Biological Response to Bard Clamshell Septal Occluders in the Canine Heart
Background The Clamshell Septal Occluder has been used to close various congenital heart defects. The purpose of this study was to evaluate the long-term biological response to this device after placement in the canine heart. Previous in vivo studies with device placement were limited to 60 days.
Methods and Results An atrial septal defect was created in dogs by blade septostomy followed by balloon dilation. Both old and new (redesigned) devices were placed. Angiographic follow-up was performed at 1, 3, and 6 months and 1 and 2 years after device placement with groups of dogs euthanitized at the same intervals. Gross and microscopic assessment was done on the explanted devices. The implants were covered at least 50% by neointima at 1 month and covered completely by 3 months. There was no thrombus formation. Areas of focal hemorrhage were evident at 1 month and were not present at 3 months. The fibrous capsule that covered the device became more densely organized and neovascularized by 2 years. A focal foreign body reaction at the device-tissue interface persisted for 2 years. There were no arm fractures with either the old or new devices in these dogs.
Conclusions The Bard Clamshell Septal Occluder is well tolerated in the canine heart for at least 2 years and elicits a normal healing process.
The Clamshell Septal Occluder was first described in 1987 as a nonsurgical method for closure of atrial septal defects.1 The device has since been used in the transcatheter closure of multiple defects, including muscular ventricular septal defects, coronary fistulae, paravalvular leaks, and surgical fenestrations.2 3 4 5 Previous in vivo studies showed a normal healing response to device placement, but these observations were limited to 60 days after implantation.1 Thus, the purpose of this study was to evaluate the long-term in vivo effects of device placement in the canine heart.
The study was designed specifically to evaluate the host biological response to device placement. Both the original and redesigned Clamshell Septal Occluders were used. Both devices have identical Dacron fabric but differ in arm design and metal alloy used for manufacture. The implant study period was 2 years, with both devices studied long term. Inappropriately large devices (>23 mm) and incorrectly positioned devices were excluded from the study. The use of mongrel dogs and the experimental protocols were approved by both the University of Nebraska Medical Center and the Cleveland Clinic Foundation Animal Research committees.
Creation of the Atrial Septal Defect
The dogs’ weights ranged from 20 to 35 kg. Pentothal (20 to 25 mg/kg IV) was given for analgesia. Ketamine (22 mg/kg IM) was given, and the dogs were intubated. Isoflurane was administered as a general anesthetic. The right femoral vein was entered percutaneously. An 8F Mullins transseptal sheath was advanced to the right atrium under fluoroscopy, and the left atrium was accessed transseptally. A blade atrial septostomy was performed by use of a 9.4-mm Park blade septostomy catheter. The newly created defect was dilated further with a 20-mm×4-cm balloon dilation catheter. Early in the study, the device was placed immediately after the defect was created in 6 dogs. Original devices were placed in these dogs long term (>6 months). To ensure that an inflammatory reaction to the septostomy did not alter the short-term histological response to the device, the remaining devices were placed 2 weeks after creation of the atrial septal defect.
An 11F Mullins sheath was advanced across the atrial septal defect. Heparin (100 U/kg IV) was given. A 23-mm Bard Clamshell Septal Occluder was loaded into the delivery catheter and advanced into the left atrium. The distal device arms were deployed into the left atrium and withdrawn against the septum. The sheath was pulled back, exposing the proximal arms in the right atrium. The device was released after it was appropriately positioned. Device position was confirmed by fluoroscopy. The device was considered to be appropriately positioned if all device arms were separated by the atrial septum and opposed each other across the septum (Fig 1⇓). A cineangiocardiogram, with injection into the right atrium, was performed to evaluate for shunting.
Two cineangiocardiograms were performed, with injection into the right atrium at 1, 3, and 6 months and 1 and 2 years. The injections were filmed through levophase. The camera was positioned to view the device from the side perpendicular to the atrial septum and en face. The films were reviewed for any shunting, filling defect, or arm fracture.
Groups of dogs were euthanatized at 1, 3, and 6 months and 1 and 2 years. The dogs were euthanatized with a combination of pentobarbital, amytal, and thiopental (5000 mg) by rapid infusion. Postmortem fluoroscopy was performed to rule out any device arm fractures. The right and left atrial gross anatomic features were photographed. The explanted heart was sectioned and embedded in paraffin for histological examination. The histology slides were stained with hematoxylin and eosin to evaluate cellularity, Masson’s trichrome and picrosirius red to evaluate collagen, and fibrous capsule and phosphotungstic acid–hematoxylin to evaluate fibrin and thrombosis. Two experienced pathologists (B.M. and J.M.A.) reviewed the gross and microscopic specimens.
Comparison of arm fractures, shunting, and filling defects between the new and old devices was performed with χ2 analysis. Because no differences were noted between the two groups, a probability value could not be generated.
Implantation and Physiology
Twenty-nine dogs underwent placement of the device. Ten dogs had original devices; the remainder receivedredesigned devices. There were no complications during device placement. Shunting in either direction, filling defects, and arm fractures were not seen angiographically in the 10 dogs with old devices and the 19 dogs with new devices.
