Journal of Veterinary Cardiology (2006) 8, 109e114

www.elsevier.com/locate/jvc

Angiographic classification of patent ductus

arteriosus morphology in the dog*
Matthew W. Miller, DVM, Dipl ACVIM a,*, Sonya G. Gordon, DVM,
Dipl ACVIM a, Ashley B. Saunders, DVM, Dipl ACVIM a, Wendy G.
Arsenault, DVM, Dipl ACVIM a,y, Kathryn M. Meurs, DVM, Dipl ACVIM
Linda B. Lehmkuhl, DVM, Dipl ACVIM b,x, John D. Bonagura, DVM,
Dipl ACVIM b, Philip R. Fox, DVM, Dipl ACVIM, Dipl ECVIM c

a,z

Department of Small Animal Medicine and Surgery, and The Michael E. DeBakey Institute of
Comparative Cardiovascular Sciences in Biomedical Devices, College of Veterinary
Medicine and Biomedical Sciences, Texas A&M University, College Station, TX, USA
b
Department of Clinical Sciences, College of Veterinary Medicine, The Ohio State University,
Columbus, OH, USA
c
Department of Medicine, The Bobst Hospital, and Caspary Research Institute of
the Animal Medical Center, New York, USA
Received 5 June 2006; received in revised form 5 July 2006; accepted 10 July 2006

KEYWORDS
PDA;
Ductus;
Angiography;
Morphology;
Dog

Abstract Objectives: To characterize angiographic morphology and minimum internal transverse diameter of left-to-right shunting patent ductus arteriosus
(PDA) in a large series of dogs.
Background: PDA is the most common congenital cardiac malformation in the dog.
Transarterial ductal occlusion is increasingly performed to close this defect. While accurate assessment of ductal morphology and luminal diameter is important to assure

A unique aspect of the Journal of Veterinary Cardiology is the emphasis of additional web-based images permitting the detailing of
procedures and diagnostics that previously were limited with still figures. These images can be viewed (by those readers with subscription access) by going to http://www.sciencedirect.com/science/journal/17602734. The issue to be viewed is clicked and the
available PDF and image downloading is available via the Summary Plus link. The supplementary material for a given article appears at the end of the page. Downloading the videos may take several minutes. Another means to view the material is to go to
www.doi.org and enter the below doi number unique to this paper.
* Corresponding author.
E-mail address: mmiller@cvm.tamu.edu (M.W. Miller).
y
Dr. Arsenault is currently affiliated with Southwest Florida Veterinary Cardiology, Bonita Springs, FL, USA.
z
Dr Meurs is currently affiliated with the Department of Veterinary Clinical Sciences, College of Veterinary Medicine, Washington
State University, Pullman, WA, USA.
x
Dr Lehmkuhl is currently affiliated with MedVet Associates, Worthington, OH, USA.

1760-2734/$ - see front matter 2006 Elsevier B.V. All rights reserved.
doi:10.1016/j.jvc.2006.07.001

110

M.W. Miller et al.

optimal occlusion using catheter-delivered devices, such information is currently
limited.
Animals, materials and methods: In 246 dogs representing 31 breeds with left-to-right
shunting PDA, right lateral selective aortic angiograms were recorded and reviewed.
Results: PDA morphology conformed to four general phenotypes (types I, IIA, IIB, and
III) which varied according to degree of ductal tapering, and the presence, absence, or
location of abrupt ductal narrowing. Minimum internal ductal diameter for all dogs
averaged 2.9 mm (median, 2.5 mm; range, 1.0e9.5 mm) and was not correlated to
age or body weight. There was no significant difference in minimum internal diameters between types I, IIA or IIB PDA, whereas, type III PDA was significantly wider
(p 0.024) than other phenotypes. The most frequently-encountered variant (type
IIA) was identified in 54.4% of cases (average minimum internal diameter, 2.3 mm
[median, 2.2 mm; range, 1.0e5.5 mm]).
Conclusions: PDA angiographic morphology was categorized based upon the degree,
presence, or absence of ductal narrowing, and the location of ductal attenuation.
When planning PDA repair, this information should assist planning, selection and deployment of transcatheter occluding devices.
2006 Elsevier B.V. All rights reserved.

