Morphologic characterization of the patent ductus arteriosus in the premature infant and the choice of transcatheter occlusion device

Abstract

Objectives:
The aim of this study was to describe and differentiate the morphology of patent ductus arteriosus (PDA) seen in children born prematurely from other PDA types.

Background:
PDAs are currently classified as types A-E using the Krichenko's classification. Children born prematurely with a PDA morphology that did not fit this classification were described as Type F PDA.

Methods:
A review of 100 consecutive children who underwent transcatheter device closure of PDA was performed. The diameter and length (L) of the PDA and the device diameter (D) were indexed to the descending aorta (DA) diameter.

Results:
Comparison of 26 Type F PDAs was performed against, 29 Type A, 7 Type C and 32 Type E PDAs. Children with Type F PDAs (median 27.5 weeks gestation) were younger during the device occlusion compared with types A, C, and E (median age: 6 vs. 32, 11, and 42 months; P = 0.002). Type F PDAs were longer and larger, requiring a relatively large device for occlusion than types A, C, and E (Mean L/DA: 1.88 vs. 0.9, 1.21, and 0.89, P ≤ 0.01 and Mean D/DA: 1.04 vs. 0.46, 0.87, and 0.34, P ≤0.01). The Amplatzer vascular plug-II (AVP-II) was preferred for occlusion of Type F PDAs (85%; P <0.001).

Conclusions:
Children born prematurely have relatively larger and longer PDAs. These "fetal type PDAs" are best classified separately. We propose to classify them as Type F PDAs to add to types A-E currently in use. The AVP-II was effective in occluding Type F PDAs.

Full text

Morphologic Characterization of the Patent Ductus
Arteriosus in the Premature Infant and the Choice
of Transcatheter Occlusion Device
Ranjit Philip,
1,2
*MD, B. Rush Waller III,
1,2
MD, Vijaykumar Agrawal,
3
MD,
Dena Wright,
1,2
RN, Alejandro Arevalo,
1,2
MD, David Zurakowski,
4
PhD,and
Shyam Sathanandam,
1,2
MD
Objectives: The aim of this study was to describe and differentiate the morphology of
patent ductus arteriosus (PDA) seen in children born prematurely from other PDA
types. Background: PDAs are currently classified as types A-E using the Krichenko’s
classification. Children born prematurely with a PDA morphology that did not fit this
classification were described as Type F PDA. Methods: A review of 100 consecutive
children who underwent transcatheter device closure of PDA was performed. The di-
ameter and length (L) of the PDA and the device diameter (D) were indexed to the de-
scending aorta (DA) diameter. Results: Comparison of 26 Type F PDAs was performed
against, 29 Type A, 7 Type C and 32 Type E PDAs. Children with Type F PDAs (median
27.5 weeks gestation) were younger during the device occlusion compared with types
A, C, and E (median age: 6 vs. 32, 11, and 42 months; P50.002). Type F PDAs were lon-
ger and larger, requiring a relatively large device for occlusion than types A, C, and E
(Mean L/DA: 1.88 vs. 0.9, 1.21, and 0.89, P0.01 and Mean D/DA: 1.04 vs. 0.46, 0.87,
and 0.34, P0.01). The Amplatzer vascular plug-II (AVP-II) was preferred for occlusion
of Type F PDAs (85%; P<0.001). Conclusions: Children born prematurely have relatively
larger and longer PDAs. These “fetal type PDAs” are best classified separately. We
propose to classify them as Type F PDAs to add to types A-E currently in use. The
AVP-II was effective in occluding Type F PDAs. V
C2015 Wiley Periodicals, Inc.
Key words: congenital heart disease; patent ductus arteriosus; pediatric interventions;
preterm infant; prematurity; transcatheter device closure
INTRODUCTION
Patent Ductus Arteriosus (PDA) accounts for 5–10%
of all congenital heart diseases with an incidence of
1 in 2,000 live births in children born at term [1].
Throughout fetal life, the morphology of the ductus
arteriosus remains consistent [2]. However, the ana-
tomic features such as length and diameter of the
PDAs that are found in infants and children who are
born at term are quite variable. Krichenko et al.
described these differences based on findings from lat-
eral aortography, classifying them into five types, A
through E [3]. This classification was proposed when
transcatheter PDA closure was typically reserved for
children and adults, but not in premature neonates.
Children born prematurely often have a PDA. About
half of premature infants with a birth weight
(BW) <1 kg and a third of those with BW <1.5 kg
have a PDA [4]. Historically, treatment for a sympto-
matic PDA in premature infants has been either medi-
1
Division of Pediatric Cardiology, University of Tennessee
Health Science Center, Le Bonheur Children’s Hospital,
Memphis, Tennessee
2
Department of Pediatrics, University of Tennessee Health
Science Center, Le Bonheur Children’s Hospital, Memphis,
Tennessee
3
Department of Radiology, University of Tennessee Health
Science Center, Le Bonheur Children’s Hospital, Memphis,
Tennessee
4
Department of Biostatistics, Harvard Medical School, Boston,
Massachusetts
Conflict of interest: Nothing to report.
*Correspondence to: Ranjit Philip, University of Tennessee Health
Science Center, Le Bonheur Children’s Hospital, 848 Adams
Avenue Memphis, TN 38103. E-mail: rphilip@uthsc.edu
Received 17 July 2015; Revision accepted 3 October 2015
DOI: 10.1002/ccd.26287
Published online 2 November 2015 in Wiley Online Library
(wileyonlinelibrary.com)
V
C2015 Wiley Periodicals, Inc.
Catheterization and Cardiovascular Interventions 87:310–317 (2016)

