1. No category

PEDIATRICS and NEONATOLOGY
Volume 55
Number 4
August 2014
Contents
Editorials
239
Management of Patent Ductus Arteriosus in Premature Babies: The Art and The Sciences
Jieh-Neng Wang
241
Renal Excretion of Water-soluble Contrast Media after Enema in Neonates and Small Infants
Ji-Nan Sheu
Review Article
243
Epidemiology of Enterovirus 71 Infections in Taiwan
Min-Yuan Chia, Pai-Shan Chiang, Wan-Yu Chung, Shu-Ting Luo, Min-Shi Lee
Original Articles
250
Novel Treatment Criteria for Persistent Ductus Arteriosus in Neonates
Hiroyuki Nagasawa, Daisuke Terazawa, Yoshinori Kohno, Yutaka Yamamoto, Masashi Kondo,
Masami Sugawara, Toshinari Koyama, Ryosuke Miura
256
Renal Excretion of Water-Soluble Contrast Media after Enema in the Neonatal Period
Hee Sun Kim, Bo-Kyung Je, Sang Hoon Cha, Byung Min Choi, Ki Yeol Lee, Seung Hwa Lee
262
Cobedding of Twin Premature Infants: Calming Effects on Pain Responses
Zohreh Badiee, Zohreh Nassiri, Amirmohammad Armanian
269
Sonographic Finding of Persistent Renal Pelvic Wall Thickening in Children
Nai-Chia Fan, You-Lin Tain
275
Emergency Department Visits for Food Allergy in Taiwan: A Retrospective Study
Chan-Fai Chan, Po-Hon Chen, Ching-Feng Huang, Tzee-Chung Wu
282
Changes in Amplitude-Integrated Electroencephalograms in Piglets during Selective Mild HC after HI
Ji-Mei Wang, Guo-Fu Zhang, Wen-Hao Zhou, Ze-Dong Jiang, Xiao-Mei Shao
291
Changes in Outcome and Complication Rates of Very-Low-Birth-Weight Infants in One Tertiary Center in
Southern Taiwan between 2003 and 2010
Shen-Dar Chen, Yung-Chieh Lin, Chin-Li Lu, Solomon Chih-Cheng Chen
297
Histopathological Evaluation of Horse Serum-Induced Immune Complex Vasculitis in Swine: Implication to
Coronary Artery Lesions in Kawasaki Disease
Saji Philip, Wen-Chuan Lee, Mei-Hwan Wu, Cherian Kotturathu Mammen, Hung-Chi Lue
306
Cyclic Pamidronate Infusion for Neonatal-Onset Osteogenesis Imperfecta
Chia-Hsuan Lin, Yin-Hsiu Chien, Shinn-Forng Peng, Wen-Yu Tsai, Yi-Ching Tung, Cheng-Ting Lee,
Chun-Ching Chien, Wuh-Liang Hwu, Ni-Chung Lee
Case Reports
312
Normal Uricemia in Lesch–Nyhan Syndrome and the Association with Pulmonary Embolism in a Young
Child—A Case Report and Literature Review
Jeng-Dau Tsai, Shan-Ming Chen, Chien-Heng Lin, Min-Sho Ku, Teng-Fu Tsao, Ji-Nan Sheu
316
Vertebral Artery Dissection Complicated by Basilar Artery Occlusion
Chia-Yin Kuan, Kun-Long Hung
PEDN_v55_i4_FM.indb 3
11-07-2014 20:39:49
Contents continued from the previous page
Brief Communications
320
Severe Anaphylaxis in Children: A Single-center Experience
Erdem Topal, Arzu Bakirtas, Ozlem Yilmaz, Ilbilge H. Ertoy Karagöl, Mustafa Arga, Mehmet S. Demirsoy,
Ipek Turktas
323
Comparison of Risk for Early Onset Sepsis in Small-for-Gestational-Age Neonates and Appropriate-forGestational-Age Neonates Based on Lower Levels of White Blood Cell, Neutrophil, and Platelet Counts
Nora Hofer, Silvia Edlinger, Bernhard Resch
Letters to the Editor
326
Pulmonary Hemorrhage in Very-Low-Birth-Weight Infants
Bai-Horng Su, Hsiang-Yu Lin, Fu-Kuei Huang, Ming-Luen Tsai
328
Reply: Pulmonary Hemorrhage in Very Low Birth Weight Infants
Ting-An Yen, Po-Nien Tsao
PEDN_v55_i4_FM.indb 4
11-07-2014 20:39:49
Pediatrics and Neonatology (2014) 55, 239e240
Available online at www.sciencedirect.com
ScienceDirect
journal homepage: http://www.pediatr-neonatol.com
EDITORIAL
Management of Patent Ductus Arteriosus in
Premature Babies: The Art and The Sciences
Patent ductus arteriosus (PDA) is the most common cardiovascular abnormality in premature infants, with the
incidence depending on the estimated gestational age.
Although ductus arteriosus is important for prenatal and
immediate postnatal circulation, its persistence beyond the
transitional period is associated with neonatal morbidity
and mortality.1 There is no consensus among neonatologists
on the management of PDA; the reason for this variation is
that current evidence does not mandate one treatment
over any other. Therefore, “to treat or not to treat, when
to treat and how to treat” are still the main questions.2
To answer these questions, the best way is to establish
treatment criteria. Some scholars have used clinical parameters to sum up a cardiovascular dysfunction score,3 and
others have used serum biomarkers such as B-type natriuretic
peptide to predict the course of PDA,4 whereas most researchers have preferred to apply echocardiography evaluation to quantify the impact of significant PDA. There are many
echocardiographic criteria, such as increased left atrium to
ascending aorta diameter ratio, increased end diastolic flow
of the left pulmonary artery, left ventricular output (LVO) to
right ventricular output (RVO) ratio (LVO/RVO), LVO to superior vena cava flow ratio, PDA size or its flow, and PDA/left
pulmonary artery ratio.5 However, the indications for the
treatment of premature very low birth weight infants
(VLBWIs) with PDA have not been well established.
In the current issue of this Journal, Nagasawa et al
report6 their experience using the value of the measured
left ventricular end-diastolic dimension (LVDd), which was
divided by the normal LVDd as an index (LVDd ratio), to
compare 30 patients who underwent PDA ligation to 30
patients treated with indomethacin and 30 patients who did
not undergo radical therapy. They found that the percentage of patients with <115% in the LVDd ratio was 90% in
the no-radical-therapy patients. The LVDd ratios of 130%
and 115% were regarded as cut-off values for surgical ligation and indomethacin treatment. Therefore, the LVDd
ratio can be considered a useful measure to determine the
treatment of VLBWIs with PDA.
Theoretically, it is very reasonable that the LVDd in PDA
patients is commonly increased because of its extended
pulmonary flow, as a consequence of the flow volume
through the ductus arteriosus. Therefore, LVDd measurement could be a useful criterion for determining treatment
in VLBWIs with PDA. However, due to different gestational
ages and body weights, the size of LVDd varies. Therefore,
the authors needed to establish a normal value for LVDd
and used the LVDd ratio to set up the criteria.7 Before we
adopt these novel criteria, there are some concerns with
regard to applying the results of this study. The first one
would be that this study was a retrospective design. Even if
the inter- or intraobserver error was low, selection bias
might still exist. As mentioned in the article,6 the treatment choice did not depend on the LVDd ratio, the sex and
appropriate for gestational age/large for gestational age
distributions were not similar between groups, and differences existed on the examination day. The second concern
is that the authors collected the cases over a 13-year period
(from February 1999 to February 2012); therefore, it is
necessary to convince others that technical or advanced
concepts did not change the clinical neonatal care practice
over the decade.
Although these minor concerns exist, the authors provide many solid data to explain the results. In the future,
we believe that a multicenter study, or a prospective
randomized study using the LVDd ratio, may help to prove
that ratio as a useful measure to determine the treatment
of VLBWIs with PDA.
Conflicts of interest
The author declares no conflicts of interest.
Jieh-Neng Wang*
Department of Pediatrics, National Cheng Kung University
Hospital, Tainan, Taiwan
* Department of Pediatrics, National Cheng Kung University
Hospital, 138 Sheng Li Road, Tainan 70421, Taiwan.
E-mail address: [email protected]
http://dx.doi.org/10.1016/j.pedneo.2014.04.003
1875-9572/Copyright ª 2014, Taiwan Pediatric Association. Published by Elsevier Taiwan LLC. All rights reserved.
Apr 18, 2014
240
References
1. Hermes-DeSantis ER, Clyman RI. Patent ductus arteriosus:
pathophysiology and management. J Perinatol 2006;26:S14e8.
2. Abdel-Hady H, Nasef N, Shabaan AE, Nour I. Patent ductus
arteriosus in preterm infants: do we have the right answers?
Biomed Res Int 2013;2013:676192.
3. Yeh TF, Raval D, Luken J, Thalji A, Lilien L, Pildes RS. Clinical
evaluation of premature infants with patent ductus arteriosus:
a scoring system with echocardiogram, acid-base, and blood
gas correlations. Crit Care Med 1981;9:655e7.
4. Hsu JH, Yang SN, Chen HL, Tseng HI, Dai ZK, Wu JR. B-type
natriuretic peptide predicts responses to indomethacin in
Editorial
premature neonates with patent ductus arteriosus. J Pediatr
2010;157:79e84.
5. Harling S, Hansen-Pupp I, Baigi A, Pesonen E. Echocardiographic prediction of patent ductus arteriosus in need of
therapeutic intervention. Acta Paediatr 2011;100:231e5.
6. Nagasawa H, Terazawa D, Kohno Y, Yamamoto Y, Kondo M,
Sugawara M, et al. Novel treatment criteria for persistent
ductus arteriosus in neonates. Pediatr Neonatol 2014;55:
250e5.
7. Nagasawa H. Novel regression equations of left ventricular
dimensions in infants less than 1 year of age and premature
neonates obtained from echocardiographic examination.
Cardiol Young 2010;20:526e31.
Pediatrics and Neonatology (2014) 55, 241e242
Available online at www.sciencedirect.com
ScienceDirect
journal homepage: http://www.pediatr-neonatol.com
EDITORIAL
Renal Excretion of Water-soluble Contrast
Media After Enema in Neonates and Small
Infants
Gastrografin is an ionic, hyperosmolar, and water-soluble
contrast for radiologic examination of the gastrointestinal
tract. It is administered either orally or rectally in cases of
suspected intestinal perforation or in the early postoperative
evaluation following gastrointestinal surgery. It is recognized
that approximately 3% of iodinated orally ingested contrasts
may be absorbed into portal venous circulation of the intact
gastrointestinal tract.1,2 Absorption occurs through paracellular and transcellular routes.3 Once absorbed, the iodine
undergoes the same physiological pathway as intravenously
administered contrast, including renal excretion.
Previous reports showed that the presence of ingested
water-soluble contrast media in the urinary tract, detected
on abdominal radiographs, was regarded as a strongly suggestive of finding of gastrointestinal perforation or anastomotic leakage after gastrointestinal surgery.4 However, since
the introduction of computed tomography, some studies
have indicated that the urinary excretion of orally administered water-soluble contrast media could be easily observed
on computed tomography in the absence of perforation in
various bowel diseases involving the small bowel mucosa like
inflammatory bowel disease, ischemia, and radiation enteritis.2,5 Computed tomography is a more sensitive imaging
modality for identifying the presence of urinary excretion of
contrast media; however, if performed, computed tomography has a limitation for neonates and small infants due to
the attendant sedation hazard and radiation exposure.
The mechanism of the increased urinary excretion of
orally or rectally ingested Gastrografin in patients with
underlying bowel diseases but without bowel perforation
was due to the diminished speed of transit of contrast
media through the bowel and the increased contrast media’s dwell time in the bowel lumen, which resulted in
increased absorption of contrast and thus a greater concentration in the urine of this absorbed contrast.
In this issue, Kim et al6 present an experience on the
investigation of the role of renal excretion of water-soluble
contrast materials after enema on the abdominal distension
of neonates and small infants. The authors have demonstrated that the patients with severe gastrointestinal disorders causing bowel obstruction may increase watersoluble contrast media’s dwell time in the bowel lumen
and increase absorption of contrast, and thus increase
urinary excretion of this absorbed contrast. They found
that their patients in the renal excretion group showed no
free air in the abdomen cross table lateral views. Therefore, they assumed that the urinary excretion of Gastrografin in the renal excretion group was mainly due to
increased absorption of contrast via the intestinal wall,
rather than to leakage via bowel perforation.
The authors have demonstrated a positive correlation
between severe gastrointestinal diseases and increased
urinary renal excretion of water-soluble contrast media in
the neonatal period. The findings may extend the clinical
implications of water-soluble contrast media on future
management for neonatal bowel disease presenting with
abdominal distension. However, Sohn et al7 reported renal
excretion of Gastrografin in 21% of those who had normal
healthy bowel, suggesting that such phenomenon is not rare
in healthy adults. Therefore, the renal excretion of orally
or rectally ingested contrast media is strongly suggestive
but not pathognomonic of bowel wall disease or perforation. This should emphasize the importance of clinical
correlation and careful interpretation in determining the
significance of this radiological finding in neonates and
small infants. Some limitations exist in this study that may
affect the final result and conclusion should be addressed
and discussed. Firstly, the enrolled case numbers of the
study were too small with the result that they may be
insufficient to draw a firm conclusion. Secondly, patients
with associated underlying disease such as severe infection,
sepsis, or complications of prematurity that could disturb
or slow the bowel mobility were not excluded, which may
be confounding factors. Lastly, the volume of instilled
http://dx.doi.org/10.1016/j.pedneo.2014.04.004
1875-9572/Copyright ª 2014, Taiwan Pediatric Association. Published by Elsevier Taiwan LLC. All rights reserved.
242
Editorial
contrast media varied markedly in each patient. The volume of contrast media was determined by the subjective
feeling of the examiner according to the bowel opacification but it was not dependent on the patient’s body weight
or a rule, which also might lead to a selection bias.
Finally, the authors should be congratulated for their efforts. It is useful regarding the application of contrast media
enema on the management of neonates and small infants
with abdominal distension in practice. In the future, gastrointestinal tract management in cases of gastrointestinal disorder will require more sophisticated and individualized care.
Conflicts of interest
All contributing author declares no conflicts of interest.
Ji-Nan Sheu*
School of Medicine, Chung Shan Medical University,
Taichung, Taiwan
Department of Pediatrics, Chung Shan Medical University
Hospital, Taichung, Taiwan
* No. 110, Section 1, Jianguo North Road, Taichung 402,
Taiwan.
E-mail address: [email protected]
Apr 18, 2014
References
1. Low VH, Chu BK. Diagnostic error due to vicarious excretion
of rectal iodinated contrast. Australas Radiol 2006;50:
369e72.
2. Hay M, Cant PJ. Renal excretion of enteral Gastrografin in the
absence of free intestinal perforation. Clin Radiol 1990;41:
137e8.
3. Andersen R, Tvedt KE, Nordby A, Laeaerum F. Contrast medium
concentration in epithelial mucosal cells after colonic instillation of Iodixanol: a semiquantitive study indicating the route
of permeation. Acad Radiol 2002;9:379e85.
4. Highman JH. Urinary excretion of Gastrografin as a sign of intestinal perforation. Br J Radiol 1964;37:697e700.
5. Apters S, Gayer G, Amitai M, Hertz M. Urinary excretion of
orally ingested gastrografin on CT. Abdom Imaging 1998;23:
297e300.
6. Kim HS, Je BK, Cha SH, Choi BM, Lee KY, Lee SH. Renal
excretion of water-soluble contrast media after enema in the
neonatal period. Pediatr Neonatol 2014;55:256e61.
7. Sohn KM, Lee SY, Kwon OH. Renal excretion of ingested Gastrografin: clinical relevance in early postoperative treatment
of patients who have undergone gastric surgery. AJR Am J
Roentgenol 2002;178:1129e32.
Pediatrics and Neonatology (2014) 55, 243e249
Available online at www.sciencedirect.com
ScienceDirect
journal homepage: http://www.pediatr-neonatol.com
REVIEW ARTICLE
Epidemiology of Enterovirus 71 Infections
in Taiwan
Min-Yuan Chia, Pai-Shan Chiang, Wan-Yu Chung, Shu-Ting Luo,
Min-Shi Lee*
National Institute of Infectious Diseases and Vaccinology, National Health Research Institutes, Taiwan
Received Jul 5, 2013; received in revised form Jul 16, 2013; accepted Jul 17, 2013
Available online 10 October 2013
Key Words
enterovirus 71;
epidemiology;
vaccine
Enterovirus 71 (EV71) was first described in USA in 1969 but retrospective studies in The
Netherlands further detected EV71 in the clinical specimens collected in 1963. EV71 has one
single serotype measured by using hyperimmune animal antisera but can be phylogenetically
classified into three genogroups (A, B, and C) including 11 genotypes (A, B1eB5, C1eC5). In
Taiwan, EV71 caused a large-scale nationwide epidemic in 1998. Retrospective studies further
detected EV71 in clinical specimens collected from handefootemouth disease patients in 1980
and 1986. Therefore, EV71 may have circulated in Taiwan prior to 1980. Since 1998, EV71 has
cyclically caused nationwide epidemics with different predominant genotypes in 1998 (genotype C2), 2000e2001 (B4), 2005 (C4), 2008 (B5), and 2012 (B5). Phylogenetic analysis revealed
that C4 viruses isolated in 2005 were probably from China, B5 viruses isolated in 2008 were
probably from South Eastern Asia, and B5 viruses isolated in 2012 were probably from Xiamen,
China. Several studies have collected postinfection sera from children to measure crossreactive neutralizing antibody titers against different EV71 genotypes and found that antigenic
differences between genogroup B and C viruses did not have a clear pattern but that genotype
A virus was antigenically different from genogroup B and C viruses. In conclusion, EV71 cyclically caused nationwide epidemics through international importations. EV71 surveillance in
Taiwan should combine genetic and serological methods.
Copyright ª 2013, Taiwan Pediatric Association. Published by Elsevier Taiwan LLC. All rights
reserved.
1. Introduction
* Corresponding author. National Institute of Infectious Diseases
and Vaccinology, National Health Research Institutes, Taiwan.
E-mail address: [email protected] (M.-S. Lee).
Enteroviruses (EVs) are single-stranded, positive-sense RNA
viruses in the Picornaviridae family. They cause various
clinical manifestations, including cutaneous, visceral, and
neurological diseases.1 Human EV can be further classified
1875-9572/$36 Copyright ª 2013, Taiwan Pediatric Association. Published by Elsevier Taiwan LLC. All rights reserved.
http://dx.doi.org/10.1016/j.pedneo.2013.07.007
244
into four species including human EV A (20 serotypes such as
EV71 and coxsackievirus A), human EV B (59 serotypes),
human EV C (21 serotypes), and human EV D (4 serotypes).