Introduction
Patent ductus arteriosus (PDA) is the most common
congenital heart defect in the dog, and more breeds
are now affected compared with breed surveys
conducted 30 years ago.1 Increasingly, transcatheter intravascular techniques offer less invasive
alternatives to thoracic surgery for correction of
this condition.2e8 Importantly, these techniques
rely upon accurate angiographic characterization
of ductal morphology and diameter. This information is required in order to facilitate selection and
optimal placement of occluding devices (e.g., the
Gianturco vascular occlusion coil, GianturcoeGrifka
vascular occlusion device, or Amplatzer vascular occlusion plug).9 Although a detailed anatomical description of angiographic ductal morphology has
been published for PDA in humans,10 angiographic
descriptions of ductal morphology in the dog with
PDA are limited. The purpose of the present study
was to describe and characterize the angiographic
morphology of PDA in dogs, and to develop a morphological classification scheme to facilitate intravascular ductal occlusion.

Animals, materials and methods

Study animals
From March 1993 through January 2006, we
evaluated 246 dogs with left-to-right shunting
PDA that required ductal closure, and who underwent aortic angiography prior to transcatheter
ductal occlusion with either GianturcoeGrifka
vascular occlusion devices, Gianturco vascular

occlusion coilsd or an Amplatzer vascular occlusion

plug.e All studies were performed by cardiology
specialists at Veterinary Medical Teaching Hospitals
(Texas A&M University, University of Missouri, Ohio
State University, or Animal Medical Center, USA).
Thirty-one breeds of dogs were represented in this
study population. The average age was 4.6 months
(median, 4.1 months; range, 2e96 months) and average body weight was 9.3 kg (median, 6.2 kg;
range, 0.9e31.8 kg). There were 189 females (28
neutered) and 57 males (4 neutered).

Angiographic technique
Following routine anesthesia, arterial access was
obtained either percutaneously, or by surgical cutdown of the femoral artery. Following insertion of
a vascular introducer, an angiographic catheter or
transeptal sheath was positioned in the descending
aorta or directly within the PDA. Angiograms were
performed in the right-lateral projection by injection of 0.5e2 mL/kg of radiographic contrast material
(Hexabrix,f
Angio-Conray,g
Omnipaque
h
Renographin, or Oxilani) via power injector with
300e1050 PSI pressure, or by vigorous hand injection. Angiograms were recorded either digitally,
on 110 mm film, on video tape, or by combined techniques. These were carefully reviewed frame
by frame off line by the same investigator (MM). Special attention was directed to ductal morphology,
d

Cook Cardiology, Bloomington, Indiana, USA.

AGA Medical, Golden Valley, Minnesota, USA.
f
Mallinckrodt, Hazelwood, MO, USA.
g
Mallinckrodt, Hazelwood, MO, USA.
h
GE Healthcare, Fairfiled, CT, USA.
i
Guerbet LCC, Bloomington, IN, USA.
e

Angiographic classification of PDA morphology in the dog

and to determine the minimum internal diameter of
the PDA. The absolute minimal internal ductal diameter was established by comparing the angiographic width of the ductus at its most narrow
diameter, to known angiographic measuring
markers that were positioned within the imaging
field. Uncertainties were resolved by consensus.
An angle was measured that represented the degree
of distal ductal taper from its point of aortic origin.
This angle was defined by the intersection of two
lines that were drawn beginning where the ductus
departs from the aorta, and then extended distally
along the anterior and posterior ductal borders.