cal therapy with indomethacin or surgical ligation.
With the advent of smaller occlusion devices and cath-
eters, transcatheter therapy has been extended to treat
PDAs in infants born prematurely, with reports of even
treating patients <1 kg [5–7]. The PDAs in these pre-
mature infants are typically long and tortuous without
significant stenosis, similar to the ductus arteriosus
seen during fetal life. Although they have been classi-
fied as either Types A, C, or E [8,9], the morphology
of the PDAs seen in infants born prematurely does not
fit well into the widely accepted angiographic classifi-
cation. Therefore, we propose to characterize the mor-
phology of these “Fetal type” PDAs. There were two
main objectives to this study. The primary objective
was to compare the morphology of the PDAs found in
children born prematurely with the other PDA types.
The secondary objective was to describe the choice of
transcatheter occlusion device based on the PDA mor-
phology.
MATERIALS AND METHODS
A retrospective review of 100 consecutive children
who underwent transcatheter device closure of PDA at
Le Bonheur Children’s Hospital between June 2012
and Dec 2014 was performed. Only patients with iso-
lated PDAs were included in the study. Children with
congenital heart disease including right aortic arch
with an isolated PDA were excluded. Approval for
this study was obtained from the University of Ten-
nessee Health Science Center institutional review
board. Patient demographics including gestational age
were included. The morphology and dimensions of the
PDA were determined by a lateral aortogram and clas-
sified according to Krichenko et al. For the purpose
of this study, Children born prematurely with a PDA
morphology that did not fit into the Krichenko classi-
fication were grouped in a separate category – Type
F, to add to the current classification of Types A–E
(Fig. 1). The narrowest diameter along the length of
the PDA was designated as the minimal luminal diam-
eter (MLD). The diameters at the aortic (A-diameter)
and the pulmonary artery (PA; P-diameter) ends as
well as the length (L) of the PDA were measured.
The diameter of the device (D-diameter) used for
occlusion was recorded. For the Amplatzer
TM
duct
occluder-I (ADO-I; St. Jude Medical, St. Paul, MN),
the diameter of the device by the aortic end (not the
retention skirt) was used as the device diameter. These
measurements were indexed to the diameter of the de-
scending aorta (DA) just distal to the PDA to account
for the varying patient size. The MLD, A-diameter, P-
diameter, length, and D-diameter indexed to the DA
diameter were labelled MLD/DA, A/DA, P/DA, L/
DA, and D/DA ratios, respectively. In very low birth
infants, in whom arterial access was not obtained,
angiograms were performed with a catheter across
the PDA. As the catheter could have altered the
ductal morphology, measurements were made from
2D-echocardiography. Morphologic characterization
and measurements were cross-referenced between two
interventional cardiologists, an interventional radiolog-
ist, and a pediatric cardiologist all of whom were
blinded to the patient’s gestational age for interob-
server reliability. Disagreements in measurements, and
PDA type between observers were reconciled by mu-
tual discussion before analysis of data. Patients were
grouped according to their PDA type for comparison
of the indexed PDA size. There was no correction for
gestational age in the premature patients. The choice
of device type was compared among the groups. An-
giography following device placement was reviewed
for residual shunt size, left PA (LPA) obstruction, and
aortic obstruction.
Statistical Analysis
Normally distributed, continuous variables were
expressed as a mean standard deviation. Skewed,
continuous variables were expressed as a median and
interquartile range (IQR). Categorical variables were
expressed as a number with a percentage of the total.
Av
2
analysis was used to test for differences in
categorical variables, and Wilcoxon rank sum test was
used for continuous variables. Analysis of variance
(ANOVA) was used to analyze the differences
between group means and their associated device clo-
sure procedure. In addition, the Bonferroni method
was used for correcting for multiple comparisons. A
P-value <0.05 was considered statistically significant.
Kappa statistics was used to calculate interobserver
agreement for PDA types between the 4 observers.
Intraclass correlation coefficient (ICC), a measure of
reliability and consistency between observers was
determined using a two-way mixed effects model.
For statistical analysis, SPSS-22.0.1 for Mac (SPSS,
Chicago, IL) was used.
RESULTS
Charts and angiograms were reviewed on 100 con-
secutive patients who underwent transcatheter PDA
closure; 33 patients were born prematurely. Of the 67
patients who were born at term, the PDAs were classi-
fied as Type A in 27 (40%), Type B in 1 (1.5%), Type
C in 7 (10%), Type D in 3 (4.5%), and Type E in 29
(44%). The PDAs in the 33 premature infants were
classified as type A in 2 (6%), type D in 2 (6%) and
PDA in the Preterm Infant 311
Catheterization and Cardiovascular Interventions DOI 10.1002/ccd.
Published on behalf of The Society for Cardiovascular Angiography and Interventions (SCAI).

