The data in this manuscript were presented at the 27th Annual American College of Veterinary Internal Medicine Forum, Montreal, Canada, June 2009.
Utility of Transesophageal Echocardiography for Transcatheter Occlusion of Patent Ductus Arteriosus in Dogs: Influence on the Decision-Making Process
Article first published online: 24 AUG 2010
Copyright © 2010 by the American College of Veterinary Internal Medicine
Journal of Veterinary Internal Medicine
Volume 24, Issue 6, pages 1407–1413, November/December 2010
How to Cite
Saunders, A.B., Achen, S.E., Gordon, S.G. and Miller, M.W. (2010), Utility of Transesophageal Echocardiography for Transcatheter Occlusion of Patent Ductus Arteriosus in Dogs: Influence on the Decision-Making Process. Journal of Veterinary Internal Medicine, 24: 1407–1413. doi: 10.1111/j.1939-1676.2010.0587.x
- Issue published online: 3 NOV 2010
- Article first published online: 24 AUG 2010
- Submitted April 19, 2010; Revised June 18, 2010; Accepted July 9, 2010.
Background: Appropriate device selection for transcatheter occlusion of patent ductus arteriosus (PDA) is essential to procedural success.
Objectives: To determine if transesophageal echocardiography (TEE) influences device selection for PDA occlusion and to report benefits, limitations, and complications associated with TEE.
Animals: Twenty-two client-owned dogs with left-to-right shunting PDA.
Methods: PDA dimensions were obtained via transthoracic echocardiography (TTE) and then TEE followed by angiography. Based solely on information from TTE and angiography, an initial device type and size were selected. After initial device selection, TEE measurements were disclosed and changes in device selection were recorded. After device release, angiography, TEE, or both were performed to assess occlusion.
Results: An Amplatz canine duct occluder (ACDO) was securely positioned and released in 21 dogs and an embolization coil was deployed in 1 dog. Based on TEE evaluation, initial selected device type was unchanged but ACDO size was changed in 3 dogs. TEE was utilized throughout the procedure allowing real time visualization of device deployment, release and assessment of closure in 17 dogs. No complications occurred related to TEE. Complete PDA closure was achieved in all dogs.
Conclusions and Clinical Importance: TEE provided anatomic information regarding PDA morphology that closely approximated angiographic ductal dimensions while aiding in device deployment, release and confirmation of closure. We conclude that TEE provides complementary anatomical and intraprocedural information and is well tolerated in dogs.
Amplatz canine duct occluder
patent ductus arteriosus
Transcatheter occlusion of patent ductus arteriosus (PDA) is an interventional cardiac procedure that frequently is performed in veterinary medicine. Several devices including the Gianturco embolization coil, Amplatzer vascular occlusion plug and Amplatz canine duct occluder (ACDO) have been used for transcatheter occlusion of PDA in dogs.1–5 The type of occlusion device selected is based on a combination of factors including PDA morphology, minimal ductal diameter, ductal length, ampulla width, and patient size. Histologic sections of canine PDA document incomplete smooth muscle in the ductal wall, preventing normal constriction and commonly resulting in an asymmetrical lumen.6 The most common PDA morphology (type II) tapers as the ductus communicates with the pulmonary artery, thereby providing a shelf of tissue useful for securely deploying an occlusion device.7 Device embolization may be more likely if the PDA does not taper (type III) or if estimates of ductal dimensions are inaccurate prompting deployment of an inappropriately sized device.4,7,8 Minimal ductal diameter measurements provide a reference commonly used for selecting device size. Measurements can be obtained by a combination of transthoracic echocardiography (TTE), angiography, and transesophageal echocardiography (TEE).9,10 Several studies in children with PDA have documented the utility of TEE for providing useful anatomical and intraprocedural information related to device type and size selection, facilitation of device deployment and release, and confirmation of closure.11,12 Although TEE is an established imaging modality for interventional cardiac procedures in humans, limited information exists regarding the utility of TEE for interventional cardiac procedures in veterinary medicine.9,13
The purpose of this prospective case series was to determine if TEE influences device selection and sizing in transcatheter occlusion of PDA and to report the benefits, limitations, and complications associated with the use of TEE in this clinical cohort.
