Author + information
- Received May 2, 2006
- Revision received November 13, 2006
- Accepted November 16, 2006
- Published online June 5, 2007.
- Thomas K. Jones, MD, FACC⁎,a,⁎ (, )
- Larry A. Latson, MD, FACC†,a,
- Evan Zahn, MD, FACC‡,a,
- Craig E. Fleishman, MD, FACC§,
- Joth Jacobson, MS¶,
- Robert Vincent, MD, FACC∥,
- Kirk Kanter, MD∥,
- Multicenter Pivotal Study of the HELEX Septal Occluder Investigators#
- ↵⁎Reprint requests and correspondence:
Dr. Thomas K. Jones, Children’s Heart Center, 4800 Sand Point Way NE, Seattle, Washington 98015.
Objectives This study sought to compare the safety and efficacy of the HELEX septal occluder (HSO) with surgical repair of atrial septal defect (ASD).
Background The HSO is a low-profile, double-disk occluder device for percutaneous closure of secundum ASD.
Methods Patients were enrolled (HSO arm prospectively, surgery arm prospectively/retrospectively) from 14 U.S. sites and followed up for 12 months postprocedure. Investigator-reported outcomes were evaluated, including closure success (no or clinically insignificant residual shunt) and the incidence of adverse events. The first 3 HSO patients at each site were considered training cases and were excluded from analysis.
Results Between March 2001 and April 2003, 119 nontraining cases received an HSO and 128 had surgical repair. The groups were similar with statistical but clinically unimportant differences in median age, weight, and preprocedural echocardiographic defect size. Anesthesia time and hospital stay were significantly shorter in the HSO group. Closure success, defined as complete closure or a clinically insignificant residual shunt, was similar in both groups. Major and minor adverse events rates were not statistically different. The most common major adverse events for the HSO group was device embolization requiring catheter retreival (1.7%), and in the surgery group was postpericardiotomy syndrome (6.3%), including one death because of tamponade. The primary end point, clinical success, a composite of closure success and no major adverse events at 12 months, satisfied the noninferiority hypothesis comparing device closure with surgery.
Conclusions Closure of ASD with the HELEX septal occluder is safe and effective when compared with surgical repair, with reduced anesthesia time and hospital stay. (U.S. Multicenter Pivotal Study of the HELEX Septal Occluder for Percutaneous Closure of Secundum Atrial Septal Defects; this study was approved by the Food and Drug Administration before the National Institutes of Health website was active, so there is not a URL or registration number.)
Secundum atrial septal defect (ASD) is a common congenital cardiac anomaly accounting for approximately 10% of all congenital heart disease (1,2). It is one of the most common congenital heart defects to present in adulthood (3). Untreated, ASD produces right heart volume overload and progressive impairment over time, including reduced aerobic capacity, atrial dysrhythmias, congestive heart failure, and pulmonary hypertension (4–6). In the U.S. alone it is estimated that approximately 10,000 new patients per year can be expected to have an ASD. Successful surgical repair of ASD has been performed for 50 years with continued improvement in technique and outcomes (7). King and Mills (8) reported the first transcatheter closure of ASD in 1976, but the delivery system was quite large and impractical, especially for younger patients. With time, improvements in design concepts and materials discoveries have led to improved results in transcatheter closure systems (9). However, it has been only recently that a transcatheter ASD closure device was deemed suitably effective and safe to receive U.S. Food and Drug Administration approval for market release (10).
The HELEX septal occluder (HSO) (W. L. Gore & Associates, Flagstaff, Arizona) is a low-profile, double-disk occluder device designed to close secundum ASDs. The device is composed of an expanded polytetrafluoroethylene membrane bonded to a single nitinol wire frame and can be delivered though a 9-F femoral venous sheath. The device is packaged with its integral delivery system (Fig. 1).Because it is a compliant, non–self-centering device, it is capable of conforming well to the curvilinear surfaces of the atrial septum. The delivery system allows for repositioning or retrieval of the device after deployment. A safety cord attached to the device provides an additional level of patient safety by allowing for removal of the occluder even after device release. Previous reports have shown the apparent safety of this device in initial clinical evaluations with high levels of clinical success (11–15).