type E in 3 (9%). PDA morphology in 26 premature
patients did not fit into the Krichenko classification
and were classified as Type F (26/33, 79%; P<0.001)
as shown in Fig. 2A. The relative incidence of PDA
Types A-F are shown in Fig. 2B and Table I. Although
more female patients underwent PDA closure (61%,
P<0.05), there was no difference in the relative inci-
dence of the PDA types between gender (P¼0.114).
There were no PDAs in children born at term with
morphology considered to be Type F. Therefore, all
patients with Type F PDAs were born prematurely
with a median gestational age of 27.5 weeks (23–34
weeks). Patients with Type F PDAs were younger and
smaller at the time of device closure compared with
patients with other PDA types (Median age 6 vs. 34
months, P¼0.002; Median weight 3.8 vs. 24.7 kg,
P<0.001; Median BSA 0.31 vs. 0.81 m
2
,P<0.001).
Morphologic features were compared between Types
A, C, E, and F. Since Types B and D were found only
in a few patients, comparisons were not performed for
these two types.
Type A vs. Type F Morphology
Type A PDAs were conical in shape with a large
aortic ampulla (A-diameter: 9.90 3.80 mm) and a
Fig. 1. Proposed Classification of PDAs: PDA morphology of
premature children that did not fit the Krichenko et al. classifi-
cation were grouped as Type F. To the left of the descriptive
text is a figure of the different PDA types with their companion
lateral aortograms before and after transcatheter device clo-
sure. To the right of the text, is the concomitant 2D and color
Doppler echocardiogram image of the different PDA types. The
Type F PDAs were relatively larger and longer compared with
other types with a tortuous connection to the PA giving an
appearance of a hockey-stick. [Color figure can be viewed in
the online issue, which is available at wileyonlinelibrary.com.]
312 Philip et al.
Catheterization and Cardiovascular Interventions DOI 10.1002/ccd.
Published on behalf of The Society for Cardiovascular Angiography and Interventions (SCAI).

