently, other putative risk factors were reported as physicians modified their practices and oversizing became less common.15,28 The supplemental echocardiogram review envisioned for this study was ambitious, aiming to collect as much information as possible while recognizing that there was likely to be substantial missing data. It was somewhat of a surprise that many of the supplemental variables could not be identified in the control cohort drawn from patients enrolled in the PAS study, indicating the need for standard imaging techniques to be adopted widely. Recently, the American Society of Echocardiography and Society for Cardiovascular Angiography and Interventions jointly published a set of guidelines for echocardiographic evaluation of ASD, which include most of the features evaluated in this study.27 In the future, these guidelines should facilitate more thorough and consistent insight into risk factors for erosion, and, ultimately, lead to reduction of the risk of this complication.
For rare events such as ASO-associated erosion, it is critical to distinguish between absolute and relative risk. By all estimates, the absolute risk of cardiac erosion after ASD closure with an ASO device is substantially <1%, and likely <1 in 1000, even if fewer than half of actual erosions have been ascertained. Given that nearly all of the reported erosions have been in patients with a deficient aortic rim, which is present in ≈25% to 50% of patients undergoing ASD closure, the absolute risk of erosion in patients with a deficient aortic rim is 2- to 4-fold higher than in the general ASD population, still <1% and potentially <0.1%. One of the goals of future studies in this area should be to identify more precise relative risk factors for erosion so that appropriate preprocedural and follow-up information can be presented to patients, families, and referring physicians, acknowledging the risks that do exist without grouping lower-risk patients in with those at higher or particular risk.
This study was limited by potential ascertainment, selection, survival, and other biases. This was unavoidable given the nature and sources of the data. As a result of limitations in the data, the resolution of the analysis was necessarily constrained as well, and there may be important risk factors for erosion that could not be discerned from this analysis.
In addition, because the control cohort was drawn from the PAS cohort, and implant criteria were generally similar to the contemporaneous IFU (ie, 2008–2012), they likely reflected changes in the IFU between device approval and the onset of the PAS. As a result, implant and evaluation practices in the control cohort may have differed from the earlier portion of the erosion group. Although all the controls underwent ASD closure in 2008 or later but only 58 of the 125 erosions, which necessarily introduces bias, characteristics of erosion patients were similar during the 2 time periods, and it is unlikely that risk factors for erosion changed over time, even if implant practices did. To address the possibility of bias resulting from this difference, an era-adjusted analysis was performed. As expected, this adjusted analysis, which effectively excluded the 54% of erosions diagnosed before 2008, and was thus reduced in power, yielded several differences from the global analysis, but these did not bear directly on the major findings.
Although cardiac erosion has been reported predominantly after ASD closure with the ASO device, it has also been reported with other occlusion devices, and after patent foramen ovale closure, but the findings of the present study cannot necessarily be extended to those circumstances.2,12,29
Deficiency of the aortic rim was the predominant relative risk factor for cardiac erosion after ASD closure with the ASO device in this case-control study, which included all the erosion cases that have been adjudicated by the manufacturer of the device. Although inherent limitations in the available data on these patients precludes definitive interpretation of these findings, leaving open the possibility of other important relative risk factors, these findings confirm and extend our collective understanding of patients at the greatest risk of this rare complication. Additional relative risk factors identified in these analyses are of uncertain significance, but generally support the hypothesis that there are potentially modifiable procedure-related factors associated with the risk of erosion. Although it is a challenging mandate to investigate risk factors for erosion in a prospective population-based manner, the findings of this study, with regard both to limitations in the available data for the largest cohort of erosion cases and to definition of additional or modifying relative risk factors, support the potential utility and incremental value of the ADVANCE ASO Post Market Surveillance Study. Finally, the general inadequacy of recorded echocardiographic data in erosion cases and PAS controls, which limited the ability of this study to provide deeper insight into risk factors for erosion, highlights the importance of comprehensive evaluation before and during ASD closure, as also reflected in the protocol for the ADVANCE trial and the recently published guidelines.28 It is our hope that ongoing and future studies will capture sufficiently detailed and extensive data to take the next step and advance our collective understanding of risk factors for erosion, and thus to facilitate further reduction in the risk of this uncommon but potentially serious complication of transcatheter ASD closure.
Sources of Funding
St. Jude Medical prepared and delivered the data and echocardiogram studies to the investigators. St. Jude Medical also provided support for one of the investigators to travel to the institutions of the participating echocardiographers to conduct instruction and quality control to ensure standardization of the supplemental echocardiogram measurements.
Dr McElhinney serves as a consultant and proctor for Medtronic, Inc. Dr Amin serves as a proctor for St. Jude Medical. The other authors report no conflicts.
The online-only Data Supplement is available with this article at http://circ.ahajournals.org/lookup/suppl/doi:10.1161/CIRCULATIONAHA.115.019987/-/DC1.