Materials and Methods
Client-owned dogs referred for definitive treatment of PDA were evaluated by the cardiology service at the Texas A&M University Veterinary Teaching Hospital. Dogs weighing >2.5 kg with a left-to-right shunting PDA were included. Complete TTE was performed with a GE Vivid 7a with an appropriately selected 3.5–10 MHz phased-array transducer. Measurements of the PDA were acquired from the right parasternal short-axis basilar or left cranial short-axis view depending on which provided the clearest image of the PDA in each individual dog. Measurements included minimal ductal diameter and ampulla width (Fig 1). The dogs were anesthetized and placed in right lateral recumbency. An incision was made over the right femoral artery. Once the artery was isolated, a modified Seldinger technique was used to place a long sheath and dilatorb either over a guide wirec directly into the artery or through an appropriately sized introducer.d,e The long sheath was advanced to the descending thoracic aorta immediately distal to the PDA. While femoral artery access was being obtained, a member of the cardiology team performed a TEE study to determine PDA morphology and dimensions including minimal ductal diameter and ampulla width and length with a GE multiplane TEE probef (Fig 2). A mouth guard was placed, and the probe lubricated before insertion into the oral cavity. The probe was carefully advanced to a midesophageal position until a long-axis 4-chamber view was obtained. The probe was slowly withdrawn to a cranial esophageal position, maintained in a neutral to retroflexed position and rotated between 40° and 80° until the PDA was visualized between the aorta and pulmonary artery. Ductal morphology and measurements were recorded but not disclosed to members of the cardiology team performing the transcatheter occlusion procedure. Once the long sheath was in place, angiography was performed by injecting approximately 1 mL/kg of a nonionic contrast agent.g Morphology, minimal diameter, and ampulla width and length of the PDA were recorded (Fig 3). Angiographic measurements were based on comparing diameter of the contrast within the ductus to a measuring catheterh placed within the esophagus. Based on TTE and angiographic information, the members of the cardiology team performing the transcatheter occlusion procedure made a preliminary selection of an occlusion device type and size. The preliminary selection was recorded and then the ductal morphology and measurements acquired with TEE were revealed. Any change in device type or size made based on the additional information acquired by TEE was recorded. Also recorded was whether or not TEE was used during the procedure to monitor device deployment, release, and ductal closure.
Transcatheter occlusion with an ACDOi or embolization coilj was performed as described previously.5,14 Angiography and TEE (when utilized intraoperatively) were performed within 10 minutes of device release to assess closure. The morning after the procedure, a complete TTE study was performed and residual ductal flow was recorded if present. Residual ductal flow documented with angiography, TEE, or TTE was recorded as none, trivial, mild, moderate, or severe as described previously.3
Data are reported as mean ± standard deviation when normally distributed, or as median and range when not normally distributed. Comparison of ductal dimensions obtained with angiography, TTE, and TEE was performed by Bland-Altman analysis with commercially available software.k
Twenty-two dogs were included in the series. Seventeen dogs were female, 6 of which were spayed. Five dogs were intact males. The median age at presentation was 6 months (range, 2.9–115 months), and median weight was 8.5 kg (range, 2.6–24 kg). Breeds represented included 3 Weimaraners (14%), 3 German Shepherds (14%), 2 Bichon Frise (9%), 5 mixed breed (23%), and 1 each of Pomeranian, Border Collie, Newfoundland, West Highland White Terrier, Boston Terrier, Maltese, Coton de Tulear, Italian Greyhound, and Labrador Retriever. Four dogs presented with clinical signs including exercise intolerance (n = 2) and cough (n = 2). Physical examination findings including cardiac auscultation were consistent with the presence of a left-to-right shunting PDA in all dogs. Six dogs had concurrent heart disease which included subaortic stenosis diagnosed by the presence of a subvalvular ridge and persistently increased left ventricular outflow tract velocities after PDA occlusion (n = 3), degenerative mitral valve disease (n = 2), and ventricular septal defect with persistent left cranial vena cava (n = 1). One of the dogs with degenerative mitral valve disease had an estimated systolic right ventricular-to-right atrial pressure gradient of 70.6 mmHg before occlusion and 40.2 mmHg after ductal occlusion. Degenerative mitral valve disease was diagnosed in the 2 oldest small breed dogs (68.4 and 115 months).