smaller PA end (P-diameter: 3.90 0.10 mm). How-
ever, the A/DA ratio was larger for Type F PDAs than
Type A PDAs (0.91 0.23 vs. 0.82 0.25 mm;
P¼0.01; Table II). Type F PDAs, did not taper from
the aortic to the PA end (0.91 0.23 to 0.78 0.20;
P¼0.35). Therefore, the P/DA ratio was also signifi-
cantly larger for Type F PDAs than Type A PDAs
(0.78 0.20 vs. 0.25 0.08; P0.001). The MLD for
24 of 29 Type A PDAs (83%) was at the PA end and
just posterior to it in the remainder. The MLD for 20
of 26 Type F PDAs was at the anterior half toward the
PA end, but the MLD was not the P-diameter. The
indexed length of Type F PDAs was longer than Type
A PDAs (L/DA ratio: 1.88 0.40 vs. 0.90 0.37;
P<0.001).
Type C vs. Type F Morphology
Type C PDAs were tubular without any narrowing,
with almost similar A-diameter, P-diameter and MLD
(mean: 6.6 mm, 5.4 and 4 mm, respectively; P¼0.65).
However, Type F PDAs had larger A/DA, P/DA, and
MLD/DA ratios compared with Type C PDAs
(P¼0.01 for all indices, Table II). Furthermore, the
indexed lengths of Type F PDAs were longer than
Type C PDAs (L/DA ratio: 1.88 0.40 vs. 1.21 0.46;
P¼0.01). In further contrast to Type C PDAs, 19 of
26 Type F PDAs (73%) had tortuosity of the segment
closest to the PA end. This morphologic feature typi-
cally showed first a mild cranial angulation followed
by a final caudal turn into the connection with the PA.
The shape resembles that of a “hockey-stick,” which
Fig. 2. Relative incidence of PDA Types: (A) Relative incidence of PDA Types A–F based on
gestational age. B: Overall incidence of PDA Types A–F. [Color figure can be viewed in the
online issue, which is available at wileyonlinelibrary.com.]
TABLE I. Demographics According to the Type of PDA (100 Consecutive Patients)
Variable Type A Type B Type C Type D Type E Type F P value
Patients 29 1 7 5 32 26
Gender 0.114
Female 20 (69%) 1 (100%) 6 (86%) 2 (40%) 14 (44%) 18 (69%)
Male 9 (31%) 0 (0%) 1 (14%) 3 (60%) 18 (56%) 8 (31%)
Gestation <0.001
a
Preterm 2 (7%) 0 (0%) 0 (0%) 2 (40%) 3 (9%) 26 (100%)
Term 27 (93%) 1 (100%) 7 (100%) 3 (60%) 29 (91%) 0 (0%)
Age (months) 0.002
a
Median (IQR) 32 (12–90) 210 (–) 11 (5–24) 35 (22–123) 42 (24–114) 6 (2–21)
Weight (Kg) 0.003
a
Mean SD 21.5 17.6 69 7.7 4.0 23.6 12.8 29.1 19.8 4.2 2.6
BSA (m
2
)0.004
a
Mean SD 0.76 0.45 1.78 0.40 0.13 0.80 0.50 0.87 0.49 0.34 0.08
Device <0.001
a
ADO 29 (100%) 0 (0%) 2 (29%) 0 (0%) 0 (0%) 4 (15%)
AVP 0 (0%) 1 (100%) 5 (71%) 5 (100%) 6 (19%) 22 (85%)
Coil 0 (0%) 0 (0%) 0 (0%) 0 (0%) 26 (81%) 0 (0%)
a
Significant differences between the types of PDAs.
IQR: interquartile range; SD: standard deviation; ASO: Amplatzer
TM
duct occluder; AVP: Amplatzer
TM
vascular plug (II and 4).
PDA in the Preterm Infant 313
Catheterization and Cardiovascular Interventions DOI 10.1002/ccd.
Published on behalf of The Society for Cardiovascular Angiography and Interventions (SCAI).