Statistical analysis
Continuous data was reported as mean  standard
deviation, median, and range. Spearman rank correlation was used to evaluate the relationship
between minimum internal ductal diameter, age
and body weight. KruskaleWallis one-way analysis
of variance on ranks was used to compare
minimum internal ductal diameters between categories of ductal morphology. A multiple comparison
procedure (Dunns method) was used to isolate the
group or groups that differed from the others. Linear regression was used to identify a statistical
relationship between age or body weight and minimal ductal diameter. Statistical analyses were performed using commercial software (SigmaStat 3.1).
A p value <0.05 was considered to be statistically
significant.

Results
In the right lateral view the angiographic position
of the PDA relative to the ventral margin of the
descending aorta and the dorsal margin of the main
pulmonary artery was noted as follows. The ductus
coursed in a cranioventral direction in 214/246
dogs (87%), and in a dorsal to ventral direction
(essentially perpendicular) in 32/246 dogs (13%).
The ductus tapered substantially (>30%) from the
point of the descending aorta towards the main
pulmonary artery in 226 of the 246 dogs (92%). This
conferred to the ductus an overall conical or
funnel shaped appearance. Narrowing was most
prominent at the point where the ductus inserted
into the pulmonary artery. These changes were
observed most clearly at or close to the ventral
border of the trachea. In the remaining 8% of the
dogs there was very little alteration in the ductal
diameter, and its morphology appeared as an unrestricted, tubular configuration.

111

Although there was considerable minor variation of morphology, the angiographic appearance
of each ductus arteriosus could be classified into
one of four phenotypes (Fig. 1). Type I PDA was recorded in 12 dogs (5%), type IIA PDA in 134 (54.5%),
type IIB PDA in 80 (32.5%), and type III PDA in 20
dogs (8%). These types were distinguished by the
following morphological characteristics. A type I
PDA gradually tapered from the aorta towards
the point of pulmonary insertion, with no abrupt
alteration in ductal diameter. The angle formed
by the walls of the PDA was typically less than 15
degrees. A type II PDA was characterized by an
abrupt and dramatic distal narrowing in ductal
diameter of more than 50%. This was subcharacterized into one of two phenotypes based
upon morphology of the proximal portion of the
PDA. In type IIA PDA the proximal portion of the
ductus retained a constant dimension (that is,
the walls remained parallel or nearly so) before
it abruptly narrowed distally at the point of insertion into the pulmonary artery. In contrast, type
IIB PDA morphology displayed a conical ductal
shape with the presence of abrupt distal narrowing
that occurred just before it inserted into the pulmonary artery (Fig. 2). The angle formed by the
type IIB PDA was between 30 and 60 degrees. In
contrast, type III PDA morphology was characterized by a tubular appearance without substantial
(<20%) attenuation in ductal diameter.
The internal minimal ductal diameter is compared for all morphologic PDA types in Fig. 3. For
the entire study population, the minimum ductal diameter averaged 2.9 mm (median, 2.5 mm; range,
1.0e9.5 mm). There was no significant difference
between the minimum internal diameter of type I,
IIA, or IIB (p 0.865) PDA. However, the minimal internal diameter of type III PDA was significantly
greater (p 0.012) than that of all other morphologic types. There was no significant correlation
between minimum internal ductal diameter and
patient age (p 0.671) or body weight (p 0.097).

Discussion
Schneider described angiographic features of PDA
in dogs from Germany9 in the context of human
PDA phenotypes reported by Krichenko.10 Based
upon the series of cases in the present report,
our type IIA PDA morphology was most comparable
to the Krichenko type E,10 our type IIB PDA was
most similar to the Krichenko type A, and our
type III PDA was similar to the Krichenko type C
morphology. Both our data as well as findings
reported by Schneider9 suggest that type IIA

Angiographic classification of PDA morphology in the dog

113

Figure 2 Representative angiogram of a type IIB ductus demonstrating the technique by which the angle of ductal
taper was measured. This angle is defined by the intersection of two lines originating where the ductus departs from
the aorta, and drawn along the anterior and posterior ductal borders.