Based on angiographic and TEE assessment, PDA morphology was recorded as type I (n = 1) or type II (n = 21).7 PDA dimensions could not be obtained with TTE in 1 dog and with TEE in 2 dogs because of poor image quality. Median ductal ampulla width was 8.1 mm (range, 2.7–17.1 mm), 6.4 mm (range, 4.0–17.1 mm), and 6.5 mm (range, 2.7–17.3 mm) with angiography, TTE, and TEE, respectively. Median ductal length was 15.8 mm (range, 9.1–38.3 mm) and 14.4 mm (range, 6.5–26.4 mm) with angiography and TEE, respectively. Median minimal ductal diameter was 2.9 mm (range, 1.1–5.7 mm), 3.1 mm (range, 1.4–10.7 mm), and 2.6 mm (range, 1.0–7.0 mm) with angiography, TTE, and TEE, respectively. Bland-Altman analysis comparing minimal ductal diameter acquired with TTE and TEE to angiography documented a mean bias of 0.92 and −0.02 mm, respectively (Fig 4). The 95% limits of agreement were narrower with TEE (−0.57 to 0.52) than with TTE (0.23–1.60). An ACDO was released in 21 dogs and a coil in 1 dog. Coil embolization was selected for 1 of the smallest dogs (3.4 kg) because femoral artery size prevented placement of an appropriately sized sheath required for ACDO device delivery. The median ACDO size was 6 mm (range, 3–14 mm), and median ACDO proximal disc diameter was 11.5 mm (range, 8–22 mm). Three ACDO-to-minimal ductal diameter ratios were calculated. The first 2 utilized the minimal ductal diameter derived from the angiogram and TEE. Final ACDO size selection was based on either angiogram or TEE minimal ductal diameter measurements, depending on which was considered more accurate by the members of the cardiology team performing the transcatheter occlusion procedure. The 3rd ratio reflected the actual minimal ductal diameter used to select the ACDO. Mean ACDO-to-angiographic minimal ductal diameter ratio was 2.0 ± 0.5 (range, 0.9–3.0). Mean ACDO-to-TEE minimal ductal diameter ratio was 2.0 ± 0.5 (range, 1.1–2.8). Mean ACDO-to-minimal ductal diameter (either angiographic or TEE based on final ACDO device selection) was 2.0 ± 0.4 (range, 1.4–3.0). Median ACDO proximal disc-to-angiographic ampulla width ratio was 1.4 (range, 0.8–3.0), and median ACDO proximal disc-to-TEE ampulla width ratio was 1.8 (range, 1.0–3.0). In the 21 dogs in which it was recorded, the median amount of contrast administered during the procedure was 1.3 mL/kg (range, 0.5–3.5 mL/kg). The 6 dogs with the largest volume administered (>2 mL/kg) either did not have TEE available during the procedure which required additional contrast injections to confirm catheter location and device placement (n = 3), required an additional angiogram after retraction of the TEE probe (n = 1) (Fig 5), or had multiple left ventricular or aortic root injections to evaluate concurrent subaortic stenosis or ventricular septal defect (n = 2). The median fluoroscopy time was 5.3 minutes (range, 3.4–16.3 minutes). After device release, residual ductal flow was documented in 9 dogs with angiography (3 trivial, 3 mild, 3 moderate) and in 9 dogs with color Doppler TEE (6 trivial, 2 mild, 1 moderate). Five dogs had similar severity of residual flow documented with both angiography and TEE. Three dogs had more severe residual flow documented with angiography and 2 dogs had trivial residual flow detected with TEE and not with angiography. One dog had trivial residual flow with angiography and did not have TEE assessment of residual flow performed because of poor image quality. The next morning, complete occlusion was documented with color Doppler TTE in all of the dogs.