makes the Type F PDA a morphologically different
type than Type C.
Type E vs. Type F Morphology
Type E PDAs were narrow and elongated in appear-
ance with a highly constricted PA end. The MLD in
29 of 32 Type E PDAs (91%) was at the PA end. The
L/DA ratio of Type F PDAs, however, was longer than
Type E PDAs (1.88 0.40 vs. 0.89 0.36; P<0.001).
Type F PDAs were distinct from Type E PDAs since
the diameters by the aortic and PA ends and the MLD
were significantly larger (P<0.001 for all measure-
ments and ratios).
Interobserver Agreement
The Kappa statistics for interobserver agreement
among all observers for the PDA classification of Type
F was 0.88 and among the interventional cardiologists
was 0.96. The mean ICC reliability (rho) for all meas-
urements among the 2 observers compared with the
interventional cardiologist’s measurements was 0.917
(P<0.001).
Choice of Occlusion Device
The choice of transcatheter device based on PDA
morphology is listed in Table I. The first 4 Type F
PDAs in this series were occluded using the ADO-I.
Since these patients were smaller, and there was a pos-
sibility of the retention disk protruding into the aorta
(Fig. 3A and B), the AVP was preferred in the 22 sub-
sequent Type F PDAs (18 AVP-II and 4 AVP-4). In
Type F PDAs, the diameter of the device relative to
the DA diameter was larger when compared with all
the other PDA types (D/DA ratio P0.01 for all). The
device diameter when compared to the diameter of the
aorta was nearly 1:1 (D/DA ratio ¼1.04 0.18) for
Type F PDAs. On follow-up echocardiogram, there
was no significant (>2 m/s) Doppler flow acceleration
in any of the Type F PDAs. At latest follow-up, there
was one mortality from sepsis in a patient with Type
F PDA, which was unrelated to the procedure or the
device.
PDA Closure in Premature Infants 2.5 Kg
There were 14 premature infants, who were
2.5 Kg with Type F PDAs that had transcatheter de-
vice closures, of which 12 were ventilator dependent.
The Median GA of this subgroup was 27 weeks (23–
32 weeks). The median age and weight at the time of
the procedure were 6 weeks (4–16 weeks) and 1.6 Kg
(1.06–2.5 Kg), respectively. The primary indication for
closure was pulmonary hypertension associated with
chronic lung disease in 9 (65%), and left ventricular
volume overload in the rest. All of them had previ-
ously failed at least one course of medical therapy.
The median fluoroscopy time for these patients was
9.2 min (4–24.5 min) and procedural time was 56 min
(42–117 min) that included a hemodynamic pulmonary
hypertension study. The median total contrast dose was
3 mL/Kg (0–4.5 mL/Kg). Arterial access was obtained
in 50% of these patients to perform a retrograde aorto-
gram. Ultrasound was used for vascular access in all
patients. Patients in whom arterial access was not
obtained, the PDA were closed using transthoracic
echo guidance. All were closed via an antegrade
approach using either the AVP-II (n¼10) or the AVP-
4(n¼4). There were no procedure-related complica-
tions, including no access vessel injuries, and no
obstruction of the LPA or the aorta. Nine of these
patients have since been discharged home without ven-
tilator support.
DISCUSSION
Transcatheter PDA closure is being utilized more
frequently in smaller patients including preterm infants
[6,8–10]. Comorbidities associated with prematurity
including prolonged ventilatory support, bronchopul-
monary dysplasia, pulmonary hypertension, and
TABLE II. Comparison of PDA Size Indexed to DA Diameter for Types A, C, E, and F
ANOVA with Bonferroni
Ratios
Type A
(n¼29)
Type C
(n¼7)
Type E
(n¼32)
Type F
(n¼26)
Type F vs. A
Pvalue
Type F vs. C
Pvalue
Type F vs. E
Pvalue
A/DA 0.82 0.25 0.79 0.15 0.41 0.21 0.91 0.23 0.010
a
0.01
a
<0.001
a
P/DA 0.25 0.08 0.57 0.36 0.10 0.06 0.78 0.20 <0.001
a
0.01
a
<0.001
a
MLD/DA 0.24 0.07 0.42 0.16 0.10 0.04 0.67 0.12 <0.001
a
0.01
a
<0.001
a
L/DA 0.90 0.37 1.21 0.46 0.89 0.36 1.88 0.40 <0.001
a
0.01
a
<0.001
a
D/DA 0.46 0.11 0.87 0.36 0.34 0.08 1.04 0.18 <0.001
a
0.01
a
<0.001
a
A/DA vs. P/DA (PValue) <0.01
a
0.74 <0.01
a
0.35
Data expressed as mean SD.
a
Statistically significant.
314 Philip et al.
Catheterization and Cardiovascular Interventions DOI 10.1002/ccd.
Published on behalf of The Society for Cardiovascular Angiography and Interventions (SCAI).