morphology was the most commonly represented

phenotype of canine PDA (54.5% and 80% respectively), and that, the majority of dogs with patent
ductus arteriosus have angiographic morphology of
either type IIA or IIB.
There were certain dissimilarities between phenotypes detected in the present study and those
reported by Schneider.9 Our series did not identify
dogs with ductal constrictions that occurred at
both the aortic and pulmonic sites (which would
be compatible with Krichenko type D PDA). Moreover, we identified a type III PDA morphology in
8% of our study population. It has been our experience that this morphology is common in the
German shepherd breed in the United States. This
phenotype was not reported by Schneider whose
study population contained a large number of this
breed.9 In addition, we identified five percent of
our cases which had gradual ductal tapering (type
I PDA), whereas this phenotype was not reported
in dogs from Germany reported by Schneider.9
These differences probably represent differences
in the genetic pools of dogs from different
countries.
The morphology and luminal diameter of persistent PDA are critical variables which must be

determined prior to interventional occlusion with

a catheter-delivered device. Human clinical experience suggests that a strategic approach to ductal
occlusion based upon ductal morphology, length to
width ratio and minimum internal diameter decreases the frequency of procedure-related morbidities, such as embolization rate, incidence of left
pulmonary artery stenosis, incomplete ductal closure, and the need for a second procedure.11e13 Veterinary clinical studies also have suggested that
ductal diameter and morphology may influence outcome during transcatheter PDA occlusion.6,14 In
fact, considering the vascular occlusion devices
that are currently available, types I, IIA and IIB PDA
would seem to offer the best morphology for repair
by transarterial methods. In contrast, the type III
PDA displays lack of distal ductal taper, tends to be
large in diameter, and does not provide an area
upon which occluding devices can become engaged.
For these reasons, it has become our practice to
avoid percutaneous ductal closure for dogs with
type III PDA morphology. We generally prefer to
choose embolization coils with diameters that are
at least two times larger than the minimum ductal
diameter. When ductual morphology is type IIA, we
routinely deploy the entire coil within the ampulla

Figure 1 Representative aortic angiograms from dogs with left-to-right shunting PDA highlighting characteristic ductal morphology. To the right of each angiogram is the corresponding line drawing illustrating relevant morphologic features. All angiograms were obtained in the right lateral projection. (I) Type I PDA. Notice that the diameter of the
ductus gradually decreases in size from the aorta to the pulmonary arterial location. (IIA) Type IIA PDA. Notice
that the walls of the ductus are essentially parallel to one another, and that the ductal diameter abruptly decreases
(>50%) at the point of insertion into the pulmonary artery. (IIB) Type IIB PDA. Notice that the diameter of the ductus
decreases markedly from the aortic to the pulmonic side with the most narrow aspect being at the pulmonic insertion.
(III) Type III PDA. Notice that the diameter of the ductus does not change substantially throughout its length, producing a tubular appearance. Videos complementing this image can be downloaded from the supplementary material at
doi:10.1016/j.jvc.2006.07.001.

114

M.W. Miller et al.

The present study evaluated angiographic features of canine PDA and organized these morphologic findings into a simple classification scheme.
This may help investigators optimally select and
deploy transcatheter occlusion devices as well as
compare outcomes with different forms of PDA, or
using different techniques. Additionally, this classification scheme may help to evaluate the relative
efficacy of specific occlusion devices as newer
technologies are developed.

References
Figure 3 Vertical box plots displaying the distribution
of minimum internal ductal diameter and ductal morphology. The boxes display the 25th and 75th percentiles
while the error bars delineate the 10th and 90th percentiles. The line within the box represents the median diameter of the group. All outliers are displayed as
individual data points. The minimum internal diameter
of Type III PDA was significantly greater (p 0.012)
than that of all other ductal phenotypes.