The HELEX Pivotal Study was a nonrandomized, multicenter trial conducted in 14 medical centers in the U.S. that evaluated the safety and effectiveness of the HSO in patients with ostium secundum ASD. The primary objective of the study was to show that the HSO was not inferior to open heart surgical repair for the treatment of ostium secundum ASDs. Patients were enrolled prospectively to receive the HSO and compared with surgical patients enrolled both prospectively and retrospectively. Participating centers and investigators with experience in transcatheter and surgical treatment of congenital heart disease were selected. Individual investigators received standardized qualification and training including a series of device implantations performed in animal subjects before participation in the clinical trial. The study protocol was approved by each participating institution’s institutional review board, and informed consent was obtained from each study participant.
Inclusion criteria for enrollment included the presence of an ostium secundum ASD and evidence of right heart volume overload. Additional criteria for the device arm patients was a balloon occlusion defect diameter ≤22 mm and the presence of adequate septal rims to secure the device as judged by the individual investigator at the time of implantation. In the surgical arm, patients could be enrolled retrospectively within 12 months of institutional review board approval.
Exclusion criteria for the study included the presence of concurrent cardiac defects requiring surgical repair or significant comorbidities including a history of stroke, pulmonary hypertension, pregnancy, or the presence of multiple ASDs requiring the use of more than one device (device arm only).
Surgical repair of the ASD was carried out using accepted standard surgical techniques and individual surgical center protocols. For patients enrolled in the device arm, details of the atrial septal anatomy were evaluated by either transesophageal or intracardiac echocardiography at the time of the cardiac catheterization procedure. At that time a final determination was made by the individual investigator regarding the patient’s suitability for device implantation based on the inclusion criteria stated above.
Patients in each group were evaluated at 24 h, 4 weeks, 6 months, and 12 months after treatment. Follow-up visits included a history and physical examination, electrocardiogram, and transthoracic echocardiogram. Additionally, for patients enrolled in the device arm, fluoroscopy of the device was performed at the 6- and 12-month follow-up visit.
The primary end point of the HELEX Pivotal Study was clinical success, a composite of safety and efficacy evaluated at 12 months postprocedure. The criteria for clinical success were: 1) no major device or procedure adverse event through the 12-month follow-up; 2) no repeat procedure to the target ASD; and 3) clinical defect closure (complete occlusion or clinically insignificant leak) at the 12-month follow-up. Secondary end points evaluated included the individual outcomes of safety and efficacy. Safety outcomes for the study included assessment of device-related and procedure-related adverse events through the 12-month follow-up. Efficacy outcomes included assessment of defect closure at each of the follow-up visits. Residual defect status was classified as: 1) complete occlusion; 2) clinically insignificant leak with a small residual shunt defined as a residual defect ≤3 mm and never >6 mm accompanied by resolution of right ventricular enlargement and normalization of interventricular septal motion; or 3) clinically significant leak.
The first 3 patients undergoing an attempt at device placement at each site were considered training cases and were excluded from the primary analyses comparing outcomes with surgical controls. Subsequent analysis of the training cases identified no statistically significant differences compared with nontraining cases in patient characteristics or outcomes.
Baseline patient characteristics were compared between device nontraining and surgical control patients. Continuous measures were compared using the Wilcoxon rank sum test. Categorical measures were compared by Fisher exact test or chi-square test. The proportions of patients experiencing one or more adverse events overall, and by subgroups of interest, were compared using the Fisher exact test. The proportions of patients achieving the primary end point, clinical success, were compared using a binomial proportions test with a noninferiority margin of 10%. A value of p < 0.05 was considered evidence of noninferiority. A propensity scores analysis was used to account for baseline differences in patient populations. The SAS software (version 8, SAS Institute Inc., Cary, North Carolina) was used to perform all analyses.
A total of 321 patients were enrolled in the study from 14 U.S. sites between March 2001 and April 2003. The device arm enrolled 50 training patients and 143 nontraining patients, and the surgical control arm enrolled 128 patients. Surgical procedure dates ranged from January 2000 to December 2002. Subsequent discussion of the data is restricted to the 143 device nontraining and 128 surgical control patients.
Patient demographics are compared in Table 1,and medical histories and current medications are compared in Table 2.On average, device patients were older with comparatively smaller defects than surgical controls. These statistical differences served as the basis for the propensity scores analysis.