necrotizing enterocolitis are likely complicated by the
presence of a PDA [11]. Surgical ligation has been
routinely performed in those that fail medical therapy,
yet surgery has been associated with complications
such as pneumothorax, phrenic nerve palsy, vocal cord
paralysis, chylothorax, and scoliosis [12,13]. Transcath-
eter PDA closure may offer faster recovery compared
with surgical ligation [8].
Although authors describing transcatheter closure of
PDAs in preterm infants have typically classified them
as Types A, C, or E [8,9] it is frequently challenging
to select one of these accepted Krichenko labels.
Because these PDAs differ in morphology from other
types, we propose to reclassify these PDAs into a sepa-
rate category: Type F. These PDAs seen exclusively in
premature infants are often tortuous and distinctively
long and wide in relationship to the diameter of the
DA.
The Type F PDA tapers minimally from the aortic
to the PA end which clearly differentiates it from
Types A, B, D, and E. Though the Type F PDA does
have some morphologic similarities to Type C, this
study shows that there are other characteristic features
that differentiate these two types. Type F PDAs are rel-
atively longer than Type C PDAs. There is mild curva-
ture of the segment closest to the PA end making these
PDAs appear tortuous. This morphologic feature typi-
cally showed first a mild cranial angulation followed
by a final caudal turn into the connection with the PA.
The shape resembles that of a hockey-stick unlike the
tubular structure of the Type C PDA.
The shape of the fetal arterial duct has often been
referred to as a hockey-stick on fetal ultrasound [14].
As pregnancy progresses, the fetal arterial duct has
been reported to take a greater curvature [15]. By late
in the third trimester, the arterial duct becomes greatly
curved in majority of the fetuses and assumes a config-
uration that appears as a sharply angled C-shape or an
S-shape [16]. It was our observation that the more pre-
mature the patient at the time of the PDA closure, the
less tortuous it is. Conversely, the tortuosity was more
pronounced when closer to term. Therefore, it is possi-
ble that morphology of the PDA in a premature infant
closely resembles that of its fetal counterpart. Reese
et al have described numerous fascinating aspects of
PDAs in children born prematurely, including the
Fig. 3. Angiograms of Type F PDAs. A: arge tortuous Type F
PDA. Following deployment of an 8–6 Amplatzer Duct
Occluder (B); the device is seen partially obstructing flow
from the transverse aortic arch (white arrowhead). C: Large
Type F PDA resembling a “hockey-stick.” Following deploy-
ment of a 6 mm, AVP- II (D); the device is seen well seated,
completely within the duct with no obstruction of the aortic
arch. Note the relative size of the device to the descending
aorta (D/DA ratio 51). E: Large Type F PDA in a 3 months old
premature infant that was partially occluded with a surgical
clip (arrow). Follow up angiogram (F) after trans-catheter
occlusion with a 4 mm, AVP-4 device. Note that the device
discs are on either side of the clip.
PDA in the Preterm Infant 315
Catheterization and Cardiovascular Interventions DOI 10.1002/ccd.
Published on behalf of The Society for Cardiovascular Angiography and Interventions (SCAI).