Based on intraoperative TEE assessment of PDA morphology and dimensions, the cardiology team members performing the transcatheter occlusion procedure elected to change the device size in 3 dogs (increase in size in 2 dogs from 6 to 7 mm and 4 to 5 mm and decrease in size in 1 dog from 7 to 4 mm). Multiplane TEE also provides visualization of the PDA ampulla in transverse section frequently documenting elliptical rather than circular short-axis morphology (Fig 6). The authors frequently consider ampulla width when selecting ACDO size to ensure the proximal disc is not excessively larger or smaller than the ampulla width in long (angiographic or TEE) or short (TEE) axes. When compared with TEE measurements, angiographic measurements were believed to underestimate the true minimal ductal diameter and ampulla width in 2 dogs and overestimate true minimal ductal diameter in 1 dog. The device type selected initially was not changed in any dog. TEE was performed throughout the occlusion procedure in 17/21 dogs to provide visual confirmation that the distal disc of the ACDO was securely positioned in the main pulmonary artery and the proximal disc was within the ductal ampulla before release (Fig 7) and to assess the presence and severity of residual ductal flow. In the dog with type I PDA morphology, the smallest available ACDO device was deployed and did not assume its proper shape but could not be withdrawn into the sheath (Fig 8A). TEE was crucial in verifying that the device did not extend into the aorta (Fig 8B). Complete occlusion was confirmed via TEE, and the device was released. TEE was not performed during the procedure for device positioning and deployment due to a limited number of qualified personnel available in the catheterization laboratory to perform the study (n = 2), poor image quality (n = 1), and because the TEE probe obstructed visualization of the PDA during fluoroscopic imaging (n = 1). When TEE could not be utilized throughout the procedure, additional contrast injections were administered to confirm catheter location and secure device deployment and to assess residual ductal flow both before and after device release. No complications were recognized as occurring directly related to the use of TEE. A procedural complication occurred in 1 dog when an ACDO device was deployed in the pulmonary artery and immediately embolized to a branch of the right pulmonary artery. The authors suspect the device became unscrewed in the delivery sheath. Although the embolization appeared to occur due to premature release and not inappropriate sizing, a larger device was chosen for the 2nd occlusion attempt based on the TEE measurements. The dog recovered uneventfully and a pulmonary perfusion scan was performed within 24 hours that documented a wedge-shaped region of decreased radiopharmaceutical uptake in the caudodorsal aspect of the right caudal lung lobe corresponding with the location of the embolized ACDO device.
Successful transcatheter ductal occlusion necessitates acquisition of accurate ductal dimensions to select the most appropriate device type and size to achieve complete closure and minimize the risk of systemic or pulmonary embolization. Angiography is the standard method used to determine PDA morphology and dimensions. Angiographic images typically are recorded in a single plane, which does not always provide optimal anatomic spatial representation, a potential limitation that can be avoided if the PDA is evaluated in multiple planes. Intraoperative multiplane TEE is widely used for evaluating PDA anatomy in humans providing useful information regarding ductal morphology, size, device deployment, device release, and the presence and severity of any residual flow.11,12 When angiography was compared with TTE and TEE in a prospective evaluation of dogs with PDA, TEE more accurately estimated angiographic ductal dimensions.9 Additionally, TEE provides useful intraprocedural information.9,13 The ductal ampulla can be assessed with TEE where it frequently is identified as having an elliptical rather than circular shape in short axis which may influence device sizing.
TEE can provide unique real-time visualization of device deployment within the PDA and assessment of residual flow, which can facilitate optimal intraoperative decision making. The potential benefit of contemporaneous procedural TEE is great considering the wide individual variation in PDA morphology and size and the availability of multiple devices for PDA occlusion. The number of contrast injections required to confirm secure deployment and assess closure can be reduced or omitted when TEE is utilized during transcatheter closure of PDA in humans.11 In dogs, additional contrast injections are required to confirm secure device deployment before release and to assess closure,4 a finding that was confirmed in our study when TEE was unavailable. Reduction in the duration of fluoroscopic time required to perform the procedure leads to decreased radiation exposure to the operators and the patient. A reduction in cumulative radiation exposure to operators and staff is an important aspect of radiation safety and is beneficial in reducing the risks of long-term exposure.15 Median fluoroscopy times reported at our institution when using coils (n = 14), vascular plug (n = 31), and ACDO (n = 40) were 14.5 minutes (range, 4.5–46.1 minutes), 8.7 minutes (3.4–41.5 minutes), and 5.3 minutes (2.6–25.7 minutes), respectively.3,5,9 Procedural fluoroscopy time has decreased with the use of self-expanding devices such as the ACDO but may be further reduced with the use of TEE.