- Received October 19, 2015.
- Accepted March 14, 2016.
- © 2016 American Heart Association, Inc.
- Delaney JW,
- Li JS,
- Rhodes JF.
- El-Said H,
- Hegde S,
- Foerster S,
- Hellenbrand W,
- Kreutzer J,
- Trucco SM,
- Holzer R,
- Burch G,
- Mirani A,
- Nicolas R,
- Porras D,
- Bergersen L,
- Moore J.
- Amin Z,
- Hijazi ZM,
- Bass JL,
- Cheatham JP,
- Hellenbrand WE,
- Kleinman CS.
- DiBardino DJ,
- McElhinney DB,
- Kaza AK,
- Mayer JE Jr..
- Sarris GE,
- Kirvassilis G,
- Zavaropoulos P,
- Belli E,
- Berggren H,
- Carrel T,
- Comas JV,
- Corno AF,
- Daenen W,
- Di Carlo D,
- Ebels T,
- Fragata J,
- Hamilton L,
- Hraska V,
- Jacobs J,
- Lazarov S,
- Mavroudis C,
- Metras D,
- Rubay J,
- Schreiber C,
- Stellin G.
- Butera G,
- Romagnoli E,
- Carminati M,
- Chessa M,
- Piazza L,
- Negura D,
- Giamberti A,
- Abella R,
- Pomè G,
- Condoluci C,
- Frigiola A.
- 18.↵The U.S. Food and Drug Administration, U.S. Department of Health and Human Services. Rare Serious Erosion Events Associated with St. Jude Septal Occluder (ASO). October 17, 2013. http://www.fda.gov/MedicalDevices/Safety/AlertsandNotices/ucm371145.htm. Accessed September 8, 2015.
- 19.↵Circulatory System Devices Panel of the Medical Devices Advisory Committee. St. Jude Medical AMPLATZER® Atrial Septal Occluder Executive Summary; May 24, 2012 Advisory Committee Briefing Materials. http://www.fda.gov/downloads/AdvisoryCommittees/CommitteesMeetingMaterials/MedicalDevices/MedicalDevicesAdvisoryCommittee/CirculatorySystemDevicesPanel/UCM304944.pdf. Accessed September 8, 2015.
- 24.↵Statement of the American Heart Association to the Food and Drug Administration Circulatory System Devices Panel, May 24, 2012, Transcatheter Atrial Septal Defect Occluders. http://www.fda.gov/ucm/groups/fdagov-public/@fdagov-afda-adcom/documents/document/ucm304928.pdf. Accessed September 8, 2015.
- 25.↵Statement of the Society for Cardiovascular Angiography and Interventions to the Food and Drug Administration Circulatory System Devices Panel, May 24, 2012, Transcatheter Atrial Septal Defect Occluders. http://www.fda.gov/downloads/AdvisoryCommittees/CommitteesMeetingMaterials/MedicalDevices/MedicalDevicesAdvisoryCommittee/CirculatorySystemDevicesPanel/UCM304929.pdf. Accessed September 8, 2015.
- Mallula K,
- Amin Z.
- Silvestry FE,
- Cohen MS,
- Armsby LB,
- Burkule NJ,
- Fleishman CE,
- Hijazi ZM,
- Lang RM,
- Rome JJ,
- Wang Y.
Cardiac erosion is a rare but potentially serious complication of transcatheter atrial septal defect closure using the Amplatzer Septal Occluder. The absolute risk of erosion after Amplatzer Septal Occluder implant has been estimated to range from 0.043% to 0.3%, but little is known about specific mechanisms and risk factors for this complication. Therefore, we undertook the current case-control study, in which all erosion cases that have been reported and entered into the St. Jude Medical database were analyzed against an age- and sex-matched control cohort. A total of 125 erosion events were reported and adjudicated between 2002 and 2014, and were diagnosed a median of 14 days after atrial septal defect closure, with 9 deaths. Deficiency of the aortic rim was an almost universal finding in cases, and deficiency of any rim (ie, aortic, superior vena cava, inferior vena cava) was the predominant risk factor for erosion. There were a number of other differences between cases and controls that implicate device oversizing as an additional contributing factor and support the hypothesis that there are potentially modifiable procedure-related factors associated with this complication. The findings of this study, with regard both to limitations in the available data for the largest cohort of erosion cases and to definition of additional or modifying relative risk factors, support the potential utility and incremental value of the prospective ADVANCE ASO Post Market Surveillance Study. Finally, the general inadequacy of recorded echocardiographic data in both cases and controls, which limited the ability of this study to provide deeper insight into risk factors for erosion, highlights the importance of comprehensive evaluation before and during atrial septal defect closure, as also reflected in the protocol for the ADVANCE trial and recently published guidelines.