property of vasomotion which contributes to the spasm
noted in these patients [17]. Most Type F PDAs, in
this series had failed medical therapy prior to trans-
catheter closure. It is also possible that the medical
therapy made some of these PDAs appear tortuous.
Our hypothesis for the different PDA morphologies
is as follows. The arterial duct appears almost similar
in morphology in all human fetuses before birth, unless
there is an associated congenital heart disease [16]. In
a premature infant, when the duct fails to close, it con-
tinues to resemble the ductal morphology of the fetus.
These “fetal type” (Type F) PDAs are typically long
with mild anterior tortuosity and void of any constric-
tions. When a child is born at term and the ductus arte-
riosus remains patent, its morphology undergoes
certain transformations. When it shortens without con-
stricting at either end, it resembles the “tubular ductus”
(Type C). In this series, children with Type C PDAs
were born at term and were relatively young (median
age 11 months) at the time of device closure. Since
they do not have any stenoses, these children become
symptomatic and present earlier for PDA closure. After
birth, when the fetal duct constricts only at the PA
end, it resemble the “conical ductus” (Type A) with a
wide aortic ampulla. The Type A is the most common
PDA morphology described in many studies. When the
fetal arterial duct, after birth, shortens and constricts
only by the aortic end, it resembles the “window
ductus” (Type B). After birth, if the fetal arterial duct
constricts at both the aortic and the PA ends with a
wide center, it resembles the “saccular ductus” (Type
D). In a child born at term, the ductus arteriosus starts
constricting from the PA end [18] before becoming dif-
fusely narrow. When it fails to close at this stage, and
remains fairly long and narrow with a highly con-
stricted PA end, it resembles the “elongated ductus”
(Type E). Many of these PDAs do not contribute
to a hemodynamically significant shunt, or remain
“silent” [19].
There is fairly a large percentage of Type F PDAs
in this series. This was because of a recent trend of a
high number of referrals from the neonatologists for
transcatheter PDA closure on sick premature infants.
There is data to suggest that early surgical ligation in
the premature neonate does not prevent the develop-
ment of chronic lung disease [20]. Consequently, a
large number of older premature infants with estab-
lished chronic lung disease or pulmonary hypertension
also presented for device closure beyond the neonatal
period adding to this series of Type F PDAs. Though
indexing the PDA size to the patient’s body surface
area yielded comparable results, we decided to repre-
sent the PDA size relative to the DA diameter as it
may have clinical value. For example in Type F PDAs,
the D/DA ratio was 1:1. This information may be
useful for implanters as the DA diameter could help
decide on the appropriate size of the occlusion device.
The ADO-I and the Nit-occlud
V
R
coil system and the
Nit-occlud
V
R
PDA-R (pfm Medical) are best suited for
ducts that have an aortic ampulla. In PDAs without a
significantly large aortic ampulla end, the ADO-I with
a relatively wide retention skirt could potentially cause
obstruction to flow in the DA (Fig. 3A and B). There-
fore, these devices are not suitable for the Type F
PDAs as we have learned from our initial experience.
The AVP-II, with retention disks on either end of the
device that are of the same diameter as the central
occlusion portion makes it a very versatile device. In
this series of 12 PDAs occluded in children 2.5 Kg,
the device stayed almost completely within the PDA
with discs not obstructing the aorta or the LPA in all
(Fig. 3C and D). Therefore, currently, the AVP-II may
be best suited for PDA closure in the premature infant.
The AVP-4 was used in a few premature infants
because of its ability to be delivered via a smaller cath-
eter. In two patients that were not included in this
study who had residual shunts after surgical ligations,
we found it advantageous to use the AVP-4 advanced
through a 4-French catheter across the ligature with the
device discs staying on either side of it (Fig. 3E and
F). The AVP-4 however, is a fairly long device and
could potentially protrude into the aorta or the PA in a
small infant. In addition, its shape may not make it the
right option for Type F PDAs. The ADO-II, though
appears to be an attractive option for the Type F
PDAs, has retention disks that are larger than the cen-
tral occlusive portion (6mm). This feature may make
the terminal ends interfere with aortic and branch PA
blood flow. Hence, we have not opted to use it to
occlude Type F PDAs. The ADO-II AS device that is
not currently available in the United States, features
both shorter lengths (2–6 mm) and retention disks only
slightly larger than the central occlusive portion (1–
1.5 mm larger), which are useful attributes for neonatal
PDA closure. There are reports [21] with encouraging
early results using this device in this setting.
CONCLUSIONS
The morphology of the ductus arteriosus that fails to
close in a child born prematurely resembles its fetal
counterpart. These PDA are relatively large and long
with a tortuous connection to the PA giving an appear-
ance of a hockey-stick. They do not fit into the con-
ventional classification described by Krichenko et al.
These “fetal type” PDAs are therefore best classified
independently in a distinct category of their own. We
propose to classify them as Type F PDAs. Currently,
316 Philip et al.
Catheterization and Cardiovascular Interventions DOI 10.1002/ccd.
Published on behalf of The Society for Cardiovascular Angiography and Interventions (SCAI).