Overall, human EV has more than 100 serotypes.1,2 For
many years, polioviruses were the most important EVs
because they caused large outbreaks of paralytic disease
before poliovirus vaccines were available. EV71 was first
described in California, USA, in 1969. Since then, EV71 has
been detected worldwide.2e5 Globally, two patterns of
EV71 epidemic have been reported: small-scale outbreaks
with few central nervous system (CNS)-complicated cases
and deaths, and large-scale outbreaks with frequent CNScomplicated cases and deaths. The latter pattern
occurred in Bulgaria (1975), in Hungary (1978), in Malaysia
(1997), in Taiwan (1998), in Singapore (2000), in southern
Vietnam (2005, 2007e2009, and 2011), in Brunei (2006), in
Korea (2009), and in China (2007e2009).2,6 Because EV71
mortality rates are heavily affected by healthcare accessibility and standards, a more relevant clinical definition,
such as CNS complication should be also used to quantify
disease burden of EV71 infections in prevalent areas.
Although EV71 was first isolated in 1969, a retrospective
analysis shows that this virus circulated in The Netherlands
as early as 1963.5 Molecular evolution studies further predicted that EV71 could have emerged in the human population around 1941.5,7 Recently, EV71 repeatedly caused
life-threatening outbreaks of handefootemouth disease
(HFMD) with neurological complications in Asian children.
The neurological manifestations progress very quickly and
range from aseptic meningitis to acute flaccid paralysis and
brainstem encephalitis.2 Due to its tremendous impact on
healthcare systems, development of EV71 vaccines is a
national priority in Taiwan and other Asian countries.
Several vaccine candidates are being evaluated in humans
and some will be licensed in the near future. Epidemiological characteristics are critical to the design of vaccine
trials and formulation of vaccination policy when vaccines
are licensed. Therefore, this review will focus on epidemiology of EV71 infections in Taiwan.
M.-Y. Chia et al
infections with neurological complications and the national
surveillance data would underestimate the proportion of
EV71 infections with neurological complications. In a cohort
study conducted from 2006 to 2012 in northern Taiwan,
about 100 EV71 primary infections were detected serologically in children younger than 5 years and none of them
developed neurological complications (Lee MS et al, unpublished data). Therefore, the proportion of EV71 infections with neurological complications would be between
0.21% and 1%.
3. Laboratory diagnosis of recent EV71
infections
EVs include over 100 serotypes and usually cause selflimited infections with nonspecific symptoms in children,
with the exceptions of polioviruses and EV 71 which
frequently cause neurologic complications. Therefore,
early detection and serotyping of EVs are critical in clinical
management and disease surveillance. Moreover, herpangina and HFMD are major clinical manifestations for EV71
and other EV species A viruses. Therefore, laboratory
diagnosis is critical to differentiate EV71 infections from
other EVs. There are several methods for laboratory diagnosis of recent EV71 infections, including viruses isolation,
molecular methods, and serology (Table 1). The traditional
methods for detection and serotyping of EV infections are
virus isolation and immunofluorescence assay, which are
time-consuming and labor-intensive.11,12 Several clinical
studies have documented that molecular diagnosis based on
polymerase chain reaction saves time and is more sensitive
than virus isolation for the detection of EV infections in
hospitalized patients,13e16 but few studies have been conducted on outpatients. Moreover, no study has compared
molecular tests and virus isolation/immunofluorescence
Table 1 Laboratory diagnosis methods for recent EV71
infections.
Method
2. Clinical spectrum of EV71
According to previous clinical studies conducted in northern
Taiwan, symptomatic EV71 infections progress through four
stages: HFMD/herpangina (Stage 1), CNS involvement
(Stage 2), cardiopulmonary failure (Stage 3), and convalescence (Stage 4).2 This classification was recently recommended in a World Health Organization (WHO) report.6
Follow-up studies further demonstrated that EV71 infection can cause long-term sequelae including neurological
development and cognitive function.8 In a prospective
hospital-based case-finding study, 21% of 183 EV71 infections in children aged < 18 months of age developed
neurological complications such as meningitis and encephalitis.9 Based on national severe EV surveillance and two
cross-sectional serological surveys, Lu et al10 estimated
that 130,617 Taiwanese children aged < 3 years were
infected with EV71 infections in 1998 and that 273 (0.21%)
of these infected children developed neurological complications. Overall, the prospective hospital-based casefinding study would overestimate the proportion of EV71
Advantages
Disadvantages
Virus isolation/IFA Provide virus
Low sensitivity
isolates
Time consuming
for further study Requires skillful
manpower
Nested RT-PCR
High sensitivity
Requires skillful
Time-saving
manpower
False positivity
Multiple primers for
different genotypes
CODEHOP RT-PCR High sensitivity
Requires skillful
Time saving
manpower
Requires sequencing
Serology:
High sensitivity
Requires skillful
neutralizing
High specificity
manpower
antibody
Requires pair sera
Serology: ELISA
Rapid diagnosis
False positivity
IgM
ELISA
Z
enzyme-linked
immunosorbent
assay;
IFA Z immunofluorescence assay; RT-PCR Z reverse transcription polymerase chain reaction.
Epidemiology of enterovirus 71 in Taiwan
assay for serotyping of human EVs using clinical specimens.
Although these methods have been used to detect EVs in
clinical specimens including throat swabs, stool samples,
and cerebrospinal fluid,12e14,16e19 reports elucidating the
comparison among the diagnostic approaches are limited.
Molecular tests for the detection of human EVs in clinical
specimens usually target highly conserved sites in the 50
untranslated region.20 Due to low virus titers in clinical
specimens, several reverse transcription-nested or reverse
transcription-seminested polymerase chain reaction have
been developed to further increase its sensitivity and specificity.21e23 However, serotyping of EVs based on 50 untranslated region sequences directly amplified from clinical
specimens have not been well evaluated. In addition, EV VP1
capsid gene has recently been proposed to be an ideal target
for the detection and serotyping of EVs using the consensus
degenerate hybrid oligonucleotide primer (CODEHOP).24 A
recent study compared virus isolation and the two molecular
tests for detection and serotyping of EVs in clinical samples
and found that the VP1 CODEHOP test performed well for
detection and serotyping of EVs in clinical specimens and
that it could reduce unnecessary hospitalization during EV
seasons.25 Although molecular methods are much more
sensitive than the virus isolation for the detection of EV infections, they require skillful manpower and are not suitable
for clinics and community hospitals.
There are two serological methods available for laboratory diagnosis of recent EV71 infections, including
neutralization assay and enzyme-linked immunosorbent
assay (ELISA) immunoglobulin M (IgM). The neutralization
assay is the most reliable method but it requires collection
of pair sera, which is not feasible for most situations.
Several studies have tried to develop ELISA serum IgM
methods for rapid diagnosis of EV71 infections but these
serum IgM assays all share the drawback of frequent false
positive reactions (>20%) in patients infected with other
human EVs such as CVA6 and CVA16 (Lee MS et al. unpublished data).26e28 Serum IgM tests with low false positive
reactions are desirable to reduce unnecessary hospitalization during EV seasons.
245
Table 2 EV71-related severe and fatal cases in Taiwan,
1998e2012.
Year
Severe
case no.*
Death no.
(case fatality
ratey)
Predominant
genotype
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
405
6
152
181
58
44
20
82
2
7
344
25
12
59
144
78
1
25
27
8
4
5
8
0
2
14
2
0
3
2
C2
B4
B4
B4
B4
B4
C4
C4
C5
C5
B5
B5
C4
C4/B5
B5
(19)
(17)
(16)
(15)
(14)
(9)
(25)
(10)
(29)
(4)
(8)
(5)
(1)
Data are presented as n (%).
Data source: Taiwan Centers for Disease Control.
* Different laboratory methods were used for different years
so these data should be interpreted with caution.
y
Case fatality rate Z number of deaths/number of severe
cases.
EV71 epidemics occurred in 1998 with 405 severe cases
including 78 fatal cases. Nationwide epidemics occurred again
in 2000e2001, 2005, 2008, and 2012 with different predominant genotypes. Overall, clinical spectra of these epidemics
were similar although different genotypes predominantly
circulated in different years.32,33 The laboratory diagnosis
methods only employed virus isolation in early years but were
revised to include serum IgM and molecular methods later.
Therefore, it is not feasible to compare the magnitude of EV71
epidemics based on number of EV71 confirmed severe cases.
However, case fatality rates decreased significantly in recent
years, which may be related to an early warning of EV71 epidemics and stage-based clinical management.6,32
4. EV surveillance in Taiwan
5. Age-specific incidence rates
A nationwide EV71 epidemic occurred in Taiwan in 1998,
which may have caused the largest number of severe cases
and deaths in human history. Since then, a national EV surveillance system has been established by Taiwan Centers for
Disease Control. The national EV surveillance system includes
three components: (1) a sentinel physician network to collect
weekly number of HFMD and herpangina, which was recently
replaced by the Real-time Outbreak and Disease Surveillance
through the National Health Insurance Database;29,30 (2) a
laboratory network for virus identification by collecting
throat swabs from EV-like patients (herpangina and HFMD);
and (3) mandatory notification of EV-like severe cases, which
collects throat swab, serum, and contact information through
an epidemiological investigation. Based on these surveillance
data, nationwide EV71 epidemics have occurred cyclically in
Taiwan since 1998 (Table 2). Retrospective studies further
found that EV71 epidemics may have occurred in Taiwan in
1980 and 1986.31,32 As shown in Table 2, the first recognized
To design clinical trials of EV71 vaccines, age-specific
incidence rates of EV71 infections are required to identify
target populations, estimate disease burdens, define endpoints of clinical efficacy, and calculate the sample size for
efficacy trials. Age-specific incidence rates of EV71-related
severe infections during the 1998 epidemic have been
estimated to be 27.3/100,000, 37.1/100,000, 30.0/100,000,
and 23.1/100,000 for children aged <6 months, 6e11
months, 12e23 months, and 24e35 months, respectively,
which are too low to be a suitable endpoint of efficacy
trials.34 Alternatively, EV71-related mild illness such as
herpangina and HFMD could be suitable clinical endpoints.
In an infant prospective cohort study initiated in 2006 in
northern Taiwan and employing serum neutralization assay
to detect EV71 infections, the age-specific incidence rates
of EV71 infection during the 2008e2009 epidemic were
observed to increase from 1.71/100 personeyears at age
246
M.-Y. Chia et al
0e6 months infants to 4.09/100 personeyears, 5.74/100
personeyears, and 4.97/100 personeyears in children aged
7e12 months, 13e24 months, and 25e36 months, respectively. In addition, the cumulative incidence rate was 15%
by age 36 months, 39% of EV71 infections developed HFMD/
herpangina, and 29% of EV71 infections were asymptomatic
in young children.35 A retrospective serosurvey also found
that 37% of seropositive children reported to develop
HFMD/herpangina during the 1998 nationwide epidemic in
Taiwan.34 Overall, development of EV71 vaccines should
target children age < 6 months, and EV71-related mild
illness such as herpangina and HFMD could be used as the
clinical endpoints of efficacy trials.36
6. Seroprevalence of serum EV71 neutralizing
antibody
In addition to age-specific incidence rates that require
time-consuming prospective cohort studies, seroprevalence
studies of neutralizing antibody can be readily conducted
to estimate cumulative infection rates. Several seroprevalence studies have been completed in Taiwan using sera
collected in 1994, 1997, and 1999 (Table 3). Overall, seropositive rates of EV71 neutralizing antibody in sera
collected in 1994 were 39% in 5-year-old children and 57% in
8-year-old children, which indicates that EV71 was prevalent from 1986 to 1994. Seropositive rates of EV71
neutralizing antibody in children aged <3 years in 1997 in
Taipei City were about 0% (age 0.5e0.9 years) to 15% (age 3
years), which are much lower than in the same age groups
in 1999. In a large-scale cross-sectional serosurvey conducted in 1999 in six areas including two urban (Taipei City
and Kaohsiung City) and four rural areas, seropositive rates
were higher in the preschool children in rural areas than in
the urban areas but similar in school children in both areas,
which indicates more intrafamily transmission affecting
preschool children occurred in rural areas.
Table 3
7. Genetic and antigenic evolutions
According to analysis of VP1 sequences, EV71 was phylogenetically divided into three distinct genogroups: A, B, and
C.1,2 Genogroups B and C can be further divided into genotypes B1eB5 and C1eC5, respectively, and genotype C4 is
further classified into subgenotypes C4a and C4b.36
Recently, genogroups D, E, and F were identified in India.37
Genogroup A includes the EV71 strain (BrCr-CA-70), which
was identified in 1970 in the USA but was not detected again
until 2008. In an investigation of the HFMD outbreak in Anhui,
China in 2008, five EV71 isolates were identified and they
were closely related to genotype A based on analysis of VP1
genes.38 In 2009, genotype A viruses were further detected in
Beijing, Hubei, and Yunan (Figure 1). Reasons for the reemergence of genotype A in China are not clear and should
be clarified. By contrast, genotypes B and C continued to
circulate around the world after the 1970s and the 1980s,
respectively. Recently, retrospective studies have shown
that a genotype B0 virus circulated in The Netherlands as
early as 1963 and a genotype C0 virus circulated in Japan as
early as 1978.4,5 Interestingly, genogroup B viruses seemed
to evolve through positive selection and only genotype B5
viruses are circulating. By contrast, genogroup C viruses
have evolved through neutral selection and multiple genotype C viruses including C2, C4, and C5 are cocirculating
globally.5,7,36 Moreover, EVs frequently recombine to
generate new genotypes and serotypes, which could only be
elucidated by complete genome analysis. Therefore, more
efforts are required to generate complete genome data
especially early EVs isolated before 1980.
In Taiwan, different predominant genotypes were identified in different epidemics (Table 2; Figure 1). Based on
retrospective studies, genotype B1 viruses circulated in
Taiwan from 1980 to 1986. In the 1970s, genotype B1 viruses
circulated widely in Europe, Japan, and USA. Phylogenetic
analysis of VP4 genes found that the genotype B1 viruses
Age-specific seroprevalence of EV71 neutralizing antibody in Taiwan.
ref
Year
Findings
Taipei city10
Taipei city10
Taipei city10
1994
1997
1999
Taipei city34
1997
Taipei city34
1999
Ilan county34
1999
Taoyuan
county34
Taichung
county34
Kaohsiung
city34
Kaohsiung
county34
1999
39% for 5 y, 57% for 8 y, 67% for 10e20 y, 58% for 31e50 y
44% for <6 mo, 0% for 6e11 mo, 5% for 1 y, 16% for 2 y, 15% for 3 y, 38% for 5 y, 47% for 10e20 y
38% for <6 mo, 15% for 6e11 mo, 20% for 1 y, 22% for 2 y, 21% for 3 y, 29% for 4 y, 29% for 5 y,
33% for 6 y, 51% for 7 y, 47% for 8 y, 64% for 9 y, 63% for 10e20 y, 69% for 21e30 y, 69% for 31e50 y,
77% for >50 y
36% for <0.5 y, 4% for 0.5e0.9 y, 4% for 1e1.9 y, 22% for 2e2.9 y, 36% for 3e5.9 y, 63% for 6e11 y,
66% for 12e19 y, 57% for 20e29 y
7% for <0.5 y, 0% for 0.5e0.9 y, 8% for 1e1.9 y, 11% for 2e2.9 y, 34% for 3e5.9 y, 56% for 6e11 y,
54% for 12e19 y, 60% for 20e29 y, 48% for 30e49
8% for <0.5 y, 15% for 0.5e0.9 y, 18% for 1e1.9 y, 15% for 2e2.9 y, 49% for 3e5.9 y, 79% for 6e11 y,
74% for 12e19 y, 78% for 20e29 y, 50% for 30e49 y
13% for <0.5 y, 15% for 0.5e0.9 y, 30% for 1e1.9 y, 36% for 2e2.9 y, 49% for 3e5.9 y, 58% for 6e11 y,
60% for 12e19 y, 55% for 20e29 y, 47% for 30e49 y
12% for <0.5 y, 0% for 0.5e0.9 y, 14% for 1e1.9 y, 30% for 2e2.9 y, 51% for 3e5.9 y, 65% for 6e11 y,
81% for 12e19 y, 73% for 20e29 y, 75% for 30e49 y
10% for <0.5 y, 3% for 0.5e0.9 y, 5% for 1e1.9 y, 15% for 2e2.9 y, 26% for 3e5.9 y, 57% for 6e11 y,
56% for 12e19 y, 58% for 20e29 y, 72% for 30e49 y
24% for <0.5 y, 9% for 0.5e0.9 y, 12% for 1e1.9 y, 25% for 2e2.9 y, 40% for 3e5.9 y, 61% for 6e11 y,
68% for 12e19 y, 63% for 20e29 y, 67% for 30e49 y
Location
1999
1999
1999
Epidemiology of enterovirus 71 in Taiwan
247
100 C4a-KF154310-00184-TW-2012
97
C4a-KF154293-01585-TW-2011
100
C4a-JX986739-Wuhan1143-HuB-CHN-2011
C4a-KF154356-00704-TW-2010
100
C4a-AY905614-ZJ-CHN-1-CHN-2003
C4a-GQ231932-TW-2871-TW-2004
98
C4b-AB465409-763/Toyama-JPN-1997
C4b-EU131776-N3340-TW-2002
93
C4b-JN230523-Xiangyang-Hubei-CHN-2009
99
94
C4b-AF302996-SHZH-CHN-1998
C3-DQ341355-06-KOR-2000
C1-AF135945-2623-AUS-1986
C1-JN874553-4215-TW-1998
99
89
90
C1-FJ868281-20.2.04-TW-2004
C5-AM490161-933V-VNM-2005
84
C5-EF063152-E2005125-TW-2006
C5-KC222965-HCM84-VNM-2011
100
C5-HQ676173-FI11/ns/T/09-Finland-2009
99
C5-EU527983-07364-TW-2007
82
C2L-HM622391-00643-TW-2008
C2-EV71-AB665746-JPN-2010
91
91
C2-AF135949-2644-AUS-1995
C2-AF176044-TW-1998
A-GU434678-Hubei-CHN-2009
100
A-JN408343-10-Yunnan-CHN-2009
A-U22521-BrCr-USA-1970
A-JQ410995-CMU3-1-Beijing-CHN-2009
73
79 A-GQ117124-001-Luan-Anhui-CHN-2008
B2-AF135888-2952-SD-USA-1981
99
86
B2-JQ766157-B27423-CHN-2011
75
B1-AB575913-11977-NED-1971
B1-FJ357380-237-TW-1986
100
100
B3-DQ341367-MY821-3-SAR-MAL-1997
B3-JQ950555-26M/4/99/GuaR1-AUS-1999
B4-HQ283893-3351-TW-1999
72
83
B4-GQ150746-E59-TW-2002
B5-JN316089-PM-11977-MAL-1999
86
B5-EU131740-N2838-TW-2003
76
100
B5-JN874552-141-TW-2008
B5-EU527985-TW-2007
B5-JN964686-Xiamen-CHN-2009
85
B5-KF154308-03939-TW-2011
79
98
B5-KF154353-03419-TW-2012
CVA16-U05876-G-16-RSA-1951
0.05
Figure 1 Phylogenetic analysis of VP1 genes of representative EV71 strains. The phylogenetic tree was generated by the
neighbor-joining method with 1000 bootstraps based on nucleotide sequence from 2481 to 3278 using genome of EV71 (accession
number is U22521) as reference. Only bootstrap values over 70% are shown. Coxsackievirus A16 strain was used as the outlier.