Delivery of the HSO was attempted in 135 (94.4%) of the 143 enrolled device patients. The 8 patients with no delivery attempt were found during catheterization to have a balloon-occlusion defect size >22 mm. Therefore, these patients were excluded per study eligibility criteria. Successful device delivery was achieved in 119 (88.1%) of the 135 patients with delivery attempts. Device delivery was unsuccessful in 10 of the 16 patients because of the following anatomical considerations: multiple defects requiring more than one device, 2 patients; inadequate defect rim, 1 patient; thin or floppy septum, 2 patients; impingement on cardiovascular structure, 1 patient; other characteristics (e.g., intracardiac dimensions too small compared to defect size) considered by the investigator to be inappropriate for successful placement, 4 patients. In the 6 remaining patients, the investigator was unable to achieve satisfactory device placement.
Procedural details of the 2 study groups are presented in Table 3.In the device group, the median balloon size of the defect was 14 mm, compared with a resting median diameter of 10 mm. The ratio of the balloon stretch to resting diameter of 140% is consistent with previously published reports and suggests that the investigators took care to avoid overstretching the defect by observing the minimal balloon diameter at which shunting ceased, referred to as the “stop-flow” diameter (16). The median fluoroscopic time of 22 min in the device group is typical for routine diagnostic catheterizations performed in patients with congenital heart disease and similar to the reported time for the only other approved atrial septal occluder device available in the U.S. (10). The duration of anesthesia time and of hospital stay was also significantly less in the device group compared with the surgical control group.
Major adverse events were reported in 5.9% of device patients (7 of 119) and in 10.9% of surgical control patients (14 of 128) through the 12-month visit (Table 4).In the device arm, removal of the device was required in 5 of the 7 patients. Two patients experienced postprocedure device embolization (both within 24 h), 2 patients had device sizing issues that required removal (both on day 1), and 1 patient had a potential nickel sensitivity that eventually resulted in device removal (day 129). The device was removed by percutaneous transcatheter technique in 5 of the 6 patients. The 1 patient with the reported potential nickel sensitivity underwent surgical removal and defect closure at the request of his parents. Additionally, 1 patient experienced a procedure-related major adverse event of retroperitoneal hemorrhage. The manipulation of an intracardiac echocardiogram catheter through tortuous iliac venous anatomy in this elderly anticoagulated patient was believed by the investigator to have been the etiology of the bleeding event. One patient experienced a major adverse event with a potential, but not definite, relationship to the device or procedure. The patient reported an acute, confusional migraine requiring hospitalization on day 6 after implantation. Imaging studies did not show any evidence of a cerebral embolic event, and the complaint resolved spontaneously. In the surgical control arm, the major adverse event of pericardial effusion caused by postpericardiotomy syndrome requiring percutaneous drainage was reported in 8 patients, with pericardial tamponade contributing to the death of 1 patient.
Minor adverse events were reported in 27.7% (33 of 119) of device patients with a successful delivery and in 28.1% (36 of 128) of surgical control patients through the 12-month follow-up period. In the device arm the most common minor adverse events were arrhythmias (5.0%) and headaches (4.2%). Fractures in the wire frame of the device were reported in 6 patients through the 12-month visit. None of these patients required intervention, and the Data Safety Monitoring Board adjudicated these as minor adverse events. In the surgical control arm, the most common minor adverse events were postpericardiotomy syndrome not requiring hospitalization (7.8%), followed by arrhythmias (3.9%), pericardial effusions (3.9%), and pneumopericardium (2.3%).
The residual defect status was determined on the 12-month echocardiogram by an independent echocardiography core laboratory. Successful defect closure by either method was defined by the study protocol as complete closure or the presence of a clinically insignificant residual defect. A clinically insignificant residual defect was further defined as associated with normalization of right ventricular volume and interventricular septal motion by transthoracic echocardiography, and typically <3 mm and absolutely <6 mm diameter as determined by color Doppler flow image. Of the 82 patients evaluated in the surgical control arm by the core laboratory, 100% were determined to have successful defect closure at 12 months postprocedure (Table 5).Of the 105 patients evaluated in the device arm, 98.1% were determined to have successful defect closure on final evaluation. Significant residual defects were identified in 1.9% (2 of 105) of device patients on final evaluation. Both of these patients had multifenestrated defects that the investigator believed would be incorporated into the closure during the healing response over the first year. Neither of these patients was noted to have wire frame fractures or any other device-related adverse events.
The primary end point of the pivotal study was clinical success, a composite evaluation of safety and efficacy. The clinical success end point was determined at 12 months postprocedure. A clinical success end point was determined only for patients with a successful device delivery or surgical closure and follow-up data on residual defect status and adverse events.