the AVP-II appears to be the device best suited for
trans-catheter occlusion of this PDA type.
REFERENCES
1. Mitchell SC, Korones SB, Berendes HW. Congenital heart dis-
ease in 56,109 births: incidence and natural history. Circulation
1971;43:323–332.
2. Silver MM. The morphology of the human newborn ductus arte-
riosus: a reappraisal of its structure and closure with special ref-
erence to prostaglandin E1 therapy. Hum Pathol 1981;12:1123–
1136.
3. Krichenko A, Benson LN, Burrows P, Moes CA, McLaughlin P,
Freedom RM. Angiographic classification of the isolated persis-
tently patent ductus arteriosus and implication for percutaneous
catheter occlusion. Am J Cardiol 1989;63:877–880.
4. Hamrick SE, Hansmann G. Patent ductus arteriosus of the pre-
term infant. Pediatrics 2010;125:1020–1030.
5. Francis E, Singhi AK, Lakshmivenkateshaiah S, Kumar RK.
Transcatheter occlusion of patent ductus arteriosus in pre-term
infants. JACC Cardiovasc Interv 2010;3:550–555.
6. Roberts P, Adwani S, Archer N, Wilson N. Catheter closure of
the arterial duct in preterm infants. Arch Dis Child Fetal Neona-
tal Ed 2007;92:F248–F250.
7. Betham J, Meur S, Hudsmith L, Archer N, Wilson N. Echocar-
diographically guided catheter closure of arterial ducts in small
preterm infants on the neonatal intensive care unit. Catheter
Cardiovasc Interv 2011;77:40915.
8. Abu Hazeem AA, Gillespie MJ, Thun H, Munson D, Schwartz
MC, Dori Y, Rome JJ, Glatz AC. Percutaneous closure of patent
ductus arteriosus in small infants with significant lung disease
may offer faster recovery of respiratory function when compared
to surgical ligation. Catheter Cardiovasc Interv 2013;82:526–
533.
9. Zahn EM, Nevin P, Simmons C, Garg R. A novel technique for
transcatheter patent ductus arteriosus closure in extremely pre-
term infants using commercially available technology. Catheter
Cardiovasc Interv 2015;85:240–248.
10. Delaney JW, Fletcher SE. Patent ductus arteriosus closure using
the amplatzer vascular plug II for all anatomic variants. Catheter
Cardiovasc Interv 2013;81:820–824.
11. Mouzinho AI, Rosenfeld CR, Risser R. Symptomatic patent duc-
tus arteriosus in very-low-birth-weight infants: 1987-1989. Early
Hum Dev 1991;27:65–77.
12. Noori S. Pros and cons of patent ductus arteriosus ligation: He-
modynamic changes and other morbidities after patent ductus
arteriosus ligation. Semin Perinatol 2012;36:139–145.
13. Bal S, Elshershari H, Celiker R, Celiker A. Thoracic sequels af-
ter thoracotomies in children with congenital cardiac disease.
Cardiol Young 2003;13:264–267.
14. DeVore GR. The prenatal diagnosis of congenital heart disease-a
practical approach for the fetal sonographer. J Clin Ultrasound
1985;13:229–245.
15. Benson CB, Brown DL, Doubilet PM, DiSalvo DN, Laing FC,
Frates MC. Increasing curvature of the normal fetal ductus arte-
riosus with advancing gestational age. Ultrasound Obstet Gyne-
col 1995;5:95–97.
16. Matsui H, McCarthy KP, Ho SY. Morphology of the patent ar-
terial duct: Features relevant to treatment. Images Paediatr Car-
diol 2008;10:27–38.
17. Reese J, Veldman A, Shah L, Vucovich M, Cotton RB. Inadver-
tent relaxation of the ductus arteriosus by pharmacologic agents
that are commonly used in the neonatal period. Semin Perinatol
2010;34:222–230.
18. Schneider DJ, Moore JW. Congenital Heart Disease for the
Adult Cardiologist -Patent Ductus Arteriosus. Circulation 2006;
114:1873–1882.
19. Chehab G, Saliba Z, El-Rassi I. The silent patent ductus arterio-
sus. J Med Liban 2008;56:7–10.
20. Raval MV, Laughon MM, Bose CL, Phillips JD. Patent ductus
arteriosus ligation in premature infants: who really benefits, and
at what cost? J Pediatr Surg 2007;42:69–75.
21. Kenny D, Morgan GJ, Bentham JR, Wilson N, Martin R,
Tometzki A, Oslizlok P, Walsh KP. Early clinical experience
with a modified Amplatzer ductal occluder for transcatheter ar-
terial duct occlusion in infants and small children. Catheter Car-
diovasc Interv 2013;82:534–540.
PDA in the Preterm Infant 317
Catheterization and Cardiovascular Interventions DOI 10.1002/ccd.
Published on behalf of The Society for Cardiovascular Angiography and Interventions (SCAI).