248
isolated in Taiwan in 1980 and 1986 clusters together but
could be differentiated from other genotype B1 viruses,
which indicates that genotype B1 viruses have been introduced to Taiwan for some time.31 Information about circulating genotypes is not available from 1987 to 1997 in Taiwan.
In 1998, Taiwan experienced the largest epidemic caused by
genotype C2. Genotype C2 viruses were widely detected in
the middle 1990s in Australia, Japan, and The Netherlands
and they were likely to have emerged though recombination
between EV71 and CAV8.39 The genotype C2 viruses were
replaced by genotype B4 viruses, which caused nationwide
epidemics in 2000e2001 in Taiwan. Genotype B4 viruses
were widely detected in the late 1990s and could evolve from
genotype B2 viruses. The B4 viruses were replaced by genotype C4a viruses, which caused nationwide epidemics in
2005. Genotype C4 viruses were first detected in Japan and
China in 1997 and 1998, respectively and evolved into two
subgenotypes (C4b and C4a) in 2003 (Figure 1). Since then,
genotype C4a viruses have become predominant and spread
to a lot of countries including Australia, Austria, Japan,
Taiwan, and Vietnam. Genotype C4a viruses could evolve
from recombination between genotype C4b and genogroup B
viruses. The genotype C4a viruses were replaced by genotype C5 viruses, which only spread sporadically in 2006e2007
in Taiwan. Genotype C5 viruses were first detected in Vietnam in 2003 and caused a large scale of epidemic in 2005.40
Genotype B5 viruses emerged in late 2007 in Taiwan and
caused a nationwide epidemic in 2008. Genotype B5 viruses
were first detected in South Asia (Singapore or Malaysia) in
1999 and spread to several Asian countries in the early 2000s
(Figure 1). In 2003, genotype B5 viruses were detected
sporadically in Taiwan. They were not detected again in
Taiwan until 2007. After the 2008e2009 epidemic, genotype
B5 viruses disappeared and only genotype C4a viruses were
detected sporadically in 2010 and 2011 in Taiwan. Genotype
B5 viruses re-emerged in mid-2011 in Taiwan and caused
nationwide outbreaks in 2012. Interestingly, the re-emerging
genotype B5 viruses in 2011e2012 were phylogenetically
closer to B5 viruses circulated in Xiamen, China in 2009 than
to the B5 viruses isolated in Taiwan in 2008e2009 (Figure 1;
Lee MS et al, unpublished data). International spreading of
EV71 is common in Asia and should be monitored through
international networks.
Because EV71 genetic variants have been widely identified
in Taiwan and globally, it is necessary to examine their antigenic variations, which is critical to the selection of vaccine
strains. EV71 has one single serotype as measured by using
hyperimmune animal antiserum but antigenic variations have
been reported recently in human studies. Using sera collected
from young children with a primary infection of genotype B5,
two studies detected partial antigenic differences between
genogroups B and C but not between viruses in the same
genogroup (B5 and B4 viruses).18,33 Kung et al41 did not detect
significant antigenic differences between genotypes B4 and
C4 viruses using acute-phase sera from EV71 inpatients. A
serological survey in healthy Japanese children and adults
detected partial antigenic differences between genotype B5
and A viruses but not among different genotypes in genogroups
B and C that had previously circulated in Japan.42 By constructing an antigenic map, however, Huang et al43 detected
antigenic differences between genogroups B and C, and also
between B5 and B4 viruses. A recent study found that
M.-Y. Chia et al
Taiwanese children infected with genotypes C2, C4, B4, and
B5 had lower GMTs (4-fold difference) against genotype A
than other genotypes but antigenic variations between genogroups B and C did not have a clear pattern.44 It is hard to
compare different studies that employed different human
sera and laboratory procedures, in particular the cell lines
used in the neutralization assay. A network to harmonize
laboratory procedures including standard sera and viruses is
required to make the comparison possible. Moreover, the
clinical and epidemiological significance of the observed
antigenic variation requires longitudinal serological studies
for clarification. In addition, it is not feasible to collect a large
amount of serum from children postinfection to measure
cross-reactive neutralizing antibody titers against multiple
EV71 genotypes. Recently, a rabbit model has been developed
to generate antisera for monitoring antigenic variations of
EV71, which could be integrated to EV surveillance system
(Lee MS, unpublished data).
8. Conclusions
EV71 is highly contagious and causes life-threatening outbreaks in children in Taiwan and several Asian countries. In
the past 10 years, EV71 has caused nationwide epidemics
every 3e4 years in Taiwan. Different genotypes dominate in
different epidemics but the pattern is not predictable in
Taiwan. Case-fatality rates of EV71 infections in recent
epidemics seem to decrease significantly in Taiwan due to
early warning of EV71 epidemics and stage-based clinical
management. Overall, EV71 seems to have evolved rapidly
and spread widely in Asia in the past 15 years. The Taiwanese experience on control and prevention of EV71 would
be valuable to other countries facing EV71 epidemics.
Conflicts of interest
The authors have no conflicts of interest relevant to this
article.
References
1. Solomon T, Lewthwaite P, Perera D, Cardosa MJ, McMinn P,
Ooi MH. Virology, epidemiology, pathogenesis, and control of
enterovirus 71. Lancet Infect Dis 2010;10:778e90.
2. Lee MS, Chang LY. Development of enterovirus 71 vaccines.
Expert Rev Vaccines 2010;9:149e56.
3. Bible JM, Pantelidis P, Chan PK, Tong CY. Genetic evolution of
enterovirus 71: epidemiological and pathological implications.
Rev Med Virol 2007;17:371e9.
4. Chan YF, Sam IC, AbuBakar S. Phylogenetic designation of
enterovirus 71 genotypes and subgenotypes using complete
genome sequences. Infect Genet Evol 2010;10:404e12.
5. van der Sanden S, Koopmans M, Uslu G, van der Avoort H, Dutch
Working Group for Clinical Virology. Epidemiology of enterovirus 71 in the Netherlands, 1963 to 2008. J Clin Microbiol
2009;47:2826e33.
6. World Health Organization. A guide to clinical management
and public health response for hand, foot and mouth disease
(HFMD). Geneva: WHO; 2011.
7. Tee KK, Lam TT, Chan YF, Bible JM, Kamarulzaman A, Tong CY,
et al. Evolutionary genetics of human enterovirus 71: origin,
Epidemiology of enterovirus 71 in Taiwan
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
population dynamics, natural selection, and seasonal periodicity of the vp1 gene. J Virol 2010;84:3339e50.
Chang LY, Huang LM, Gau SS, Wu YY, Hsia SH, Fan TY, et al.
Neurodevelopment and cognition in children after enterovirus
71 infection. N Engl J Med 2007;356:1226e34.
Chang LY, Tsao KC, Hsia SH, Shih SR, Huang CG, Chan WK, et al.
Transmission and clinical features of enterovirus 71 infections
in household contacts in Taiwan. JAMA 2004;291:222e7.
Lu CY, Lee CY, Kao CL, Shao WY, Lee PI, Twu SJ, et al. Incidence and case-fatality rates resulting from the 1998 enterovirus 71 outbreak in Taiwan. J Med Virol 2002;67:217e23.
Lin TL, Li YS, Huang CW, Hsu CC, Wu HS, Tseng TC, et al. Rapid
and highly sensitive Coxsackievirus A indirect immunofluorescence assay typing kit for enterovirus serotyping. J Clin
Microbiol 2008;46:785e8.
Tsao KC, Huang CG, Huang YL, Chen FC, Huang PN, Huang YC,
et al. Epidemiologic features and virus isolation of enteroviruses in northern Taiwan during 2000e2008. J Virol Methods
2010;165:330e2.
Robinson CC, Willis M, Meagher A, Gieseker KE, Rotbart H,
Glode
´ MP. Impact of rapid polymerase chain reaction results on
management of pediatric patients with enteroviral meningitis.
Pediatr Infect Dis J 2002;21:283e6.
Singh S, Chow VT, Phoon MC, Chan KP, Poh CL. Direct detection
of enterovirus 71 (EV71) in clinical specimens from a hand,
foot, and mouth disease outbreak in singapore by reverse
transcription-PCR with universal enterovirus and EV71-specific
primers. J Clin Microbiol 2002;40:2823e7.
Tsao LY, Lin CY, Yu YY, Wang BT. Microchip, reverse
transcription-polymerase chain reaction and culture methods
to detect enterovirus infection in pediatric patients. Pediatr
Int 2006;48:5e10.
Wu Y, Yeo A, Phoon MC, Tan EL, Poh CL, Quak SH, et al. The
largest outbreak of hand; foot and mouth disease in Singapore
in 2008: the role of enterovirus 71 and coxsackievirus A strains.
Int J Infect Dis 2010;14:e1076e81.
Vuorinen T, Vainionpa
¨a
¨ R, Hyypia
¨ T. Five years’ experience of
reverse-transcriptase polymerase chain reaction in daily diagnosis of enterovirus and rhinovirus infections. Clin Infect Dis
2003;37:452e5.
Huang YP, Lin TL, Hsu LC, Chen YJ, Tseng YH, Hsu CC, et al.
Genetic diversity and c2-like subgenogroup strains of enterovirus 71, Taiwan, 2008. Virol J 2010;7:277.
Tryfonos C, Richter J, Koptides D, Yiangou M, Christodoulou CG.
Molecular typing and epidemiology of enteroviruses in Cyprus,
2003e2007. J Med Microbiol 2011;60:1433e40.
Romero JR. Reverse-transcription polymerase chain reaction
detection of the enteroviruses. Arch Pathol Lab Med 1999;123:
1161e9.
Casas I, Palacios GF, Trallero G, Cisterna D, Freire MC,
Tenorio A. Molecular characterization of human enteroviruses
in clinical samples: comparison between VP2, VP1, and RNA
polymerase regions using RT nested PCR assays and direct
sequencing of products. J Med Virol 2001;65:138e48.
Nix WA, Berger MM, Oberste MS, Brooks BR, McKenna-Yasek DM,
Brown Jr RH, et al. Failure to detect enterovirus in the spinal
cord of ALS patients using a sensitive RT-PCR method.
Neurology 2004;62:1372e7.
Thoelen I, Lemey P, Van Der Donck I, Beuselinck K, Lindberg AM,
Van Ranst M. Molecular typing and epidemiology of enteroviruses
identified from an outbreak of aseptic meningitis in Belgium
during the summer of 2000. J Med Virol 2003;70:420e9.
Nix WA, Oberste MS, Pallansch MA. Sensitive, seminested PCR
amplification of VP1 sequences for direct identification of all
enterovirus serotypes from original clinical specimens. J Clin
Microbiol 2006;44:2698e704.
Chiang PS, Huang ML, Luo ST, Lin TY, Tsao KC, Lee MS.
Comparing molecular methods for early detection and
249
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
39.
40.
41.
42.
43.
44.
serotyping of enteroviruses in throat swabs of pediatric patients. PLoS One 2012;7:e48269.
Wang SY, Lin TL, Chen HY, Lin TS. Early and rapid detection of
enterovirus 71 infection by an IgM-capture ELISA. J Virol
Methods 2004;119:37e43.
Tsao KC, Chan EC, Chang LY, Chang PY, Huang CG, Chen YP,
et al. Responses of IgM for enterovirus 71 infection. J Med Virol
2002;68:574e80.
Xu F, Yan Q, Wang H, Niu J, Li L, Zhu F, et al. Performance of
detecting IgM antibodies against enterovirus 71 for early
diagnosis. PLoS One 2010;5:e11388.
Wu TS, Shih FY, Yen MY, Wu JS, Lu SW, Chang KC, et al.
Establishing a nationwide emergency department-based syndromic surveillance system for better public health responses
in Taiwan. BMC Public Health 2008;8:18.
Chuang JH, Huang AS, Huang WT, Liu MT, Chou JH, Chang FY,
et al. Nationwide surveillance of influenza during the
pandemic (2009e10) and post-pandemic (2010e11) periods in
Taiwan. PLoS One 2012;7:e36120.
Lin KH, Hwang KP, Ke GM, Wang CF, Ke LY, Hsu YT, et al. Evolution
of EV71 genogroup in Taiwan from 1998 to 2005: an emerging of
subgenogroup C4 of EV71. J Med Virol 2006;78:254e62.
Chang LY. Enterovirus 71 in Taiwan. Pediatr Neonatol 2008;49:
103e12.
Lee MS, Lin TY, Chiang PS, Li WC, Luo ST, Tsao KC, et al. An
investigation of epidemic enterovirus 71 infection in Taiwan,
2008: clinical, virologic, and serologic features. Pediatr Infect
Dis J 2010;29:1030e4.
Chang LY, King CC, Hsu KH, Ning HC, Tsao KC, Li CC, et al. Risk
factors of enterovirus 71 infection and associated hand, foot,
and mouth disease/herpangina in children during an epidemic
in Taiwan. Pediatrics 2002;109:e88.
Lee MS, Chiang PS, Luo ST, Huang ML, Liou GY, Tsao KC, et al.
Incidence rates of enterovirus 71 infections in young children
during a nationwide epidemic in Taiwan, 2008e09. PLoS Negl
Trop Dis 2012;6:e1476.
Lee MS, Tseng FC, Wang JR, Chi CY, Chong P, Su IJ. Challenges
to licensure of enterovirus 71 vaccines. PLoS Negl Trop Dis
2012;6:e1737.
Rao CD, Yergolkar P, Shankarappa KS. Antigenic diversity of
enteroviruses associated with nonpolio acute flaccid paralysis,
India, 2007e2009. Emerg Infect Dis 2012;18:1833e40.
Yu H, Chen W, Chang H, Tang R, Zhao J, Gan L, et al. Genetic
analysis of the VP1 region of enterovirus 71 reveals the
emergence of genotype A in central China in 2008. Virus Genes
2010;41:1e4.
Huang SW, Kiang D, Smith DJ, Wang JR. Evolution of reemergent virus and its impact on enterovirus 71 epidemics.
Exp Biol Med (Maywood) 2011;236:899e908.
Tu PV, Thao NT, Perera D, Huu TK, Tien NT, Thuong TC, et al.
Epidemiologic and virologic investigation of hand, foot, and mouth
disease, southern Vietnam, 2005. Emerg Infect Dis 2007;13:
1733e41.
Kung SH, Wang SF, Huang CW, Hsu CC, Liu HF, Yang JY. Genetic
and antigenic analyses of enterovirus 71 isolates in Taiwan
during 1998e2005. Clin Microbiol Infect 2007;13:782e7.
Mizuta K, Aoki Y, Suto A, Ootani K, Katsushima N, Itagaki T,
et al. Cross-antigenicity among EV71 strains from different
genogroups isolated in Yamagata, Japan, between 1990 and
2007. Vaccine 2009;27:3153e8.
Huang SW, Hsu YW, Smith DJ, Kiang D, Tsai HP, Lin KH, et al.
Reemergence of enterovirus 71 in 2008 in Taiwan: dynamics of
genetic and antigenic evolution from 1998 to 2008. J Clin
Microbiol 2009;47:3653e62.
Huang ML, Chiang PS, Chia MY, Luo ST, Chang LY, Lin TY, et al.
Cross-reactive neutralizing antibody responses to enterovirus
71 infections in young children: implications for vaccine
development. PLoS Negl Trop Dis 2013;7:e2067.
Pediatrics and Neonatology (2014) 55, 250e255
Available online at www.sciencedirect.com
ScienceDirect
journal homepage: http://www.pediatr-neonatol.com
ORIGINAL ARTICLE
Novel Treatment Criteria for Persistent
Ductus Arteriosus in Neonates
Hiroyuki Nagasawa*, Daisuke Terazawa, Yoshinori Kohno,
Yutaka Yamamoto, Masashi Kondo, Masami Sugawara,
Toshinari Koyama, Ryosuke Miura
Department of Neonatology, Gifu Prefectural General Medical Center, Gifu, Japan
Received Apr 26, 2013; received in revised form Sep 15, 2013; accepted Oct 7, 2013
Available online 25 December 2013
Key Words
left ventricular enddiastolic
dimension;
Qp/Qs;
treatment criteria;
two-dimensional
echocardiography;
very-low-birth-weight
infant
Background: The indications for ductus arteriosus ligation in very-low-birth-weight infants
(VLBWIs) with persistent ductus arteriosus (PDA) are unclear. Increased left ventricular enddiastolic dimension (LVDd) is commonly found in patients with PDA. Here, the enlargement
of LVDd in term and preterm neonates without congenital heart disease was estimated by
two-dimensional echocardiography.
Methods: The value of the measured LVDd was divided by the normal LVDd as an index (LVDd
ratio) to compare 30 patients who underwent PDA ligation with 30 patients treated with indomethacin and 30 patients who did not undergo radical therapy.
Results: An LVDd ratio between 122% and 197% (mean, 142%) was considered to be an indication for the ligation procedure. The proportion of patients exceeding 130% in the LVDd ratio
was 87% (26/30) in those patients who underwent ligation. Catecholamines and/or vasodilators
were required in 83% patients for the treatment of low ejection fraction or hypertension after
operations, suggesting that patients had been in preload and/or afterload remodeling failure
during the operation. The percentage of patients with less than 115% in the LVDd ratio was 90%
in the non-radical-therapy patients. The LVDd ratios of 130% and 115% were regarded as cut-off
values for surgical ligation and indomethacin treatment.
Conclusion: The LVDd ratio is a useful measure to determine the treatment of VLBWIs with
PDA.
Copyright ª 2013, Taiwan Pediatric Association. Published by Elsevier Taiwan LLC. All rights
reserved.
* Corresponding author. Department of Neonatology, Gifu Prefectural General Medical Center, Noisshiki 4-6-1, Gifu 500-8717 Japan.
E-mail address: [email protected] (H. Nagasawa).
1875-9572/$36 Copyright ª 2013, Taiwan Pediatric Association. Published by Elsevier Taiwan LLC. All rights reserved.
http://dx.doi.org/10.1016/j.pedneo.2013.10.009
Novel criteria for neonatal PDA treatment
1. Introduction
The treatment for very-low-birth-weight infants (VLBWIs)
with persistent ductus arteriosus (PDA) is limited to indomethacin administration or surgical ligation of the ductus
arteriosus (DA). Indications for treatment are based on
several criteria applied singly or in combinations, including
increased end-diastolic flow of the left pulmonary artery
(LPA end), increased left atrium to ascending aorta diameter ratio (LA/Ao ratio) on echocardiography, increased
cardiothoracic ratio on X-ray, increased heart rate,
decreased diastolic blood pressure, decreased urinary
output, and increased brain natriuretic peptide value.
However, the indications for the treatment of VLBWIs with
PDA have not been established.
Echocardiography has long been used to assess heart size
in children. In this study, two-dimensional echocardiography with height as an index was used to examine left ventricular end-diastolic dimension (LVDd) and its rate of
increase in neonates, including premature neonates.1,2 In
PDA patients, the flow through the DA increases mitral
valve flow causing increased LVDd. We therefore compared
the LVDd values in PDA patients to those of healthy patients
and used the difference (LVDd ratio) as an index for PDA
severity.