of the ductus itself. Conversely, in patients with type

IIB morphology, we generally deploy the initial loop
of the first coil into the pulmonary artery, and the remaining coil loops are then extruded into the ductal
ampulla. It is also our preference to more commonly
consider using a vascular plug in patients with type
IIA morphology than in a patient with type IIB morphology. Prospective studies need to be performed
in order to assess whether these clinical guidelines
confer actual benefit.
The present study contains several limitations.
We made all morphologic assessments using the
lateral angiogram. Reliance on a single angiographic
position may provide incomplete appreciation of
certain angiographic features. Moreover, ductal dimensions were measured from different image
storage modalities, and this might have influenced
slightly some of the measurements. Furthermore,
three-dimensional morphology is incompletely described by evaluating a single projection of a twodimensional imaging modality. Biplane angiography,
high speed computed axial tomography or magnetic
resonance imaging would provide better anatomic
detail but were not available during this study.
Furthermore, the present study characterized ductal morphology into one of four phenotypes. A
departure from the more detailed Krichenko system. A more complex classification scheme might be
constructed in order to address the subtle morphologic variations within the types of PDA. However,
the clinical utility of such a complex scheme is
unclear.

1. Buchanan JW. Prevalence of cardiovascular disorders. In:

Fox PR, Sisson D, Moise NS, editors. Textbook of canine
and feline cardiology: principles and clinical practice. Philadelphia: WB Saunders; 1999. p. 457e70.
2. Campbell FE, Thomas WP, Miller SJ, et al. Immediate and
late outcomes of transarterial coil occlusion of patent ductus arteriosus in dogs. J Vet Intern Med 2006;20:83e96.
3. Gordon SG, Miller MW. Transarterial coil embolization for
canine patent ductus arteriosus occlusion. Clin Tech Small
Anim Pract 2005;20:196e202.
4. Hogan DF, Green 3rd HW, Gordon S, et al. Transarterial coil
embolization of patent ductus arteriosus in small dogs with
0.025-inch vascular occlusion coils: 10 cases. J Vet Intern
Med 2004;18:325e9.
5. Sisson D. Use of a self-expanding occluding stent for nonsurgical closure of patent ductus arteriosus in dogs. J Am Vet
Med Assoc 2003;223:999e1005.
6. Schneider M, Hildebrandt N, Schweigl T, et al. Transvenous
embolization of small patent ductus arteriosus with single
detachable coils in dogs. J Vet Intern Med 2001;15:222e8.
7. Grifka RG, Miller MW, Frischmeyer KJ, et al. Transcatheter
occlusion of a patent ductus arteriosus in a Newfoundland
puppy using the Gianturco-Grifka vascular occlusion device.
J Vet Intern Med 1996;10:42e4.
8. Snaps FR, Mc Entee K, Saunders JH, et al. Treatment of patent ductus arteriosus by placement of intravascular coils in
a pup. J Am Vet Med Assoc 1995;207:724e5.
9. Schneider M, Schneider I, Hildebrand N, et al. Percutaneous
angiography of patent ductus arteriosus in dogs: techniques, results and implications for intravascular occlusion.
J Vet Cardiol 2003;5:21e7.
10. Krichenko A, Benson LN, Burrows P, et al. Angiographic classification of the isolated, persistently patent ductus arteriosus and implications for percutaneous catheter occlusion.
Am J Cardiol 1989;63:877e80.
11. Galal MO. Advantages and disadvantages of coils for transcatheter closure of patent ductus arteriosus. J Interv Cardiol 2003;16:157e63.
12. Santoro G, Bigazzi MC, Carrozza M, et al. Percutaneous
treatment of moderate-to-large patent ductus arteriosus
with different devices: early and mid-term results. Ital
Heart J 2005;6:396e400.
13. Wang JK, Hwang JJ, Chiang FT, et al. A strategic
approach to transcatheter closure of patent ductus: Gianturco coils for small-to-moderate ductus and Amplatzer
duct occluder for large ductus. Int J Cardiol 2006;106:
10e5.
14. Glaus TM, Martin M, Boller M. Catheter closure of patent
ductus arteriosus in dogs: variation in ductal size requires
different techniques. J Vet Cardiol 2003;5:7e12.