For the protocol-specified analysis, clinical success rates in eligible patients were compared between nontraining device and surgical control patients. Clinical success was determined in 109 of the 117 eligible device patients and in 86 of the 124 eligible surgical control patients (Table 6).Of the 46 patients not evaluated, 20 discontinued study follow-up without a major adverse event and before final defect evaluation (2 device patients, 18 surgical control patients), and 26 missed final core laboratory evaluation of residual defect status (6 device patients, 20 surgical control patients). Clinical success was achieved in 91.7% (100 of 109) of device patients and in 83.7% (72 of 86) of surgical control patients. Based on these data, the null hypothesis of HSO inferiority was rejected in favor of the noninferiority alternative hypothesis with p < 0.001.
Differences in patient age, body size, estimated defect size before the procedure, and certain medical histories were observed between the 2 arms (Tables 1 and 2). A propensity score analysis was conducted as a method to adjust for factors that could confound the differences in clinical success rates discussed above. The results show that the clinical success rate for the HSO remains not inferior to surgical closure when controlling for these baseline differences in treatment arms (Table 7).
This was not a randomized comparison trial, nor would it be possible or appropriate in the current era to conduct such a trial when alternative transcatheter closure methods are commercially available. However, every attempt was made to conduct a contemporaneous comparison with surgical closure within the same study institutions. Demographic differences existed between the 2 groups but were shown by a propensity score analysis to have no significant impact on the study results. The other limitation is the number of surgical control patients lost to follow-up, with only 67% of the enrolled control patients completing the 12 month evaluations. However, site-reported outcomes of residual defect status were reviewed for all patients who did not have a final core laboratory review. Six-month or 12-month site evaluations were available for most device patients missing a final core laboratory review. No clinically significant residual leaks were reported in any of these device patients. Site-reported evaluations for the surgical patients who missed final core laboratory review were available from the 12-month visit for 8 and from the predischarge evaluation for all of the remaining patients. In all of these cases, the sites reported complete defect closure.
This study shows that the HSO is a safe and effective alternative for the transcatheter closure of secundum ASDs with a balloon occlusion diameter of <22 mm when compared with surgical closure. There is no statistical difference in clinical success between the two treatments. A reduced hospital stay and anesthesia time is associated with the use of the HSO. Additionally, cardiac-related major adverse events are less common with the use of the HSO, primarily because of the absence of significant postoperative pericardial effusions seen in the surgical treatment group.
The choice of the most appropriate method for closure of clinically important ASD includes a number of issues that relate to the compatibility of the device with specific patient anatomy. The compliant nature of the HELEX occluder and its non–self-centering construction are favorable design features for certain anatomical forms of ASD. For larger defects >18 mm to 20 mm, alternative transcatheter devices or surgical repair will still be required. However, greater freedom of choice by the congenital interventional cardiologists and their patients in the selection of devices for ASD closure is now available.
The authors acknowledge the support of all of the investigators and their staff at each of the participating institutions. A complete list is available in the Appendix. Special mention also goes to Warren Cutright, DVM, and Barbara Fisher, BS, MBA, at Gore Medical Products.
For a complete list of the participating institutes and investigators, please see the online version of this article.
↵a Additionally, Drs. Jones, Latson, and Zahn are consultants to W. L. Gore & Associates, Inc., the manufacturer of the Helex Occluder and the sponsor of this study.
All of the authors participated as site investigators at their respective institutions, with the exception of Mr. Jacobson, who is an employee of W. L. Gore & Associates, Inc. and provided statistical support.
- Abbreviations and Acronyms
- atrial septal defect
- HELEX septal occluder
- Received May 2, 2006.
- Revision received November 13, 2006.
- Accepted November 16, 2006.
- American College of Cardiology Foundation
- Dickinson D.F.,
- Arnold R.,
- Wilkenson J.L.
- Carlgren L.E.
- Konstantinides S.,
- Geibel A.,
- Olschewski M.,
- et al.
- Horvath K.A.,
- Burke R.P.,
- Collins J.J.,
- Cohn L.H.
- Du Z.D.,
- Hijazi Z.M.,
- Kleinman C.S.,
- et al.
- Zahn E.M.,
- Wilson N.,
- Cutwright W.,
- Latson L.A.
- Amin Z.,
- Hijazi Z.M.,
- Bass J.L.,
- Cheatham J.P.,
- Hellenbrand W.E.,
- Kleinman C.S.