We had three objectives in this study: (1) to determine
the usefulness of the LVDd ratio in patients with PDA
compared to that of LPA-end and LA/Ao ratio; (2) to
determine the cut-off value of the LVDd ratio for the selection of different treatment strategies; and (3) to propose a new treatment algorithm based on the LVDd ratio as
an index in VLBWI with PDA.
2. Materials and methods
2.1. Study design
A retrospective case note review of the treatment of
VLBWIs with PDA was performed. Patients included in the
study were all born and/or treated at the Gifu Prefectural
General Medical Center. The indications for treatment were
analyzed in correlation with the LVDd ratio, LPA-end, and
LA/Ao ratio.
The treatment choice did not depend on the LVDd ratio.
We chose PDA ligation based on two criteria, bleeding in
the head, lungs, or intestine, and the absence of an
effective response to indomethacin treatment.
The patients receiving PDA ligation therapy were
included in the Ligation Group. Patients who underwent
indomethacin therapy were included in the Indomethacin
Group. Patients who were not treated by PDA ligation and
those who were not treated by PDA ligation or indomethacin therapy were included in the Control Group. These two
groups matched the ligation group with respect to sex,
gestational age, and birth body weight. Prophylactic indomethacin treatment was not regarded as a treatment for
PDA. We investigated the relationship between the groups
and LVDd ratio and calculated the mean and standard deviation (SD) of the LVDd ratio, LPA end, and LA/Ao ratio in
each group.
251
The necessary number of neonates included in the study
was determined according to Cohen’s method,3 and that
showed 29 patients are needed to compare two groups with
a Type 1 error of 0.05 in two-sided tests and a statistical
power of 0.85, when the absolute difference divided by SD
is presumed to be 0.80.
2.2. Study patients
The inclusion criteria were as follows: (1) birth weight of
1500 g or less; (2) gestational age less than 30 weeks; (3)
patients with no congenital heart disease and no evidence
of coronary artery lesions as determined by echocardiogram; and (4) patients who survived.
Only VLBWIs who were born from February 1999 to
February 2012, at the Gifu Prefectural General Medical
Center were included in the study.
The characteristics of the patients are shown in Table 1.
The Ligation Group consisted of 30 neonates who underwent PDA ligation. The Indomethacin Group consisted of 30
neonates treated with indomethacin alone; and the Control
Group consisted of 30 neonates who were treated with
palliative therapies but did not receive radical treatments.
All groups were matched by sex, gestational age, and birth
weight.
Table 1
Characteristics of 90 patients.
Groups
Ligation
Indomethacin Control
No. of patients
Sex (male/female)
Gestational age (wk)
Birth weight (g)
AGA/LGA
Birth height (cm)
Age at ligation (d)
Age at exam (d)
Height at exam (cm)
RDS
Surfactant
administered
Ventilation/CPAP
at exam
Inotropes before
treatment
Prophylactic
indomethacin
Bleeding tendencya
30
11/19
26.6 1.7
891 277
26/4
33.9 3.9
20 11
15 11*
34.8 3.8
27
25**
30
11/19
27.1 1.8
969 243
27/3
34.0 3.1
30
12/18
27.0 1.6
958 249
26/4
33.9 3.4
68
34.6 3.3
26
20
34
34.2 3.2
22
13
28/2
24/6
26/4
17***
4
0
6
5
7
14****
5
4
*p Z 0.0004, <0.0001 for Indomethacin Group and Control
Group, respectively.
**p Z 0.039 for Control Group.
***p Z 0.013, <0.0001 for Indomethacin Group and Control
Group, respectively.
****p Z 0.038, 0.015 for Indomethacin Group and Control
Group, respectively.
AGA Z adequate for gestational age; CPAP Z continuous positive airway pressure; LGA Z light for gestational age;
RDS Z respiratory distress syndrome.
a
Bleeding tendency means the patient is vulnerable to bleed
in either the cranium, lung, and intestine.
252
2.3. Data acquisition
H. Nagasawa et al
exceeding 130% in the LVDd ratio were 87% (26/30) and 20%
(6/30) in the Ligation Group and the Indomethacin Group,
respectively. The ratios of patients with less than 115% in
the LVDd ratio were 17% (5/30), and 90% (27/30) in the
Indomethacin Group and the Control Group, respectively.
The two-dimensional echocardiographic studies were performed using the iE33 or Sonos 5500 apparatus (Philips Ultrasound, Bothell, WA, USA) in the left parasternal shortaxis view at the chordae tendinae level immediately below
the mitral valve with a 12e4 MHz transducer or a
7.5e5.5 MHz transducer, respectively, at rest. LVDd was
measured as the distance between the endocardial surface
of the intra-ventricular septum and the left ventricular
posterior wall at the QRS complex onset on the electrocardiogram by M-mode recordings. We also measured PDA
size. Each dimension was measured by the “center
convention”, i.e., using the middle bright contour
excluding the inner and outer areas, with calibrated electronic calipers.1
All measurements were performed by two observers
(HN, DT). The interobserver error was 4.5%, and the intraobserver errors were 3.2% and 4.1%, respectively, calculated on the basis of data from 30 participants. The final
values were obtained from an average of more than three
successive beats.
The distributions of LPA end and LA/Ao ratio in the three
groups are shown in Figure 1. The means and SDs of LPA end
in the Ligation Group, Indomethacin Group, and Control
Group were 25 12, 14 12, and 5 8, respectively. There
was a significant difference (p < 0.001) between any two of
the three groups.
The means and SDs of LA/Ao ratio in the Ligation Group,
Indomethacin Group, and Control Group were 1.82 0.31,
1.40 0.31, and 1.20 0.22, respectively. There was a
significant difference (p < 0.001) between any two of the
three groups. However, the distribution of values in the
latter two categories was wider than that of the LVDd ratio,
and overlapping data were observed in many patients.
2.4. Statistical analyses
3.4. Indications for PDA ligation
The F-test and Student t test were used to determine the
differences in sample variance and means of two categories, respectively. Bonferroni correction was used to
determine statistical equivalence of means in more than
the three groups. The chi square analysis was used to
ensure statistical equivalence of the occurrence rates of
the two categories. A p value < 0.05 was considered statistically significant.
The inotropic treatments after PDA are shown in Table 2.
Catecholamine and/or vasodilator supplementation was
needed by most patients (25/30) because of decreased EF
or systemic hypertension after the operation. This indicates
that approximately 83% of the patients in the Ligation
Group were in preload and/or afterload remodeling failure
after the operation. Moreover, because patients of the
Ligation Group had a bleeding tendency or indomethacin
ineffectiveness, our definitions for the selection of ligation
therapy were almost always appropriate.
3. Results
3.1. Patient characteristics and treatments
The characteristics of the 90 patients are shown in Table 1.
The indication for ligation therapy was either no or
minimal effectiveness of indomethacin in 25 of 30 patients
in the Ligation Group. Indomethacin was not administered
to the remaining five patients because of intracranial
bleeding. The DA was closed by the end of the first
consecutive indomethacin administration in 18 of 30 patients in the Indomethacin Group. Repeat indomethacin
treatment was necessary for the other 12 patients, and the
therapy was not fully effective for two of the 12 patients.
Closed DA was confirmed in the 90 patients before
discharge.
3.2. Distribution of the LVDd ratio in each group
The distribution of the LVDd ratio in each group is shown in
Figure 1. The means and SDs of the LVDd ratio in the
Ligation Group, Indomethacin Group, and Control Group
were 142 16% (range, 122197%), 121 9% (range,
106142%), and 105 8% (range, 86122%), respectively.
There were significant differences (p < 0.0001) between
any two of the three groups. The ratios of patients
3.3. Comparison of LVDd ratio with end-diastolic
left pulmonary artery flow and LA/Ao ratio
4. Discussion
Although the indications for the treatment of VLBWIs with
PDA have been studied extensively, there are currently no
clear guidelines for the treatment of these patients. This can
be partly attributed to the difficulty of establishing criteria to
decide disease severity. Several criteria are currently used to
determine the optimal treatment of VLBWIs with PDA,
including LPA end, LA/Ao ratio, LVO to superior vena cava flow
(SVC) ratio (LVO/SVC), PDA size or its flow, PDA/LPA ratio, left
ventricular output (LVO) to right ventricular output (RVO)
ratio (LVO/RVO) determined by echocardiography, increased
cardiothoracic ratio on the chest X-ray, increased heart rate,
bounding pulse, decreased urinary output, and increased
brain natriuretic peptide value. However, no previous study
has reported a confirmed criterion for determining the management of VLBWIs with PDA as some of the criteria require
special techniques or the data show very wide variance or are
within the normal range in patients with significant PDA.
Because of the ambiguous criteria involved in the treatment
(mentioned above), deciding the appropriate treatment
approach is difficult.
Clinical symptoms, electrocardiography, and chest X-ray
are regarded as neither accurate nor specific.4,5 Although
echocardiography is commonly used to evaluate PDA
Novel criteria for neonatal PDA treatment
253
Figure 1 Distribution of the LVDd ratio, the end-diastolic flow of left pulmonary artery, and the LA/Ao ratio. The LVDd ratio is the
measured LVDd divided by the normal LVDd derived from the equation. LA/Ao Z the ratio of left atrium diameter divided by
ascending aorta diameter; LVDd Z left ventricular end-diastolic dimension.
severity, other methods have been proposed in different
studies.
Hojjar et al6 reported that LVO/SVC was the most
appropriate criterion and that LA/Ao, DA diameter, mean,
and LPA end were accurate and easy to measure compared
with left ventricular output divided by superior vena cava
flow (LVO/SVC) in 23 neonates. However, LVO/SVC measurement requires a special technique, and LA/Ao is
considerably influenced by the three-dimensional shape of
the LA, which varies significantly between patients. The
LPA end also has wide variance as shown in the present
study. Condo
` et al7 reported that the size of the DA and the
flow through it were both appropriate for evaluating PDA
severity in 97 extremely low birth weight infants (ELBWI).
The DA size estimation is often difficult because it is not a
simple straight vessel but has a complex shape. There was
no apparent relation between the DA size and LVDd ratio
(Figure 2). Importantly, flow volume through the DA depends on two parameters, the diameter of the DA at the
narrowest point and the pressure gradient between the
aorta and the pulmonary artery. The blood flow through the
DA is variable, as the flow is influenced not only by the
pressure gradient, but also the angle between the Doppler
echocardiography echo beam and blood flow. Ramos et al8
assessed 115 ELBWIs and reported that a moderate to large
Table 2
PDA determined from the PDA/LPA ratio at or before days
after birth can identify neonates <27 weeks’ gestation who
subsequently require PDA closure. However, this study
assessed the criteria for treatment only in neonates <27
weeks’ gestation. Phillipos et al9 reported that LVO/RVO
was an appropriate measure of PDA severity. These authors
determined that the value of the LVO minus the RVO equals
the flow through the DA. However, because the flow
through the oval foramen influences RVO in premature
neonates, the formula is not always applicable.4 Furthermore, previous reports have not included a sufficient
number of patients for an accurate evaluation.
Although the evaluation of neonates using the criteria
mentioned above can detect statistically significant
Characteristics of patients after treatment.
Groups
Ligation
Indomethacin
Control
Catecholamines
Vasodilators
At least one therapy
22
6
25*
4
0
4
0
0
0
*p < 0.0001 for Indomethacin Group and Control Group.
Figure 2 Relationship between the DA size and the LVDd
ratio. No proportional relationship was observed between the
DA size and the LVDd ratio.
254
H. Nagasawa et al
differences between neonates with and without significant
PDA, this is only one of the requirements for determining
treatment. The important criterion used should have good
sensitivity and specificity in practical use, in other words,
the ideal criterion is associated with few exceptions and
can be measured at any institution. We assessed two
representative conventional criteria that have been used to
determine the severity of PDA, LPA end and LA/Ao ratio.
Although these criteria showed statistical significance for
determining the therapy for PDA, the data obtained using
these methods as indexes were associated with several
exceptions.
4.1. The superiority of the LVDd ratio as a criterion
Nagasawa1 reported novel regression equations for the
determination of LVDd in premature neonates. The LVDd in
PDA patients is commonly increased because of its
extended pulmonary flow, as a consequence of the flow
volume through the DA. Therefore, LVDd measurement
should be a useful criterion for determining treatment in
VLBWIs with PDA, which are characterized by increased left
ventricular end-diastolic volume. We estimated that the
measured LVDd divided by the normal LVDd (LVDd ratio)
might be the most appropriate index, which can be
measured at most institutions without the need for expert
personnel or special techniques.
The LVDd ratio was assessed in VLBWIs with PDA retrospectively. We selected 30 patients who had undergone DA
ligation, and the number of patients necessary for an accurate assessment was determined according to Cohen.3
We also selected VLBWIs treated using indomethacin and
patients who had not received radical therapy for PDA. Our
results showed that the LVDd ratio has good sensitivity
(86%) and specificity (84%) for disease staging to determine
the need for PDA ligation in VLBWIs.
4.2. Calculation of cut-off index for decisionmaking
The cut-off index to evaluate the indications for ligation
operation and indomethacin therapy in PDA patients was
investigated.
Qp/Qs >2.0 is the standard indication volume ratio to
operate patients with ventricular septal defect.10 An LVO/
RVO ratio close to 2 was reported to be associated with
increased intraventricular hemorrhage and periventricular
leucomalacia.9 The increase in LVDd when the LV is
enlarged to twice its volume has not been determined. We
measured the conventional fractional shortening (FS) and
the FS of the long dimension (Lg-FS) in the LV in 155 normal
full-term neonates. The mean Lg-FS/FS ratio was 0.69,
when the increased LVDd value is defined as “x”, the
equation is as follows: (1þx)2 (1þ0.69x) Z 2; x Z 0.29.
We calculated that the cut-off value of the LVDd ratio for
DA ligation was 130% of the normal ratio.
The criterion value to administrate indomethacin was
calculated in two ways. When Qp/Qs Z 1.5 and the LVDd ratio
was defined as “y”, the equation was (1 þ y)2 (1þ0.69 y) Z 1.5; y Z 0.16. The mean LVDd ratio in the
Indomethacin Group minus its SD was 112% and that of the
Figure 3 Treatment algorithm for VLBWIs with PDA. Dotted
lines show alternative ways to the primary choices.
ICH Z intra-cranial hemorrhage.
Control Group plus its SD was 113%. The cut-off value of the
LVDd ratio for indomethacin treatment was 115% of the normal
value.
4.3. Algorithm for the assessment and treatment of
VLBWIs with PDA
An algorithm for the assessment and treatment of VLBWIs
with PDA was proposed and is shown in Figure 3.
An LVDd ratio of 130% of the normal value or higher indicates the need for PDA ligation. If the LVDd ratio is between 115% and 130%, indomethacin treatment should be
initiated. In cases in which indomethacin treatment is not
effective or contraindicated in patients with intracranial
hemorrhage, DA ligation should be performed. If the ratio is
less than 115%, the patient should be treated with conventional therapy.
Koch et al reported that at least one-third of neonates
with a birth weight of 1000 g or less had spontaneous DA
closure and did not need medical treatment.11 Previous
reports have shown that the prognosis of patients with
indomethacin treatment is not necessarily affirmative,12e15
and adverse effects have been reported in association with
its use, which suggests that indomethacin administration
should be avoided whenever possible. The indication of
indomethacin treatment should be based on both different
indexes and physiological symptoms.
LVDd ratio as a criterion for treatment has some limitations. Differences in the measurement of the LVDd ratio
according to sex, race, institution, and equipment have not
been determined. The ratio should be applied carefully
within 24 hours after birth owing to its elliptic shape in the
short axis view of the LV.
A prospective study analyzing the LVDd ratio will be
performed in the future.
Conflicts of interest
All authors have no conflicts of interest to declare.
Novel criteria for neonatal PDA treatment
255
References
1. Nagasawa H. Novel regression equations of left ventricular
dimensions in infants less than 1 year of age and premature
neonates obtained from echocardiographic examination. Cardiol Young 2010;20:526e31.
2. Mertens L, Seri I, Marek J, Arlettaz R, Barker P,
McNamara P, et al. Targeted neonatal echocardiography in
the neonatal intensive care unit: practice guidelines and
recommendations for training. Writing Group of the American Society of Echocardiography (ASE) in collaboration
with the European Association of Echocardiography (EAE)
and the Association for European Pediatric Cardiologists
(AEPC). J Am Soc Echocardiogr 2011;24:1057e78.
3. Cohen J. Statistical power analysis for the behavioral sciences. 2nd ed. Hillsdale, NJ: Lawrence Erlbaum Associates;
1988. pp. 19e74.
4. Evans N. Diagnosis of patent ductus arteriosus in the preterm
newborn. Arch Dis Child 1993;68:58e61.
5. Kupferschmid C, Lang D, Pohlandt F. Sensitivity, specificity and
predictive value of clinical findings, m-mode echocardiography
and continuous wave Doppler sonography in the diagnosis of
symptomatic patent ductus arteriosus in preterm infants. Eur J
Pediatr 1988;147:279e82.
6. El Hajjar M, Vaksmann G, Rakza T, Kongolo G, Storme L.
Severity of the ductal shunt: a comparison of different
markers. Arch Dis Child Fetal Neonatal Ed 2005;90:
F419e22.
7. Condo
` M, Evans N, Bellu
` R, Kluckow M. Echocardiographic
assessment of ductal significance: retrospective comparison
8.
9.
10.
11.
12.
13.
14.
15.
of two methods. Arch Dis Child Fetal Neonatal Ed 2012;97:
35e8.
Ramos FG, Rosenfeld CR, Roy L, Koch J, Ramaciotti C. Echocardiographic predictors of symptomatic patent ductus arteriosus in extremely-low-birth-weight preterm neonates. J
Perinatol 2010;30:535e9.
Phillipos EZ, Robertson MA, Byrne PJ. Serial assessment of
ductus arteriosus hemodynamics in hyaline membrane disease.
Pediatrics 1996;98:1149e53.
Bernstein Daniel. Nelson textbook of pediatrics. 16th ed.
London: WB Saunders Co. Ltd; 2000. pp. 1369e71.
Koch J, Hensley G, Roy L, Brown S, Ramaciotti C, Rosenfeld CR.
Prevalence of spontaneous closure of the ductus arterisosus in
neonates at a birth weight of 1000 grams or less. Pediatrics
2006;117:1113e21.
Schmidt B, Davis P, Moddemann D, Ohlsson A, Roberts RS,
Saigal S, et al. Long-term effects of indomethacin prophylaxis
in extremely-low-birth-weight infants. N Engl J Med 2001;344:
1966e72.
Laughon MM, Simmons MA, Bose CL. Patency of the ductus
arteriosus in the premature infant: is it pathologic? Should it be
treated? Curr Opin Pediatr 2004;16:146e51.
Schmidt B, Roberts RS, Fanaroff A, Davis P, Kirpalani HM,
Nwaesei C, et al. Indomethacin prophylaxis, patent ductus
arteriosus, and the risk of bronchopulmonary dysplasia: further
analyses from the trial of indomethacin prophylaxis in preterms (TIPP). J Pediatr 2006;148:730e4.