Based on TEE assessment in our study, the initial ACDO size selection was changed in 3 dogs. In 2 dogs, the size was increased based on TEE measurements of minimal ductal diameter and because the proximal disc of the ACDO more closely approximated the ampulla width measurement. In 1 dog, the size was decreased (7–4 mm) based on better visualization of the PDA with TEE compared with angiography. Whether this change in size resulted in improved outcome is impossible to determine. Regardless, the complete occlusion rate in this study was 100% based on TTE. Nguyenba and colleagues reported an ACDO-to-angiographic minimal ductal diameter (oversize factor) range of 1.5–2.7 in 18 dogs.4 The ACDO-to-angiographic or TEE minimal ductal diameter ratios in this study had values that exceeded the range, but the ratios were not representative of the final ACDO device size selected for each dog. The range for ACDO diameter-to-minimal ductal diameter using the most appropriate minimal ductal diameter value from either angiography or TEE was 1.4–3.0. The oversize factor was >2.7 in the dog with the type I ductal morphology in this report. The morphology was not considered amenable to coil embolization and the smallest available ACDO (3 mm) was deployed and released.
After device release, a reduction in flow or complete resolution of flow through the PDA typically occurs. TEE frequently documents a progressive reduction in flow with time. After deployment but before release, when the device is attached to the delivery cable, the nitinol wire mesh may be distorted, resulting in a small amount of residual ductal flow that may resolve when the device is released.8 Substantial residual flow before device release may indicate improper device size or deployment resulting in reassessment leading to repositioning or device exchange to ensure secure deployment before device release. Although residual flow was documented with angiography and TEE after device release in this series of dogs, it was not estimated as severe in any dog and TTE the day after the procedure was used to assess and confirm successful closure. In 2 dogs, TEE was more sensitive than angiography for detecting trivial residual flow after device release.
Limitations to TEE include probe cost and the requirement of a skilled operator throughout the procedure. Occasionally, the TEE probe had to be intermittently retracted or removed because it obstructed the angiographic view of the PDA. A limitation that was not encountered in this study but that has occurred in the authors' experience is that occasionally the TEE probe physically displaces the aorta and PDA, making angiographic evaluation of ductal morphology impossible and necessitating probe removal (Fig 9). Reported complications associated with TEE in humans are uncommon and include thermal esophageal burns, gastrointestinal perforation, bacteremia, and transient bradycardia with introduction of the TEE probe.16,17 In summary, TEE provided useful anatomic information regarding PDA morphology and closely approximated angiographic ductal dimensions while aiding in device deployment, release, and assessment of ductal closure. In some instances, TEE findings can independently influence device sizing. We conclude that TEE provides complementary anatomical and intraprocedural information and is well tolerated in dogs.
a GE Vingmed Ultrasound, Horten, Norway
b 4F, 5F, 6F, 8F Check-Flo Performer Introducer Set, Cook Inc
c Glidewire Stiff type, Terumo Medical Corp, Somerset, NJ
d 6F Pediatric Introducer, Infiniti Medical, Malibu, CA
e 8F, 11F Introducer, Boston Scientific Corp, Natick, MA
f 6T Probe, GE Medical Systems, Horten, Norway
g Oxilan, Cook Inc, Bloomington, IN
h 4F Marker Diagnostic Catheter, Infiniti Medical
i Amplatz Canine Duct Occluder, Infiniti Medical
j Cook Inc, Bloomington, IN
k SPSS 15, SPSS, Chicago, IL
The authors thank Dr Geoffery T. Fosgate for support with statistical analysis; Kathy Glaze and Katy Waddell for their technical assistance; and Jennie Lamb for assistance with figure preparation.
- 2Catheter closure of patent ductus arteriosus in dogs: Variation in ductal size requires different techniques. J Vet Cardiol 2003;5:7–12., , , et al.
- 8The Amplatz canine duct occluder: A novel device for patent ductus arteriosus occlusion. J Vet Cardiol 2007;9:109–117.,
- 15Proper use of cineangiographic equipment and contrast agents. In: BaimDS, GrossmanW, eds. Grossman's Cardiac Catheterization, Angiography, and Intervention, 6th ed. Philadelphia, PA: Lippincott Williams & Wilkins; 2000:15–34.