Ten dogs were euthanatized at 1 month (4 with old devices; 6 with new devices). On gross examination, at least 50% of the surface area of the devices was covered by neointimal endothelium (Fig 2A⇓ and 2B⇓). Tissue ingrowth was deficient in the central portion of blood-exposed surfaces of the device. Focal superficial hemorrhage was seen in these tissue-deficient areas, but thrombus formation was not observed. Areas of focal hemorrhage were seen at the edges of the device arm–tissue interface apparently caused by device motion against the atrial septum. The Dacron margins were not confluent with the septum. On microscopic examination, an expected one- to two-cell-thick foreign body reaction with macrophages and scattered foreign body giant cells was seen at the fabric-tissue interface. The fabric was encapsulated with a developing fibrous capsule consisting of collagen and granulation tissue (Fig 2C⇓). Endothelium was seen on the fibrous capsule on certain sections. No fibrin deposits were noted. The atrial septal defect underwent healing with fibrosis. In several devices, a space existed between the healing Dacron margin of the device and the underlying myocardium (Fig 2D⇓). No thrombi were noted in these spaces.
All 3 of the dogs euthanatized at 3 months had new devices in place. On gross examination, the devices were completely endothelialized (Fig 3A⇓ and 3B⇓). No focal hemorrhage or thrombus was noted. The focal hemorrhage from device motion had resolved. In most instances, the device material was contiguous with the atrial septum. Microscopically, a focal foreign body reaction was present at the device-tissue interface. A thin fibrous capsule with overlying endothelium, ie, neointima, encompassed the device (Fig 3C⇓).
Five dogs were euthanatized at 6 months (2 with old devices; 3 with new devices). The fibrous capsule had thickened. The blood-tissue interface was smooth, without thrombus or fibrin deposition. The device-tissue interface was completely covered with the fibrous capsule (Fig 4A⇓ and 4B⇓). Microscopically, the fibrous capsule had become thickened, with the focal foreign body reaction present at the device-tissue interface (Fig 4C⇓).
One- and Two-Year Follow-ups
Six dogs were euthanatized at 1 year (2 with old devices; 4 with new devices); 5 dogs were killed at 2 years (2 with old devices; 3 with new devices). The findings at the 1- and 2-year follow-ups were similar. On gross examination, the fibrous capsule was thicker than at earlier time points and, in some instances, had almost completely covered the device arms (Fig 5A⇓ and 5B⇓). The blood-tissue interface remained smooth, without signs of thrombus. Microscopically, the fibrous capsule was dense with more organization of the collagen (Fig 5C⇓). Neovascularization was seen in some instances. The one- to two-cell-thick focal foreign body reaction was present and did not differ from the reaction found in the dogs at 1 month.
Placement of the Clamshell Septal Occluder, either old or new, is followed by a normal healing response for at least 2 years. There is a focal foreign body reaction at the Dacron mesh–tissue interface. This reaction is a normal healing response and was demonstrated previously in early device healing and with other Dacron fabrics.1 6 This focal foreign body reaction at the device-tissue interface did not change in composition or intensity over the duration of observation.
This study showed that in the canine heart, tissue ingrowth and endothelialization were rapid and complete by 3 months. The device was not thrombogenic even without anticoagulation therapy. Areas of focal hemorrhage along the device arms probably were caused by device motion against the atrial septum. These areas of hemorrhage were not seen in dogs after the first month, suggesting that within the first 90 days, the device becomes well integrated into the atrial septum by the healing response. It is interesting that at 1 month, the Dacron margins of the device were not necessarily apposed to the atrial septum. Physiological space existed between the atrial septum and the Dacron margin, but no evidence of thrombosis or fibrin deposition was seen. One could speculate that normal flow dynamics kept these areas free of blood stasis or turbulence during healing. Pulmonary and systemic venous return were not compromised at any time. The blood-tissue interface remained smooth without any changes that would suggest turbulent blood flow along the device. The healing response within the fibrous capsule became more organized as time progressed. The collagen became dense with associated tissue neovascularization. These findings suggest that placement of the device would be well tolerated long term.
Although the primary purpose of this study was to evaluate the healing response, the incidence of arm (wire) fractures also was addressed. Device placement was well tolerated in the canine heart, and no fractures occurred with the old or new devices.
In conclusion, the long-term healing response to the Bard Clamshell Septal Occluder was similar to the expected foreign body reaction in the canine heart. There was no thrombus formation, and complete endothelialization occurred by 3 months. There was a normal foreign body reaction, with fibrous capsule formation and endothelialization to the Dacron mesh that persisted unchanged throughout the study period. Clinical trials of the new device are needed to evaluate its efficacy in humans.
Reprint requests to Larry A. Latson, MD, Center for Pediatric and Congenital Heart Diseases, Cleveland Clinic Foundation, Desk M40, 9500 Euclid Ave, Cleveland, OH 44195.
- Received June 28, 1995.
- Revision received October 24, 1995.
- Accepted October 30, 1995.
- Copyright © 1996 by American Heart Association
Lock JE, Rome JJ, Davis R, van Praagh S, Perry SB, van Praagh R, Keane JF. Transcatheter closure of atrial septal defects: experimental studies. Circulation. 1989;79:1091-1099.
Rome JJ, Keane JF, Perry SB, Spevak PJ, Lock JE. Double-umbrella closure of atrial defects: initial clinical applications. Circulation. 1990;82:751-758.