Carmo KB, Evans N, Paradisis M. Duration of indomethacin
treatment of the preterm patent ductus arteriosus as directed
by echocardiography. J Pediatr 2009;155:819e22.
Pediatrics and Neonatology (2014) 55, 256e261
Available online at www.sciencedirect.com
ScienceDirect
journal homepage: http://www.pediatr-neonatol.com
ORIGINAL ARTICLE
Renal Excretion of Water-soluble Contrast
Media After Enema in the Neonatal Period
Hee Sun Kim a, Bo-Kyung Je a,*, Sang Hoon Cha a,
Byung Min Choi b, Ki Yeol Lee a, Seung Hwa Lee a
a
b
Department of Radiology, Korea University Hospital, Ansan, Gyeonggido, Republic of Korea
Department of Pediatrics, Korea University Hospital, Ansan, Gyeonggido, Republic of Korea
Received Jul 11, 2012; received in revised form Dec 7, 2012; accepted Jul 29, 2013
Available online 2 December 2013
Key Words
contrast enema;
contrast media;
diatrizoate
meglumine;
ileus;
infant;
intestinal
obstruction;
newborn;
renal excretion;
water-soluble
iodinated
Background: When abdominal distention occurs or bowel obstruction is suspected in the
neonatal period, a water-soluble contrast enema is helpful for diagnostic and therapeutic purposes. The water-soluble contrast medium is evacuated through the anus as well as excreted
via the kidneys in some babies. This study was designed to evaluate the incidence of renal
excretion after enemas using water-soluble contrast media and presume the causes.
Methods: Contrast enemas using diluted water-soluble contrast media were performed in 23
patients under 2 months of age. After the enema, patients were followed with simple abdominal radiographs to assess the improvement in bowel distention, and we could also detect the
presence of renal excretion of contrast media on the radiographs. Reviewing the medical records and imaging studies, including enemas and consecutive abdominal radiographs, we evaluated the incidence of renal excretion of water-soluble contrast media and counted the stay
duration of contrast media in urinary tract, bladder, and colon.
Results: Among 23 patients, 12 patients (52%) experienced the renal excretion of watersoluble contrast media. In these patients, stay-in-bladder durations of contrast media were
1-3 days and stay-in-colon durations of contrast media were 1-10 days, while stay-in-colon durations of contrast media were 1-3 days in the patients not showing renal excretion of contrast
media. The Mann-Whitney test for stay-in-colon durations demonstrated the later evacuation
of contrast media in the patients with renal excretion of contrast media (p Z 0.07). The review of the medical records showed that 19 patients were finally diagnosed as intestinal diseases, including Hirschsprung’s disease, meconium ileum, meconium plug syndrome, and
small bowel atresia or stenosis. Fisher’s exact test between the presence of urinary excretion
and intestinal diseases indicated a statistically significant difference (p Z 0.04).
* Corresponding author. Department of Radiology, Korea University Hospital, #123, Jeokgeum-ro, Danwon-gu, Ansan-si, Gyeonggi-do
425-707, Republic of Korea.
E-mail address: [email protected] (B.-K. Je).
1875-9572/$36 Copyright ª 2013, Taiwan Pediatric Association. Published by Elsevier Taiwan LLC. All rights reserved.
http://dx.doi.org/10.1016/j.pedneo.2013.07.012
Renal excretion of water-soluble contrast media after enema
257
Conclusion: The intestinal diseases causing bowel obstruction may increase the water-soluble
contrast media’s dwell time in the bowel and also increase urinary excretion.
Copyright ª 2013, Taiwan Pediatric Association. Published by Elsevier Taiwan LLC. All rights
reserved.
1. Introduction
Diatrizoate meglumine is a water-soluble contrast medium
for the gastrointestinal (GI) tract. Usually, it is indicated for
the early postoperative assessment following GI surgery and
the evaluation of suspected GI perforation.1 This watersoluble contrast medium has been used in neonates for
the treatment of uncomplicated meconium ileus, and it was
believed to loosen the viscous, tenacious meconium
through its hyperosmolar and water-soluble properties.
Large volumes of fluid are drawn into the bowel, freeing
the meconium and allowing it to pass through the rectum.26 The patients who undergo water-soluble contrast
enemas are followed with simple abdominal radiographs to
evaluate the improvement in bowel distention and the
effectiveness of enema. In some patients, we noticed that
the urinary tracts, mainly urinary bladders, were opacified
on the radiographs.
In this study, we evaluated the cases of water-soluble
contrast enema in patients less than 2 months of age and
divided them into two groups according to the presence of
opacified urinary tracts on follow-up simple abdominal radiographs, which suggests renal excretion of contrast
media. Then, we investigated the differences between the
two groups and the causes of urinary tract opacification.
2. Methods
Since April 2004, 32 patients less than 1 year of age underwent contrast enemas with water-soluble contrast
Table 1
media at our institute: 10 examinations in 2004, four in
2005, two in 2006, two in 2007, 11 in 2008, two in 2009, and
one in 2011. All studies were performed after obtaining
documented informed consent from the parents or legal
guardians. With retrospective review of the medical records and imaging studies, we excluded patients with uncertain diagnosis, patients lost to follow-up, patients with
bowel perforation, and patients whose follow-up abdominal
radiographs were not performed until the time when the
instilled contrast media were completely evacuated.
Finally, we enrolled 23 patients in this study and patients’
demographics are presented in Table 1.
All 23 patients underwent water-soluble contrast enemas for severe abdominal distention or a lack of fecal
evacuation in spite of saline or glycerin enemas. The
contrast material used was diatrizoate meglumine and
diatrizoate sodium solution (Gastrografin; Bayer Schering
Pharma, Santa Rosa, Spain). It is a lemon-flavored, watersoluble, hyperosmolar (1750 mOsm/L), iodinated radiopaque contrast medium containing amidotrizoic acid
597 mg/mL, meglumine 159 mg/mL, sodium hydroxide
6 mg/mL, and bound iodine 367 mg/mL. Owing to the
hyperosmolarity of the contrast media, patients received
intravenous fluids to prevent any possible imbalance of
fluid or electrolytes prior to the study. Contrast enemas
were performed with the following procedure. A Foley
catheter was placed in the rectum without ballooning, and
the buttocks were strapped tightly together to prevent
leakage of contrast. At the beginning of the study, all patients were positioned for a lateral view in order to evaluate the presence of the narrowing segment in the rectum
Demographics of 23 patients according to excretory pathway of contrast.
Group
Sex
Male
Female
Gestational age
(wk)
Maturity
Preterm
Term
Birth weight
(g)
Delivery type
Vaginal
Cesarean
12 (52)
11 (48)
10 (43)
19 (57)
8 (35)
15 (65)
Age at enema
The numbers in parentheses represent the percentages.
*Student t test.
R group (N Z 12)
NR group (N Z 11)
6 (50)
6 (50)
6 (55)
5 (45)
28w41 (36 2)
26w41 (35 3)
5 (42)
7 (58)
5 (45)
6 (55)
1340w4000
(2616 507)
976w3840
(2296 573)
6 (50)
6 (50)
2 (18)
9 (82)
0w30 days
Birthw1 wk 9 (75)
1 wkw1 mo 2 (17)
1 mow 1 (8)
1w65 days
Birthw1 wk 5 (45.5)
1 weekw1 mo 1 (9)
1 mow 5 (45.5)
p
0.33*
0.24*
258
H.S. Kim et al
Figure 1 (A) A 4-day-old boy with abdominal distention. The colon study demonstrates multiple filling defects in descending and
transverse colon, slowly refluxed into the proximal colon and terminal ileum by the pressure of injected contrast. (B) Supine
abdominal radiograph taken 3 hours after the colon study demonstrated opacified urinary bladder.
and sigmoid, i.e., the transitional zone, which was helpful
for diagnosing Hirschsprung’s disease. Ten milliliters of
Gastrografin diluted with 40 mL of warm saline (20% of solution) was prepared in a 50-mL enema syringe. Under
fluoroscopic guidance, water-soluble contrast was gently
infused by hand injection. The amount of instilled contrast
media varied in each patient, from 20 to 100 mL of diluted
contrast. The amount of contrast media was determined by
the examiner during the procedure, according to the
Figure 2 A preterm baby of 34 þ 6 weeks gestational age
who suffered from a disturbance in the meconium passage. The
last film of the colon study reveals the absence of normal
haustra and a greatly decreased caliber of the colon, suggesting microcolon. A rectal tube was inflated with 3 mL of air
for effective reflux of Gastrografin into the small bowel.
Finally, a dilated small bowel was opacified, after the long
tubular filling defects in the distal ileum, which implied that
the meconium plugs were refluxed.
progression of contrast media. When small bowel dilatation
was evident on the abdominal radiograph, or when meconium obstruction was suspected clinically, we tried to send
the contrast media into the ileum via the ileocecal valve
and finally to the dilated small bowel that was proximal to
the obstructing meconium (Figure 1). Sometimes, if the
transitional zone was ruled out at the beginning of the
procedure, we inflated the Foley catheter balloon with
2e3 mL of air to send more contrast into the proximal
bowel, resulting in more effective liquefaction of the
inspissated meconium plug (Figure 2). After the enema,
patients were followed with simple abdominal radiographs
to assess the improvement in bowel distention and the
effectiveness of the enema.
One general radiologist and one pediatric radiologist
retrospectively reviewed the images and follow-up
abdominal radiographs in consensus. The renal excretion
of water-soluble contrast media was defined when the
urinary tracts, including bladders, were opacified on
consecutive abdominal supine or lateral radiographs after
enema (Figures 1 and 3). We divided the patients into two
groups: patients experiencing renal excretion as well as
anal evacuation of contrast media were categorized as
group “R”, and the remaining as group “NR”, which means
that renal excretion was not evident.
We reviewed the medical chart and imaging findings and
analyzed the differences between the two groups using the
Mann-Whitney rank sum test and Fisher’s exact test. Twotailed p values of <0.05 were considered significant. We
utilized the MedCalc v.12.1.4 statistical software package
(MedCalc Software, Mariakerke, Belgium) for analysis.
3. Results
Among the 23 patients who underwent enema with watersoluble contrast media, 12 patients (52%) showed an opacified urinary tract and bladder on follow-up abdominal
radiographs after enema and were classified as the R group.
The remaining 11 patients (48%) were assigned to the NR
group (Table 1).
Renal excretion of water-soluble contrast media after enema
259
Figure 3 (A) A 28-day-old boy with severe abdominal distention. The initial supine abdominal radiograph shows abdominal
bulging and gaseous distention of the bowel with no gas in the rectum. (B) Supine and (C) cross table lateral abdominal radiographs
are taken 8 hours after the colon study. The contrast filled the colon, and the urinary bladder is opacified in front of the rectum
that is demonstrated well on both views. (D) Cross table lateral abdominal radiograph taken on the next day of contrast study.
Although the density is decreased, the urinary bladder is still visible. The patient was diagnosed with Hirschsprung’s disease by
rectal biopsy that revealed an absence of ganglion cells.
All 12 patients of the R group showed the opacified urinary bladder on abdominal radiographs. Among them, nine
patients showed an opacified urinary bladder on both
abdomen supine and lateral views, while three patients
presented with an opacified bladder on only the lateral
views. The days of stay-in-colon as well as stay-in-bladder
were counted reviewing the abdominal radiographs. The
results for each group are presented in Table 2.
In terms of the stay-in-colon duration, the MannWhitney test between the two groups presented the tendency of earlier evacuation of contrast in the NR group than
in the R group; however, this was not statistically significant
(p Z 0.07).
In terms of the stay-in-bladder duration, the bladder was
opacified and visible on the day of the enema in all 12 patients; visible to the next day in eight patients; visible to the
3rd day in four patients; and not depicted after the 4th day.
A review of medical records determined final diagnoses,
which included various congenital GI diseases, such as six
patients with Hirschsprung’s disease, including one case
with total aganglionosis, five with meconium ileus, four
with meconium plug syndrome, one with ileal atresia, one
with ileal stenosis, one with duodenal atresia, and one with
jejunal atresia. All 12 patients of the R group had the above
mentioned congenital GI diseases. Seven patients (64%) of
the NR group were proved to have congenital GI diseases,
but not in the remaining four patients. Fisher’s exact test
between the presence of urinary excretion and the diagnosed congenital GI disease was performed. The resulting
p value was 0.037 with 100% of sensitivity, 36.4% of specificity, 63.2% of the positive predictive value, and 100% of
the negative predictive value, which indicated that there
was a statistically significant difference in the presence of
urinary excretion and GI diseases (Table 2).
4. Discussion
Non-surgical treatment of meconium ileus was first attempted using hydrogen peroxide in 1954 by Olim et al4 Dr Noblett
260
H.S. Kim et al
Table 2 Findings on follow-up abdominal radiographs and
medical record data according to excretory pathway of
contrast.
Group
R group
(N Z 12)
NR group
(N Z 11)
p
Stay-in-colon
duration
of contrast
1w10 d
w1st d 4
(33.3%)
w3rd d 4
(33.3%)
4th dw 4
(33.3%)
1w3 d
w 1st d 7
(63.6%)
w 3rd d 4
(36.4%)
4th dw 0
0.07*
Stay-in-bladder
duration
of contrast
1w3 d
w1st d 4
(33.3%)
w2nd d 4
(33.3%)
w3rd d 4
(33.3%)
Presence of
intestinal
problem
12 (100%)
7 (64%)
0.04y
*Mann-Whitney test.
y
Fisher’s exact test.
reported the use of Gastrografin for four patients with
meconium ileus in 1969 and emphasized the importance of
administering a radiopaque agent with fluoroscopic control
for safety concerns.3 Later, several articles also reported the
use of Gastrografin for the treatment of uncomplicated
meconium ileus, especially in neonates and young infants.26
Gastrografin is believed to be successful in relieving the intestinal obstruction in meconium ileus because of its high
osmolarity that draws fluid into the bowel lumen from the
plasma, which has an osmolarity of 280w300 mOsm/L,
thereby loosening the viscous, tenacious meconium. In
addition to the hypertonic nature of the contrast medium,
the wetting agent is thought to lubricate the meconium and
facilitate its passage to the distal bowel.3 In our institution,
water-soluble contrast enema is performed for neonates and
young infants who require relief from intestinal obstruction
due to a meconium plug or rarely meconium ileus that is not
relieved with glycerin or saline enema. Diatrizoate meglumine may carry some risk of dehydration, resulting in an
increased hematocrit, rising serum osmolarity, and reduction
in pulse rate and cardiac output.5 Therefore, careful attention regarding sufficient hydration is necessary before, during, and after the procedure.
After water-soluble contrast enema, most of the watersoluble contrast media is evacuated via the rectum, and
absorption from the intestine is minimal. Then, how to
explain the opacified urinary tracts (mostly urinary bladders)
in some patients of our study? Since the report by Mori and
Barrett in 1962,7 the presence of ingested water-soluble
contrast media in the urinary tract, as detected on abdominal radiographs, has been regarded as a strongly suggestive
finding of gastrointestinal perforation or anastomotic
leakage after gastrointestinal surgery.1 In these cases, orally
administered water-soluble contrast media may leak into
the peritoneum and may be absorbed into the blood stream
and then be excreted by the kidneys. However, since the
introduction of computed tomography (CT), one report
presented the urinary excretion of orally administered
water-soluble contrast media observed on CT in patients
with various bowel diseases and had no perforation.8 Our
patients in the R group showed no free air in the abdomen
cross table lateral views. Therefore, we assumed that the
urinary excretion of Gastrografin in the R group was mainly
because of the absorption from the intestinal wall rather
than bowel perforation. We admit the limitation of abdominal radiographs compared with CT to show excreted
contrast because CT is a more sensitive modality for identifying the presence of urinary excretion of contrast media.
However, if performed, CT has a limitation in neonates and
young infants owing to the radiation hazard.
In our results, the stay-in-colon duration of contrast
media after the colon study was longer in the R group than
in the NR group, but this was not statistically significant
(p Z 0.07; Table 2). The patients in the R group had more
probability of the presence of GI diseases than that of the
patients in the NR group (p Z 0.04; Table 2). Therefore, for
the R group, we assumed that the longer stay-in-colon
duration was related to or caused by the GI diseases,
resulting in colonic obstruction, which led to more Gastrografin being absorbed from the intestinal wall.
What is the proper amount of diatrizoate meglumine used
for enemas in neonates and young infants? Few reports have
described the precise amount of diatrizoate meglumine used
for enemas in neonates and young infants. O’Halloran et al2
reported that the majority of children who required Gastrografin enemas to relieve meconium ileus had been given
100-200 mL of Gastrografin in three times the volume of
water under fluoroscopic control. Garza-Cox et al9 stated
that up to 20 mL of solution was generally sufficient to outline
the colon, distal ileum, and inspissated meconium plugs on
plain abdominal radiographs in very low birth weight premature infants. In our study, although the amount was
different in each case, we instilled 20e100 mL of diluted
contrast, which was not a large amount, considering the
gestational age of the enrolled patients.
Bowel perforation is a representative, but not commonly
occurring, complication of contrast enema in children. A
survey by Kao and Franken10 demonstrated 2.75% of
perforation, and its occurrence was not correlated with the
success rate of the enema or properties of the contrast
medium. The only high risk factor for rectal perforation was
reported to be the use of an inflated balloon catheter,
similar to our result. We experienced one unfortunate case
with rectal perforation, although we did not enroll the
patient in the current study. This patient was a male of
37 þ 2 weeks of gestational age and his initial body weight
was 2900 g. He underwent colon study on the 5th day of life.
The Foley catheter balloon was inflated from the beginning
of the colon study due to the inexperience of the examiner.
In conclusion, an opacified urinary tract and bladder can
be observed on follow-up abdominal radiographs after
water-soluble contrast enema in neonates and young infants. The GI diseases causing colonic obstruction may increase water-soluble contrast dwell time in the bowel and
increase urinary excretion.
Renal excretion of water-soluble contrast media after enema
261
Contributors
Bo-Kyung Je designed the study and data collection tools,
monitored data collection for the whole trial, wrote the
statistical analysis plan, and cleaned and analysed the
data. Hee Sun Kim wrote the statistical analysis plan and
monitored data collection for the whole trial. Bo-Kyung Je,
Sang Hoon Cha, Ki Yeol Lee, and Seung Hwa Lee performed
the radiologic examination and read the radiographs. Byung
Min Choi made the clinical diagnosis and monitored the
patients’ data. Bo-Kyung Je and Hee Sun Kim drafted the
paper. Bo-Kyung Je, Sang Hoon Cha and Byung Min Choi
revised the draft paper.
Acknowledgments
This study was supported in part by a grant from Korea
University College of Medicine (K1032111).
Conflicts of interest
The authors have no conflicts of interest relevant to this
article.
References
1. Sohn KM, Lee SY, Kwon OH. Renal excretion of ingested gastrografin: clinical relevance in early postoperative treatment
2.
3.
4.
5.
6.
7.
8.
9.
10.
of patients who have undergone gastric surgery. AJR Am J
Roentgenol 2002;178:1129e32.
O’Halloran SM, Gilbert J, McKendrick OM, Carty HM, Heaf DP.
Gastrografin in acute meconium ileus equivalent. Arch Dis
Child 1986;61:1128e30.
Noblett HR. Treatment of uncomplicated meconium ileus by
Gastrografin enema: a preliminary report. J Pediatr Surg 1969;
4:190e7.
Olim CB, Ciuti A. Meconium ileus: a new method of relieving
obstruction; report of two cases with successful management.
Ann Surg 1954;140:736e40.
Rowe MI, Furst AJ, Altman DH, Poole CA. The neonatal
response to gastrografin enema. Pediatrics 1971;48:29e35.
Wagget J, Bishop HC, Koop CE. Experience with gastrografin
enema in the treatment of meconium ileus. J Pediatr Surg
1970;5:649e54.
Mori PA, Barrett HA. A sign of intestinal perforation. Radiology
1962;79:401e7.
Apter S, Gayer G, Amitai M, Hertz M. Urinary excretion of orally
ingested gastrografin on CT. Abdom Imaging 1998;23:297e300.
Garza-Cox S, Keeney SE, Angel CA, Thompson LL, Swischuk LE.
Meconium obstruction in the very low birth weight premature
infant. Pediatrics 2004;114:285e90.
Kao SC, Franken Jr EA. Nonoperative treatment of simple
meconium ileus: a survey of the Society for Pediatric Radiology. Pediatr Radiol 1995;25:97e100.
Pediatrics and Neonatology (2014) 55, 262e268
Available online at www.sciencedirect.com
ScienceDirect
journal homepage: http://www.pediatr-neonatol.com
ORIGINAL ARTICLE
Cobedding of Twin Premature Infants:
Calming Effects on Pain Responses
Zohreh Badiee*, Zohreh Nassiri, Amirmohammad Armanian
Department of Pediatrics, School of Medicine, and Child Growth and Development Research Center,
Isfahan University of Medical Sciences, Isfahan, Iran
Received Jun 5, 2013; received in revised form Nov 14, 2013; accepted Nov 26, 2013
Available online 30 March 2014
Key Words
heel lance;
pain;
preterm infant
Background: The purpose of this trial was to determine whether cobedding of preterm twins
has analgesic effects during heel lancing or not.
Methods: One hundred premature twins (50 sets) born between 26 weeks’ and 34 weeks’
gestation undergoing heel blood sampling were randomly assigned into two groups: the cobedding group (receiving care in the same incubator) and the standard care group (receiving care
in separate incubators). Pain was assessed using the premature infant pain profile score. Duration of crying was measured after heel blood sampling, and salivary cortisol was measured
prior to and after heel blood sampling.
Results: Infants in the standard care group cried for a longer time during heel lancing than
those in the cobedding group (42.6 19.8 seconds vs. 36.4 21.7 seconds, p Z 0.03). The
mean premature infant pain profile score after heel lancing was significantly higher in the standard care group (9.8 2.6 vs. 8.06 2.8, p Z 0.002). The mean salivary cortisol after heel
lancing was also significantly higher in the standard care group (24.3 7.4 nmol/L vs. 20.8
7.4 nmol/L, p Z 0.02). No significant adverse effects were seen with cobedding.
Conclusion: Cobedding is a comforting measure for twin premature infants during heel lancing,
which can be performed without any significant adverse effects.
Copyright ª 2014, Taiwan Pediatric Association. Published by Elsevier Taiwan LLC. All rights
reserved.
1. Introduction
Over the past 10e15 years, the rate of multiple births has
risen in many countries.1 Because multiple births are at
* Corresponding author. Department of Pediatrics, School of
Medicine, and Child Growth and Development Research Center,
Isfahan University of Medical Sciences, Isfahan, Iran.
E-mail address: [email protected] (Z. Badiee).
high risk of resulting in preterm birth and low birth weight,
their increasing incidence has led to a rising number of
premature infants being admitted to neonatal intensive
care units (NICUs).2e4
Technological and therapeutic advances in perinatal
care have resulted in an increased survival rate in premature infants. These premature infants are subjected to a
variety of invasive painful procedures as part of their
management. Although pain management for routine
http://dx.doi.org/10.1016/j.pedneo.2013.11.008
1875-9572/Copyright ª 2014, Taiwan Pediatric Association. Published by Elsevier Taiwan LLC. All rights reserved.
Cobedding of twin premature infants
procedural pain has improved in recent years, almost 40% of
infants undergoing heel prick for blood collection have not
received any form of intervention for pain reduction.5
Several lines of evidence suggest that early and repeated
exposure to painful stimuli during the development of nervous system leads to persistent behavioral changes, alteration in pain processing, and development.6e8
Therefore, it is essential to prevent or treat pain in
newborn infants. There are pharmacological and nonpharmacological methods for pain management in newborn
infants. Nonpharmacological pain intervention is a prophylactic and complementary approach to reduction of
pain. A number of nonpharmacological methods have been
shown to be useful in the management of mild-to-moderate
pain in newborn infants. These methods include nonnutritive sucking, breastfeeding, swaddling, facilitated
tucking, kangaroo care, music therapy, and sensory saturation including the senses of touch (massaging), sound,
smell, and vision.9 Some methods such as premature infant
pain profile (PIPP) score are used for assessing pain in
premature infants.10 Recently, Nishitani and colleagues11
showed that salivary cortisol could be a useful index of
biochemical responses to pain in infants. Application of
salivary cortisol in addition to the PIPP score appears to
assess pain more accurately in premature neonates. By
contrast, during fetal life, twins share a small, dark,
enclosed space in which their bodies touch and are very
close to each other.12 After birth, they are routinely
separated and placed in separate incubators, which may
lead to separation stress.13 Cobedding twins is the practice
of placing siblings in the same crib or incubator. Cobedding
is believed to enhance twin coregulation, improve physiological stability, decrease oxygen requirement, improve
growth and development, and decrease the length of hospitalization and number of readmissions.14e16 In addition,
cobedding provides tactile, olfactory, and auditory stimulation and may decrease pain response in premature infants.17,18 Therefore, it is proposed that the presence of a
twin who has shared the same uterine environment may
have comforting effects on a preterm infant especially
during the neonatal care. Recently Campbell-Yeo and colleagues17 evaluated the pain reducing effects of cobedding
and found that cobedding could not decrease pain score in
twin premature newborns. However, they used sucrose and
a pacifier prior to heel lancing, which might have obscured
the analgesic effects of cobedding. In addition, they did
not evaluate the biochemical responses to pain. The purpose of this study was to assess the effects of cobedding
alone on pain reactivity in premature twin newborns.
2. Materials and Methods
This open randomized controlled trial was performed between May 2012 and December 2012 at the NICU of Shahid
Beheshti University Hospital, affiliated to Isfahan University
of Medical Sciences, Isfahan, Iran.
Newborns with gestational ages of 26e34 weeks and
postnatal ages of less than 20 days who underwent heel
blood sampling for blood glucose determination were
included in the trial. Infants who had received sedatives or
analgesics within 48 hours of heel lancing, or those who had
263
major congenital malformations, an Apgar scores of less
than 6 at 5 minutes after birth, severe respiratory distress
requiring mechanical ventilation, or severe intraventricular
hemorrhage were excluded from the study. The Ethics
Committee of Isfahan University of Medical Sciences
approved the study, and written consent was obtained from
the parents of all participants.
Randomization was performed using a computergenerated random number algorithm. Allocation of
eligible newborns to intervention and control groups was
performed using a sealed opaque envelope. Participants
were randomly allocated to either the cobedding group or
the standard care group. Infants in the standard care group
remained in separate incubators following the current NICU
standards. Newborns in the cobedding group were placed
side by side in an incubator without any clothing except for
diapers so that they could touch each other freely, with
each side of the incubator pertaining to one twin. Infants
were cobedded from 24 hours prior to heel sticks to the end
of the study, and heel lancing was performed between 8 AM
and 10 AM.
The incubator temperature was adjusted according to
the weight, gestational age, and postnatal age of newborns. Each infant’s axillary temperature was closely
monitored and maintained between 36.8 C and 37.2 C for
both groups.
Blood sampling was performed in a standardized manner
by expert technicians who could not be blinded to the
study. The same technician held up each baby’s heel,
pricked it to collect the blood sample, and applied an adhesive bandage to the heel immediately afterward. Data
were collected just once for each infant. The reason for
blood sampling was diagnostic or routine screening for hypoglycemia. Oxygen saturation and heart rate were recorded on a Masimo pulse oximeter (Masimo Corporation,
Irvine, CA, USA) during the study. Salivary cortisol was
collected 1 minute prior to heel lancing and 20 minutes
after the heel lancing with a 2 mL syringe. Samples were
washed in the laboratory with WASHER ELISA instrument
(SEAC Company, Calenzano, Italy), and then they were
measured by ELISA (Human ELISA kits, Diametra, Foligno,
Italy).
Prior to starting the study, researchers were trained to
assess the PIPP. The researchers could not be blinded for
the assigned groups. The PIPP score is a pain assessment
tool that has been validated for procedural pain in premature and term infants.10 Pain assessment using the PIPP
score is based on several criteria, including gestational age,
behavioral state, heart rate, oxygen saturation, and three
facial scores. Scores range from 0 to 21, with a higher score
indicating more pain (Appendix 1).
From 1 minute prior to the start of the procedure to 2
minutes after its completion, physiological and facial
changes were video recorded. On the basis of these recordings, newborns were evaluated for the PIPP score
during 30 seconds after heel lancing by trained researchers.
The primary outcome of the study was to evaluate the
pain caused by the heel stick in preterm infants using the
PIPP score and the secondary outcome was to evaluate
cortisol secretion in response to pain. The PIPP score was
calculated based on the recordings by three researchers.
The inter-rater reliability was 0.9.
264
Z. Badiee et al
2.1. Sample size and statistical analysis
Considering a confidence level of 95%, b Z 80%, and also
minimum significant differences between groups equal to
0.6 seconds,19 we estimated that 44 infants were needed to
detect the effect of cobedding on pain response in preterm
infants.
Obtained data were analyzed by independent t test
and an analysis of variance test using SPSS software
(version 18; SPSS Inc., Chicago, IL, USA), and p < 0.05
was considered statistically significant. Statistical analyses were corrected for potential nonindependence of
outcomes between twin pairs by the generalized estimating equation procedures.
3. Results
From May 2012 to December 2012, 123 twin preterm infants
were assessed for eligibility and finally 105 of them were
included in the study. The main reason that infants could
not be included was parenteral refusal. A few infants were
excluded because there was no need for blood sampling.
The study flowchart is shown in Figure 1.
No significant differences were observed between the
two groups with regard to patient demographics (Table 1).
The mean heart rates prior to heel lancing were not
different between groups. However, the maximum heart
rates after the procedure were significantly higher in the
standard care group. The mean PIPP score was about 1.7
Figure 1
points lower in the cobedding group, supporting the
expectation that cobedding would reduce the PIPP score by
1 point. Severe pain, characterized by a PIPP score of more
than 12, was seen in 20% of newborns in the standard care
group and 6% in the cobedding group. Eighteen percent of
newborns in the cobedding group and 2% of patients in the
standard care group did not feel pain, as characterized by a
PIPP score of less than 6 (p Z 0.002). The mean duration of
crying was about 6 seconds shorter in the cobedding group
(p Z 0.03). Salivary cortisol 1 minute prior to the initiation
of heel lancing was not different between groups; however,
20 minutes after heel lancing, the mean salivary cortisol
was reduced by 3.5 nmol/L in the cobedding group, which
supported the basic theory that cobedding could attenuate
the pain caused by heel sticks. Table 2 shows the estimated
mean standard deviation of the PIPP score, crying time,
and salivary cortisol prior to and after heel lancing in the
two groups of cobedding and standard care. Our results
showed that, with increasing gestational age, the pain
reducing effects of cobedding became more prominent so
that decreases in the PIPP score, salivary cortisol, and
crying time were higher in older infants. The mean and
standard deviation of the variables, such as the PIPP score,
crying time, and salivary cortisol, prior to and after heel
lancing, based on the birth weight and gestational age, are
shown in Tables 3 and 4.
We did not find any significant increase in the incidence
of infection, necrotizing enterocolitis, apnea, or bradycardia in the cobedding group in comparison to the control
group.
Study flowchart.
Cobedding of twin premature infants
Table 1
265
Baseline characteristics of two groups.
Sex
Gestational age
(wk)
Boy
Cobedding group
Standard care group
p
Girl
N
%
N
%
26
27
0.5
52
54
24
23
48
46
Birth weight (g)
Postnatal age at
heel lance (d)
Mean
()SD
Mean
()SD
Mean
()SD
32.08
32.02
0.86
1.8
1.7
1500.6
1491.3
0.84
250.03
254.12
3.88
3.08
0.7
2.04
4.06
SD Z standard deviation.
4. Discussion
Development of new nonpharmacological pain control
methods is very important for the management of pain and
stress in preterm infants who are admitted to NICUs. We
found that cobedding of twin premature infants could
reduce their pain sensation, as assessed by the PIPP score
and salivary cortisol. In addition, cobedding decreased the
crying time during heel lancing in twin preterm neonates.
Moreover, we found that severe pain, as characterized
by a PIPP score of more than 12, was significantly lower in
the cobedding group.
Although two studies performed by the Campbell-Yeo
and colleagues17,20 did not find any significant decrease in
the PIPP scores in the cobedding group, they found that
cobedding enhanced the physiological recovery of preterm
twins after heel blood sampling.
In Campbell-Yeo et al’s17 study, the mean postnatal age
at heel lance was 18.7 (20.6) days in the cobedding group
and 12.6 (16.1) days in the standard care group, which is
much longer than the 2.88 (2.04) days in the cobedding
group and 2.08 (4.06) days in the standard care group in our
study. In these studies, the lack of decrement in the PIPP
score after cobedding may be due to a lack of full skin
contact between twins or a prolonged period of separation
between them after birth.
Cobedding has some similarities with skin-to-skin contact (SSC) because both can have tactile, olfactory, and
auditory stimulation, which may affect pain responses in
premature infants. Cong and coworkers21 evaluated the
effects of SSC on behavioral response to pain in premature
infants. They found that SSC could reduce the PIPP score
during heel sticks. Moreover, Axelin and colleagues22
demonstrated that facilitated tucking has a greater analgesic effect than oral glucose and opioids in preterm infants. The pain-reducing effects of SSC in these studies are
consistent with our results. However, it is important to note
that the mean PIPP score in our study in the cobedding
group was very close to the PIPP score in the study of
Campbell-Yeo et al. Absence of significant differences in
the PIPP scores in that study might possibly be due to the
use of sucrose in all patients, because sucrose might have
such strong analgesic effects that any additional intervention could not decrease the pain score further.
We found that salivary cortisol secretion was attenuated
with cobedding of premature twins, possibly due to the
pain-reducing effects of cobedding. Therefore, our study
demonstrated better control of stress responses after
cobedding in premature twins, which is similar to the
findings of Campbell-Yeo and colleagues.20
Cong and coworkers21 evaluated the effects of SSC on
behavioral responses to heel blood sampling in premature
neonates. They indicated lower salivary and serum cortisol
levels after heel lancing in preterm infants who underwent
SSC than the control group and concluded that SSC could
blunt stress responses to pain after heel lancing. Lower
Table 2 Mean SD of the PIPP score, crying time, and salivary cortisol before and after heel lancing in two groups of
cobedding and standard care.
Standard care group (n Z 50)
Duration of crying (s)
Salivary cortisol
1 min prior to heel lancing (nmol/L)
Salivary cortisol
20 min after heel lancing (nmol/L)
PIPP score
Baseline “O2sat”
Minimum of O2sat after heel lancing
Baseline heart rate
Maximum of heart rate during heel lancing
Cobedding group (n Z 50)
42.7 (19.8)
18.5 (6.5)
36.4 (21.7)
18.3 (6.7)
24.3 (7.4)
20.8 (7.4)
9.5
93.3
89.6
125.1
139.6
(2.3)
(1.8)
(2.5)
(8.4)
(10.3)
Data are presented as mean (SD).
PIPP Z premature infant pain profile; SD Z standard deviation.
7.9
93.2
91.8
125.4
133.6
(2.6)
(1.4)
(1.8)
(7.6)
(8.01)
p
0.03
0.303
<0.001
<0.001
0.429
<0.001
0.17
<0.001
266
Z. Badiee et al
Table 3 Mean SD of the PIPP score, crying time, and salivary cortisol before and after heel lancing in two groups of
cobedding and standard care, based on three subgroups of birth weight.
Variable
Less than 1250 g
N Z 16
Mean
Duration of crying (s)
Salivary cortisol
1 min prior to heel
lancing (nmol/L)
Salivary cortisol
20 min after heel
lancing (nmol/L)
PIPP score
Baseline O2sat
Minimum of O2sat after
heel lancing
Baseline heart rate
Maximum of heart rate
during heel lancing
SD ()
1250e1500 g
N Z 35
Mean
SD ()
More than 1500 g
N Z 49
Mean
SD ()
Standard care group
Cobedding group
Standard care group
Cobedding group
66.14
52.11
22.15
19.02
17.79
19.77
8.56
6.21
45.16
39.17
20.03
20.96
2
24.19
6.73
7.27
34.28
28.54
16.37
16.03
13.52
17.12
5.17
5.40
Standard care group
Cobedding group
27.02
20.66
8.79
5.76
25.69
23.73
6.93
9.56
22.52
18.70
7.23
5.38
Standard care group
Cobedding group
Standard care group
Cobedding group
Standard care group
Cobedding group
Standard care group
Cobedding group
Standard care group
Cobedding group
13.14
10.22
93.57
93.33
89.57
91.78
128.28
127.11
145.42
136.55
1.77
2.90
2.14
1.65
2.14
2.22
7.76
8.59
10.87
9.58
9.77
8.00
93.44
92.64
89.94
91.23
126.61
125.58
139.89
133.52
2.83
3.02
1.78
1.36
2.29
1.85
9.13
7.12
10.71
7.87
8.80
7.29
93.16
93.59
89.48
92.16
123.20
124.7
137.80
132.54
1.87
2.15
1.86
1.31
2.52
1.65
7.82
7.83
9.59
7.82
PIPP Z premature infant pain profile; SD Z standard deviation.
salivary cortisol levels and behavioral responses to pain
following minor painful procedures were also reported
after exposure to familiar odors.11,22 We postulated
that attenuated stress responses to pain after cobedding
were partially due to the calming effects of a familiar odor
and SSC.
Our results demonstrated that cobedding could significantly decrease the crying time after heel sticking.
However, even after cobedding, the mean duration of
crying was still high at 36 seconds. Goubet et al23
demonstrated a reduced crying time during painful procedures in newborns who were presented with familiar
Table 4 Mean SD of the PIPP score, crying time, and salivary cortisol before and after heel lancing in two groups of
cobedding and standard care, based on two subgroups of gestational age.
28e32 wk
N Z 18
Duration of crying (s)
Salivary cortisol
1 min prior to heel lancing (nmol/L)
Salivary cortisol
20 min after heel lancing (nmol/L)
PIPP score
Baseline O2sat
Minimum of O2sat after heel lancing
Baseline heart rate
Maximum of heart rate during heel lancing
Standard care group
Cobedding group
Standard care group
Cobedding group
Standard care group
Cobedding group
Standard care group
Cobedding group
Standard care group
Cobedding group
Standard care group
Cobedding group
Standard care group
Cobedding group
Standard care group
Cobedding group
PIPP Z premature infant pain profile; SD Z standard deviation.
32e34 wk
N Z 32
Mean
SD ()
Mean
SD ()
57.83
55.00
22.06
21.65
27.75
23.93
11.72
10.33
93.61
93.22
90.00
91.89
127.00
127.11
142.27
135.55
20.12
7.39
7.67
7.01
8.85
3.01
3.01
2.22
1.69
2.65
2.11
9.02
8.20
11.78
8.80
34.12
25.93
16.50
16.34
22.35
18.98
8.65
6.78
93.15
93.22
89.47
91.72
124.09
124.50
138.12
132.50
12.76
14.13
5.13
5.26
7.02
5.78
1.62
1.58
1.60
1.28
2.39
1.70
7.97
7.23
9.22
7.44
Cobedding of twin premature infants
267
odors, which is consistent with our results. However,
familiar odors alone cannot eliminate crying completely.
Therefore, using other pain-reducing interventions may
be necessary during heel lancing in premature infants to
reduce pain further.
Because of the small sample size, we could not assess
the incidence of infection, necrotizing enterocolitis,
apnea, and bradycardia in two controls. However, LaMar
and Dowling24 evaluated the incidence of infection in twin
preterm infants between 23 weeks and 35 weeks of
gestational age who were cared for in cobedding in the
NICUs, and concluded that cobedding was not associated
with an increased incidence of sepsis, pneumonia, and
necrotizing enterocolitis. Byers and coworkers13 had
similar findings.
A small number of studies evaluated the calming effects
of cobedding in preterm twins. The importance of this
study is that, to the best of our knowledge, it is the first
study to assess the soothing effects of cobedding without
using any other pain-reducing intervention during minor
procedures in preterm twins.
At present, demonstrating significant differences on
certain variables may be inadequate, due to small sample
size.
Cobedding appears to be a comforting measure for twin
premature infants during heel lancing, which has no significant adverse effects. This pain control method can
decrease the duration of crying, oxygen desaturation, PIPP
score, and physiological stress responses to pain after minor
painful procedures.
Acknowledgments
Steps in pain assessment
1. Familiarize yourself with each indicator and how it is to
be scored, by looking at the PIPP.
2. Score the gestational age prior to when you begin the
assessment (points are added to the premature infant’s
pain score based on gestational age to compensate for
their limited ability to behaviorally and physiologically
respond to pain).
3. Score behavioral state by observing the infant for 30
seconds.
4. Record the baseline heart rate and oxygen saturation at
the beginning of the shift.
5. Observe the infant for 30 seconds. You will need to look
back and forth from the heart monitor to the baby’s
face. Score physiological and facial changes observed
during that time and record them immediately following
the observation period.
6. Calculate the total score.
7. Scores of 0e6 generally indicate that an infant has
minimal or no pain.
8. Scores of 7e12 generally indicate slight to moderate
pain.
9. Scores of >12 may indicate severe pain.
Conflicts of interest
The authors have no conflicts of interest relevant to this
article.
References
The authors thank all staff and nurses of Shahid Beheshti
University Hospital, Isfahan, Iran for their assistance and
support for this project.
Appendix 1
1. Derom R, Orlebeke J, Eriksson A, Thiery M. The epidemiology of
multiple births in Europe. In: Keith LG, Papiernik E, Keith DM,
Luke B, editors. Multiple pregnancy: epidemiology, gestation
and perinatal outcome. New York: Parthenon; 1995. p. 145e62.
PIPP pain assessment tool.
Process
Indicator
0
1
2
3
Chart
Gestational age
36
32e35 wk,
6d
28e31 wk,
6d
<28 wk
Observe infant for 15 s
Behavioral state
Heart rate
maximum
Oxygen saturation
minimum
Brow bulge
Eye squeeze
Nasoelabial furrow
Quiet awake,
eyes open,
no facial
movements
5e15 bpm
increase
89e91%
Active sleep,
eyes closed,
facial movements
Observe baseline heart
rate and oxygen
saturation for 30 s
Active, awake,
eyes open,
facial
movements
0e5 bpm
increase
92e100%
15e24 bpm
increase
85e88%
Quiet sleep,
eyes closed,
no facial
movements
25 bpm
increase
<85%
None
None
None
Minimum
Minimum
Minimum
Moderate
Moderate
Moderate
Maximum
Maximum
Maximum
Observe infant facial
actions for 30 s
Total score
bpm Z beats per minute; PIPP Z premature infant pain profile.
Note. From “Premature infant pain profile: development and initial validation” by B. Stevens, C. Johnston, P. Petryshen, and A. Taddio,
1996, Clinical Journal of Pain, 12, p. 13e22. Copyright ª 2014, Lippincott Williams & Wilkins. All rights reserved. Reprinted with
permission.
268
2. Stoelhorst GM, Rijken M, Martens SE, Brand R, den Ouden AL,
Wit JM, et al. Changes in neonatology: comparison of two cohorts of very preterm infants (gestational age <32 weeks): the
project on preterm and small for gestational age infants 1983
and the Leiden follow-up project on prematurity 1996e1997.
Pediatrics 2005;115:396e405.
3. Blondel B, Kogan MD, Alexander GR, Dattani N, Kramer MS,
Macfarlane A, et al. The impact of the increasing number of
multiple births on the rates of preterm birth and low birth
weight: an international study. Am J Public Health 2002;92:
1323e30.
4. Joseph KS, Kramer MS, Marcoux S, Ohlsson A, Wen SW, Allen A,
et al. Determinants of preterm birth rates in Canada from 1981
through 1983 and from 1992 through 1994. N Engl J Med 1998;
339:1434e9.
5. Carbajal R, Rousset A, Danan C, Coquery S, Nolent P,
Ducrocq S, et al. Epidemiology and treatment of painful procedures in neonates in intensive care units. JAMA 2008;300:
60e70.
6. Grunau R. Early pain in preterm infants: a model of long-term
effects. Clin Perinatol 2002;29:373e94.
7. Fitzgerald M, Millard C, McIntosh N. Cutaneous hypersensitivity following peripheral tissue damage in newborn infants and its reversal with topical anaesthesia. Pain 1989;
39:31e6.
8. Simons SH, van Dijk M, Anand KS, Roofthooft D, van Lingen RA,
Tibboel D. Do we still hurt newborn babies? A prospective study
of procedural pain and analgesia in neonates. Arch Pediatr
Adolesc Med 2003;157:1058e64.
9. American Academy of Pediatrics Committee on Fetus and
Newborn, American Academy of Pediatrics Section on Surgery,
Canadian Paediatric Society Fetus and Newborn Committee,
Batton DG, Barrington KJ, Wallman C. Prevention and management of pain in the neonate: an update. Pediatrics 2006;
118:2231e41.
10. Ballantyne M, Stevens B, McAllister M, Dionne K, Jack A. Validation of the premature infant pain profile in the clinical
setting. Clin J Pain 1999;15:297e303.
11. Nishitani S, Miyamura T, Tagawa M, Sumi M, Takase R,
Doi H, et al. The calming effect of a maternal breast milk
odor on the human newborn infant. Neurosci Res 2009;63:
66e71.
12. Hayward K. Cobedding of twins: a natural extension of the
socialization process? MCN Am J Matern Child Nurs 2003;28:
260e3.
Z. Badiee et al
13. Byers JF, Yovaish W, Lowman LB, Francis JD. Co-bedding
versus single-bedding premature multiple-gestation infants
in incubators. J Obstet Gynecol Neonatal Nurs 2003;32:
340e7.
14. DellaPorta K, Aforismo D, Butler-O’Hara M. Co-bedding of
twins in the neonatal intensive care unit. Pediatr Nurs 1998;
24:529e31.
15. Miller C. Co-bedding improves outcomes for premature twins.
Adv Respir Care Pract 1998;12:17e8.
16. Lai MN, Foong SC, Foong WC, Tan K. Co-bedding in neonatal
nursery for promoting growth and neurodevelopment in stable
preterm twins. Cochrane Database Syst Rev 2012;12:
CD008313. http://dx.doi.org/10.1002/14651858.CD008313.pub2 [Review].
17. Campbell-Yeo ML, Johnston CC, Joseph KS, Feeley N,
Chambers CT, Barrington KJ. Cobedding and recovery time
after heel lance in preterm twins: results of a randomized trial.
Pediatrics 2012;130:500e6.
18. Chin SD, Hope L, Christos PJ. Randomized controlled trial
evaluating the effects of cobedding on weight gain and physiologic regulation in preterm twins in the NICU. Adv Neonatal
Care 2006;6:142e9.
19. Johnston CC, Stevens B, Pinelli J, Gibbins S, Filion F, Jack A,
et al. Kangaroo care is effective in diminishing pain response in
preterm neonates. Arch Pediatr Adolesc Med 2003;157:
1084e8.
20. Campbell-Yeo ML, Johnston CC, Joseph K, Feeley NL,
Chambers CT, Barrington KJ. Co-bedding as a comfort measure
for twins undergoing painful procedures (CComForT Trial). BMC
Pediatr 2009;9:76.
21. Cong X, Ludington-Hoe SM, Walsh S. Randomized crossover
trial of kangaroo care to reduce biobehavioral pain responses
in preterm infants: a pilot study. Biol Res Nurs 2011;13:
204e16.
22. Axelin A, Salantera
¨ S, Kirjavainen J, Lehtonen L. Oral
glucose and parental holding preferable to opioid in pain
management in preterm infants. Clin J Pain 2009;25:
138e45.
23. Goubet N, Rattaz C, Pierrat V, Bullinger A, Lequien P. Olfactory
experience mediates response to pain in preterm newborns.
Dev Psychobiol 2003;42:171e80.
24. LaMar K, Dowling DA. Incidence of infection for preterm twins
cared for in cobedding in the neonatal intensive-care unit.
J Obstet Gynecol Neonatal Nurs 2006;35:193e8.
Pediatrics and Neonatology (2014) 55, 269e274
Available online at www.sciencedirect.com
ScienceDirect
journal homepage: http://www.pediatr-neonatol.com
ORIGINAL ARTICLE
Sonographic Finding of Persistent Renal
Pelvic Wall Thickening in Children
Nai-Chia Fan, You-Lin Tain*
Division of Pediatric Nephrology, Kaohsiung Chang Gung Memorial Hospital and Chang Gung University,
College of Medicine, Kaohsiung, Taiwan
Received Jan 16, 2013; received in revised form Sep 18, 2013; accepted Oct 15, 2013
Available online 25 December 2013
Key Words
renal pelvic wall
thickening;
ultrasound;
vesicoureteral reflux;
voiding
cystourethrography
Objective: Renal pelvic wall thickening (RPWT) is a sonographic finding, which is associated
with urinary tract infection (UTI) and other genitourinary tract abnormalities. We aimed to
determine the prevalence of RPWT and whether persistent RPWT related to vesicoureteral reflux (VUR).
Materials and methods: We retrospectively reviewed sonographic findings of RPWT in children
and adolescents referred for renal ultrasound study from January 2010 to December 2011. A
total of 502 patients showing RPWT were included, 372 of whom received follow-up sonograms.
Among them, 86 children underwent both follow-up sonograms and voiding cystourethrography
studies. The association between persistent RPWT and VUR was analyzed.
Results: A total of 602 sonograms with RPWT were identified, accounting for a prevalence of
11.4%. Follow-up sonograms, revealing that these patients had recovered from RPWT, was
found in 93.7% (459/490) of renal units and in 92.7% (345/372) of the patients. Children with
persistent RPWT had a strong association with VUR occurrence (p Z 0.018) and high VUR
grading (p Z 0.006) compared to those without persistent RPWT.
Conclusion: RPWT is a common finding in children and adolescents. Persistent RPWT is associated with VUR, especially with high grade VUR. Complementary uroradiological studies should
be performed for children and adolescents with persistent RPWT.
Copyright ª 2013, Taiwan Pediatric Association. Published by Elsevier Taiwan LLC. All rights
reserved.
* Corresponding author. Department of Pediatrics, Kaohsiung Chang Gung Memorial Hospital, 123 Dabi Road, Niausung, Kaohsiung 833,
Taiwan.
E-mail addresses: [email protected], [email protected] (Y.-L. Tain).
1875-9572/$36 Copyright ª 2013, Taiwan Pediatric Association. Published by Elsevier Taiwan LLC. All rights reserved.
http://dx.doi.org/10.1016/j.pedneo.2013.10.007
270
N.-C. Fan, Y.-L. Tain
1. Introduction
Renal sonography plays an important role in acute urinary
tract infection (UTI) for the detection of genitourinary tract
abnormalities, especially in young infants.1,2 Renal pelvic
wall thickening (RPWT) is a sonographic finding associated
with UTI, urolithiasis, rejection after renal transplantation,
vesicoureteral reflux (VUR), and obstruction.3e9 We aimed
to study the prevalence of RPWT in a pediatric population
in a medical center and to determine the association between persistent RPWT and VUR.
2. Materials and Methods
We retrospectively analyzed sonographic findings of RPWT
from January 2010 to December 2011 in our hospital. All
children and adolescents aged <18 years, both hospitalized
patients and outpatients, referred to our hospital for renal
ultrasound study, were reviewed. During the study period,
patients with renal ultrasound reports showing RPWT were
included. The inclusion criteria were age <18 years and
renal echo reports showing RPWT. Among these, patients
who had both successively repeated renal ultrasound and
voiding cystourethrography (VCUG) were finally analyzed
for an association between persistent RPWT and VUR. Other
uroradiological study results, including intravenous pyelogram and dimercaptosuccinic acid, were recorded as
well.10 Indications for repeated sonographic study included
poor response to initial treatment and abnormalities identified by the first ultrasound study. These abnormalities
included RPWT, dilated pelvis, dilated ureter, nephromegaly, focal renal mass, increased renal echogenicity,
renal size discrepancy, ectopic kidney, and renal abscess.
All hospitalized patients underwent renal ultrasound on
the 1st day or 2nd day after hospitalization. Renal ultrasound
was performed with patients in the prone position, with the
transducer in horizontal and transverse sections to the renal
pelvis. RPWT was detected in both horizontal and transverse
views. RPWT was defined as thickening measuring 1 mm as
described previously.4 This sonographic sign indicated a
hypoechoic rim within the renal pelvic wall surrounded by
increased mucosal hyperechogenicity (Figure 1). All ultrasound studies were performed with the same equipment (HP
Image Point HX Ultrasound System, Philips Medical Systems,
Andover, MA, USA), with an HP 21373 A curved ultrasound
transducer under 3.5e7.5 MHz (HP Image Point HX Ultrasound System, Philips Medical Systems, Andover, MA, USA).
All results were reviewed by one pediatric nephrologist (YL
Tain) to avoid intra- and inter-operator variations. If thickening was persistent in the renal pelvic wall on successive
ultrasound studies (the second renal sonography), this finding
was defined as persistent RPWT. VCUG was not performed
routinely after the first febrile UTI.2 VCUG was indicated if
the renal sonogram revealed a dilated pelvis, dilated ureter,
and suspected duplex systems, which would suggest either
high-grade VUR or obstructive uropathy, as well as in cases
with recurrent UTI. VUR was detected by VCUG and classified
following the international VUR grading system.11
Not all enrolled patients received repeated ultrasound
studies. Only those with repeated ultrasound studies were
classified into two groups: patients with and without
Figure 1 Prone longitudinal scan of the kidney showing renal
pelvic wall thickening. The two white crosses indicate renal
pelvis and the two white arrows indicate hypoechoic rim surrounded by increased mucosal hyperechogenicity.
persistent RPWT. These two groups were further subgrouped depending on the presence of VUR. Underlying
diseases of these patients were recorded through chart
review approved by the institutional review board (IRB) at
Chang Gung Memorial Hospital, Kaohsiung, Taiwan. Data
were presented as mean standard deviation (SD). Chisquare analysis or Fisher’s exact test, t test, and binary
logistic regression were used to compare nominal data between the two groups by means of commercially available
software (SPSS 14.0; SPSS Inc., Chicago, IL, USA). A p value
<0.05 was considered statistically significant.
3. Results
A total of 5299 renal ultrasound studies from January 2010
to December 2011 were reviewed. Among them, 602 sonograms with RPWT were identified accounting for a prevalence of 11.4%. In the 602 renal ultrasound studies, there
were 502 patients, including 255 males and 247 females.
The mean age was 26.5 43 months (aged between
0.1 months and 209 months). Initial presentations of these
502 patients included febrile episodes (n Z 451, 89.8%),
pyuria (n Z 372, 74.1%), and gross hematuria (n Z 3, 0.6%).
Febrile UTI was diagnosed in 235 cases (46.8%) and nonfebrile UTI was diagnosed in four cases (0.8%). The common pathogens of UTI (n Z 239) included Escherichia coli
(n Z 159, 66.5%), E. coli extended-spectrum b-lactamases
strain (n Z 64, 26.8%), Proteus mirabilis (n Z 4, 1.7%),
Klebsiella pneumoniae (n Z 4, 1.7%), Pseudomonas aeruginosa (n Z 2, 0.8%), and others (n Z 6, 2.5%).
Of 502 patients, 372 patients (74.1%) had repeated renal
sonograms. The follow-up renal sonographies were arranged for the abnormalities in the first echo reports. These
372 patients had a total of 490 renal units with RPWT at
their initial renal ultrasound studies. A renal unit presented
each kidney separately. There was an average of
16.8 18.9 days between the repeated performances of
renal ultrasound in all enrolled patients. More than 75% of
patients received repeated renal ultrasound study in
Persistent renal pelvic wall thickening
Table 1
271
Demographic and clinical characteristics of patients with follow-up renal ultrasound.
Sex, n (%)
M
F
Age at enrollment
(months), n (%)
12
>12
Mean age in months SD
(range)
UTI, n (%)
Recurrent UTI, n (%)
Having VCUG, n (%)
Total (n Z 372)
With persistent RPWT
(n Z 27)
Without persistent
RPWT (n Z 345)
181 (48.7)
191 (51.3)
9 (33.3)
18 (66.7)
172 (49.9)
173 (50.1)
p value*
0.11
<0.001*
253 (68.0)
119 (32.0)
24.2 39.7 (0.1e205)
6 (22.2)
21 (77.8)
63.3 58.6 (3.0e205)
247 (71.6)
98 (28.4)
22.5 37.8 (0.1e192)
190 (51.1)
24 (6.5)
86 (23.1)
9 (31.3)
7 (25.0)
14 (51.9)
181 (52.5)
17 (2.2)
72 (20.9)
0.07
0.001*
p values calculated using chi-square test for categorical data, and t test for continuous data (e.g. mean age in months).
*p < 0.05, statistic significance.
RPWT Z renal pelvic wall thickening; UTI Z urinary tract infection; VCUG Z voiding cystourethrography.
underlying disease, and one patient who had received
VCUG study prior to this study.
Among 345 patients without persistent RPWT, 72 (20.9%)
underwent VCUG study. Twenty-six of 72 patients had VUR.
The characteristics of patients with VUR, either with or
without persistent RPWT, are shown in Table 2. The detection
rates of VUR in patients with and without persistent RPWT
were 79% (11/14) and 36% (26/72), respectively. The persistent RPWT group showed a higher VUR occurrence (p Z 0.006)
and higher VUR grading (VUR grading III w V; p Z 0.004)
compared to those in the group without persistent RPWT
(Table 2). We further studied the relationship between VUR
and abnormal renal sonographic findings, including persistent
RPWT, small or hypoplastic kidney, hydronephrosis, pelviectasis, ureter dilatation, globular shaped kidney, nephromegaly, and urinary bladder wall thickening in patients
with repeated renal ultrasound and VCUG. As shown in
20 days. Disappearance of the PWT occurred in 93.7% (459/
490) of renal units and in 92.7% (345/372) of the patients
who had repeated renal sonograms.
Among 372 patients with repeated renal sonograms, 27
children had persistent RPWT and 345 children did not have
persistent RPWT. Children with persistent RPWT were older
at enrollment (p < 0.001) and had a higher recurrent UTI
rate (p Z 0.001) than children without persistent RPWT
(Table 1). There was no difference in sex or UTI occurrence
between the two groups. Among 27 children with persistent
RPWT, 14 had received VCUG study. Eleven out of these 14
cases showed VUR.
The other 16 children who did not receive VCUG study
(n Z 13) or had absence of VUR (n Z 3), included 10 patients with repeated renal ultrasound performed within 7
days (median 4 days, range 1e11 days), three patients with
neurogenic bladder, two patients without significant
Table 2
Demographic and clinical characteristics of patients with follow-up renal ultrasound and voiding cystourethrography.
With VUR, n (%)
VUR grading
I
II
III
IV
V
Characteristics
Recurrent UTI
CAKUT
Status post deflux
No underlying diseases
DMSA, n (%)
With cold area, n (%)
Total (n Z 86)
With persistent
RPWT (n Z 14)
Without persistent
RPWT (n Z 72)
p value*
37 (43.0)
11 (78.6)
26 (36.1)
0.006*
8
8
11
4
6
(9.3)
(9.3)
(12.8)
(4.7)
(7.0)
2
1
2
2
4
(14.3)
(7.1)
(14.3)
(14.3)
(28.6)
6
7
9
2
2
(8.3)
(9.7)
(12.5)
(2.8)
(2.8)
12
5
3
17
14
7
(14.0)
(5.8)
(3.5)
(19.7)
(16.3)
(8.1)
5
3
1
2
5
4
(35.7)
(21.4)
(7.1)
(14.3)
(35.7)
(28.6)
7
2
2
15
9
3
(9.7)
(2.8)
(2.8)
(20.9)
(12.5)
(4.2)
0.004y,*
*p < 0.05, statistic significance.
RPWT Z renal pelvic wall thickening; VUR Z vesicoureteral reflux; UTI Z urinary tract infection; CAKUT Z congenital anomalies of the
kidney and urinary tract; DMSA Z dimercaptosuccinic acid renal scan.
y
Present p value of these two groups in high grade VUR (VUR grading IIIwV).
272
Table 3
N.-C. Fan, Y.-L. Tain
Factors associated with VUR (total renal units, n Z 169 RUs).
Characteristics
Renal US presentation,
n (%)
Abnormal
Normal
Small or hypoplastic
kidney, n (%)
Yes
No
Hydronephrosis, n (%)
Yes
No
Pelviectasis, n (%)
Yes
No
Ureter dilatation, n (%)
Yes
No
Globular shaped
kidney, n (%)
Yes
No
Nephromegaly, n (%)
Yes
No
Urinary bladder
wall thickening, n (%)
Yes
No
Persistent RPWT, n (%)
Yes
No
Age (months), n (%)
12
>12
Sex, n (%)
Male
Female
UTI recurrence, n (%)
Yes
No
Non-VUR
(n Z 121)
VUR
(n Z 48)
Unadjusted OR for
being VUR (95%CI)
p value
Adjusted OR for
being VUR (95% CI)
0.3
83 (68.6)
38 (31.4)
37 (77.1)
11 (22.9)
1.5 (0.7 to 3.3)
1 (Reference)
0.7
0.8 (0.3 to 2.5)
1 (Reference)
0.02*
2 (1.7)
119 (98.3)
5 (10.4)
43 (89.6)
6.9 (1.3 to 37.0)
1 (Reference)
5 (4.1)
116 (95.9)
5 (10.4)
43 (89.6)
2.7 (0.7 to 9.8)
1 (Reference)
54 (44.6)
67 (55.4)
21 (43.8)
27 (46.2)
1.0 (0.5 to 1.9)
1 (Reference)
47 (38.8)
74 (61.2)
19 (39.6)
29 (60.4)
1.0 (0.5 to 2.0)
1 (Reference)
0.006*
13.8 (2.1 to 90.2)
1 (Reference)
0.1
0.08
4.4 (0.8 to 23.9)
1 (Reference)
0.9
0.4
1.6 (0.5 to 5.2)
1 (Reference)
0.9
0.4
0.6 (0.2 to 1.8)
1 (Reference)
0.7
20 (16.5)
101 (83.5)
9 (18.8)
39 (81.2)
1.2 (0.5 to 2.8)
1 (Reference)
10 (8.3)
111 (91.7)
5 (10.4)
43 (89.6)
1.3 (0.4 to 4.0)
1 (Reference)
0.6
1.3 (0.5 to 3.7)
1 (Reference)
0.7
0.7
1.3 (0.4 to 4.6)
1 (Reference)
0.2
3 (2.5)
118 (97.5)
3 (6.3)
45 (93.7)
2.6 (0.5 to 13.5)
1 (Reference)
6 (5.0)
115 (95.0)
9 (74.4)
39 (25.6)
4.4 (1.5 to 13.2)
1 (Reference)
70 (57.9)
51 (42.1)
21 (43.8)
27 (56.2)
0.7 (0.3 to 1.6)
1 (Reference)
32 (26.4)
89 (73.6)
11 (22.9)
37 (77.1)
0.8 (0.4 to 1.8)
1 (Reference)
11 (9.1)
110 (90.9)
8 (16.7)
40 (83.3)
2.0 (0.8 to 5.3)
1 (Reference)
p value
0.2
3.2 (0.6 to 18.0)
1 (Reference)
0.008*
0.018*
4.5 (1.3 to 15.9)
1 (Reference)
0.4
0.5
0.8 (0.4 to 1.7)
1 (Reference)
0.6
0.3
0.6 (0.2 to 1.7)
1 (Reference)
0.2
0.7
1.3 (0.4 to 4.2)
1 (Reference)
Significant results are highlighted in bold.
OR Z odds ratio, each OR of the variable was adjusted by the other factors in the logistic regression model.
*p < 0.05 indicates a statistically significant difference.
RU Z renal unit; VUR Z vesicoureteral reflux; US Z ultrasound; RPWT Z renal pelvic wall thickening; UTI Z urinary tract infection.
Table 3, small/hypoplastic kidney and persistent RPWT had
significant associations with VUR. Moreover, persistent RPWT
was the only sonographic finding to be associated with high
grade VUR [Table 4; adjusted odds ratio (OR) Z 8.9, 95%
confidence interval (CI) Z 2.3e34.9, p Z 0.002].
4. Discussion
Our major findings are as follows: (1) RPWT is a common
sonographic finding in the pediatric population; and (2)
persistent RPWT has a significant association with high
grade VUR compared to the other abnormal renal ultrasound findings. Previously, ultrasonic studies have focused
on associated diseases of RPWT whereas we were more
concerned with elucidating the right time to detect RPWT
occurrence and the role of persistent RPWT. Previous reports showed that RPWT signs were related to UTI, urolithiasis, rejection after renal transplantation, VUR, and
obstruction.3e9 In this study, we found that persistent
RPWT signs in consecutive renal ultrasound had a strong
association with VUR, particularly high-grade VUR.
Persistent renal pelvic wall thickening
Table 4
273
Factors associated with high grade VUR (total renal units, n Z 169 RUs).
Characteristics
Renal US presentation,
n (%)
Abnormal
Normal
Small or hypoplastic
kidney, n (%)
Yes
No
Hydronephrosis, n (%)
Yes
No
Pelviectasis, n (%)
Yes
No
Ureter dilatation, n (%)
Yes
No
Globular shaped
kidney, n (%)
Yes
No
Nephromegaly, n (%)
Yes
No
Urinary bladder
wall thickening, n (%)
Yes
No
Persistent RPWT, n (%)
Yes
No
Age (months), n (%)
12
>12
Sex, n (%)
Male
Female
UTI recurrence, n (%)
Yes
No
Non-high grade
VUR (n Z 141)
High grade VUR
(n Z 28)
Unadjusted OR
for being high
grade VUR (95% CI)
p value
Adjusted OR for
being high grade
VUR (95% CI)
0.2
97 (68.8)
44 (31.2)
23 (82.1)
5 (17.9)
2.0 (0.7 to 5.8)
1 (Reference)
0.5
1.5 (0.4 to 5.9)
1 (Reference)
0.4
5 (3.5)
136 (96.5)
2 (7.1)
26 (92.9)
2.1 (0.4 to11.4)
1 (Reference)
7 (5.0)
134 (95.0)
3 (10.7)
25 (89.3)
2.3 (0.6 to 9.5)
1 (Reference)
63 (44.7)
78 (55.3)
12 (42.9)
16 (57.1)
0.9 (0.4 to 2.1)
1 (Reference)
54 (38.3)
87 (61.7)
12 (82.8)
16 (8.6)
1.2 (0.5 to 2.7)
1 (Reference)
0.3
2.9 (0.4 to 20.1)
1 (Reference)
0.3
0.7
1.4 (0.2 to 10.1)
1 (Reference)
0.9
0.6
0.7 (0.2 to 2.9)
1 (Reference)
0.7
0.9
0.9 (0.2 to 3.5)
1 (Reference)
0.5
23 (16.3)
118 (83.7)
6 (42.9)
22 (57.1)
1.4 (0.5 to3.8)
1 (Reference)
12 (8.5)
129 (91.5)
3 (10.7)
25 (89.3)
1.3 (0.3 to 4.9)
1 (Reference)
0.8
1.1 (0.3 to 3.4)
1 (Reference)
0.7
0.8
1.2 (0.3 to 5.5)
1 (Reference)
1.0
5 (3.5)
136 (96.5)
1 (3.6)
27 (96.4)
1.0 (0.1 to 9.0)
1 (Reference)
7 (5.0)
134 (95.0)
8 (82.8)
20 (8.6)
7.7(2.5 to 23.4)
1 (Reference)
78 (55.3)
63 (44.7)
13 (46.4)
15 (53.6)
0.6 (0.3 to 1.1)
1 (Reference)
35 (24.8)
106 (75.2)
8 (28.6)
20 (71.4)
1.2 (0.5 to 3.0)
1 (Reference)
15 (10.6)
126 (89.4)
4 (14.3)
24 (85.7)
1.4 (0.4 to 4.6)
1 (Reference)
p value
0.9
1.2 (0.1 to 11.8)
1 (Reference)
<0.001*
0.002*
8.9 (2.3 to 34.9)
1 (Reference)
0.1
0.8
0.9 (0.3 to 2.4)
1 (Reference)
0.7
1.0
1.0 (0.3 to 2.8)
1 (Reference)
0.6
0.4
0.5 (0.1 to 2.5)
1 (Reference)
Significant results are highlighted in bold.
OR Z odds ratio, each OR of the variable was adjusted by the other factors in the logistic regression model.
*p < 0.05 indicates a statistically significant difference.
RU Z renal unit; VUR Z vesicoureteral reflux; US Z ultrasound; RPWT Z renal pelvic wall thickening; UTI Z urinary tract infection.
It is noteworthy that most cases with RPWT were
detected during the acute phase of infection. In our hospitalized children, the first renal ultrasound was performed
within 3.8 3.4 days from the onset of fever. The follow-up
renal ultrasound was performed 16.8 18.9 days after the
first renal ultrasound. More than 90% of patients who
received follow-up renal ultrasound recovered from RPWT.
For UTI and non-UTI children in this study, the recovery rate
of RPWT was 95.3% in UTI patients and 90.1% in non-UTI
patients. Since the recovery rate of RPWT in UTI and nonUTI children was similar, it seemed that an acute UTI
episode was not the leading cause for the occurrence of
persistent RPWT. In addition, our patients did not present
urolithiasis or rejection after renal transplantation. The
trend for a high recovery rate of RPWT may come from
quickly resolved acute infections, acute inflammation or
obstruction in the majority of patients. Thus, RPWT was
considered to be an acute phase sign. If RPWT lasted longer
than the usual duration (2 weeks) after the first renal ultrasound, it suggested that VUR or genitourinary tract
obstruction should be evaluated. Given that most patients
recovered from RPWT after their clinical symptoms began
274
to improve, it is necessary to perform renal ultrasound as
early as possible to better understand the timing of RPWT
occurrence. The RPWT occurrence pattern in this study
showed a large variation, due to some patients receiving
follow-up renal ultrasound at different intervals. Thus, a
short interval follow-up may provide more information
about the recovery duration of RPWT.
In this study, we further compared persistent RPWT with
other abnormal findings on renal ultrasound which were
studied and proved to have a positive correlation to VUR
previously.7,12 Next, confounders including children’s age,
sex, and UTI recurrence9 were put together with the above
factors into logistic regression, to detect their influence on
each other and the association with VUR. Small kidney or
hypoplastic kidney was associated with VUR (adjusted
OR Z 13.8, CI Z 2.1e90.2, p Z 0.006). The adjusted OR of
children with persistent RPWT with VUR is 4.5 times as large
as those of children without persistent RPWT. Moreover, the
adjusted OR of children with persistent RPWT with high
grade VUR is 8.9 times as large as those of children without
persistent RPWT (CI Z 2.3e34.9, p Z 0.002). Our results
indicated that only persistent RPWT was significantly associated with high grade VUR compared to the other factors.
The sensitivity of renal ultrasound to first febrile UTI in
detecting VUR and renal parenchymal defects remains
questionable.13e16 Although RPWT is one of the abnormal
findings in renal sonograms indicating the possibility of
VUR,8,13 persistent RPWT has a strong association with VUR,
especially with high grade VUR. In this study, we observed
that three children in the persistent RPWT group had VUR.
All received deflux procedures later. Their RPWT recovered
following correction of their VUR. RPWT has been proposed
to be due to subepithelial edema, acute inflammatory
infiltration, fibrosis, or smooth muscle hypertrophy, followed by chronic inflammation.5,17,18 Our data suggest that
VUR may cause RPWT by increasing renal pelvic pressure,
similar to the mechanism of genitourinary obstruction.
Our study has some limitations. First, ultrasound study is
an operator-dependent technique. Although each renal
sonogram was reviewed by one pediatric nephrologist, and
RPWT is a common parameter in our hospital, we did not
measure inter- and intra-operator variations in this study.
Another limitation is the small sample size in the subgroups.
As mentioned above, VCUG is not a routine procedure to
perform after each UTI episode. There may be some missing
data in those patients without VCUG study.
In conclusion, RPWT is a common abnormal finding of
renal sonograms in the pediatric population. Persistent
RPWT is the only factor that correlates with VUR, especially with high-grade VUR. Most UTI-related RPWT recovered after about 2 weeks follow-up. Complementary
uroradiological studies could be considered if this sign is
persistent.
Conflict of interest
None declared.
N.-C. Fan, Y.-L. Tain
References
1. Mori R, Lakhanpaul M, Verrier-Jones K. Diagnosis and management of urinary tract infection in children: summary of
NICE guidance. BMJ 2007;335:395e7.
2. Subcommittee on Urinary Tract Infection. Steering Committee
on Quality Improvement and Management, Roberts KB. Urinary
tract infection: clinical practice guideline for the diagnosis and
management of the initial UTI in febrile infants and children 2
to 24 months. Pediatrics 2011;128:595e610.
3. Avni EF, Van Gansbeke D, Thoua Y, Matos C, Marconi V,
Lemaitre L, et al. US demonstration of pyelitis and ureteritis in
children. Pediatr Radiol 1988;18:134e9.
4. Tain YL. Renal pelvic wall thickening in childhood urinary tract
infectionseevidence of acute pyelitis or vesicoureteral reflux?
Scand J Urol Nephrol 2003;37:28e30.
5. Nicolet V, Carignan L, Dubuc G, He
´bert G, Bourdon F, Paquin F.
Thickening of the renal collecting system: a nonspecific finding
at US. Radiology 1988;168:411e3.
6. Peters C, Rushton HG. Vesicoureteral reflux associated renal
damage: congenital reflux nephropathy and acquired renal
scarring. J Urol 2010;184:265e73.
7. Pennesi M, L’erario I, Travan L, Ventura A. Managing children
under 36 months of age with febrile urinary tract infection: a
new approach. Pediatr Nephrol 2012;27:611e5.
8. Sorantin E, Fotter R, Aigner R, Ring E, Riccabona M. The
sonographically thickened wall of the upper urinary tract system: correlation with other imaging methods. Pediatr Radiol
1997;27:667e71.
9. Mitterberger M, Pinggera GM, Feuchtner G, Neururer R,
Bartsch G, Gradl J, et al. Sonographic measurement of renal
pelvis wall thickness as diagnostic criterion for acute pyelonephritis in adults. Ultraschall Med 2007;28:593e7.
10. Paterson A. Urinary tract infection: an update on imaging
strategies. Eur Radiol 2004;14:L89e100.
11. Lebowitz RL, Olbing H, Parkkulainen KV, Smellie JM, Tamminen-Mo
¨bius TE. International system of radiographic grading of
vesicoureteric reflux. International Reflux Study in Children.
Pediatr Radiol 1985;15:105e9.
12. Lee JH, Kim MK, Park SE. Is a routine voiding cystourethrogram
necessary in children after the first febrile urinary tract
infection? Acta Paediatr 2012;101:e105e9.
13. Alton DJ, LeQuesne GW, Gent R, Siegmann JW, Byard R.
Sonographically demonstrated thickening of the renal pelvis in
children. Pediatr Radiol 1992;22:426e9.
14. Jahnukainen T, Honkinen O, Ruuskanen O, Mertsola J. Ultrasonography after the first febrile urinary tract infection in
children. Eur J Pediatr 2006;165:556e9.
15. Mahant S, Friedman J, MacArthur C. Renal ultrasound findings
and vesicoureteral reflux in children hospitalised with urinary
tract infection. Arch Dis Child 2002;86:419e20.
16. Smellie JM, Rigden SP, Prescod NP. Urinary tract infection: a
comparison of four methods of investigation. Arch Dis Child
1995;72:247e50.
17. Han SW, Maizels M, Chou PM, Fernback SK, Cheng EY,
Furness PD 3rd. Lamina muscularis propria thickness of renal
pelvis predicts radiological outcome of surgical correction of
ureteropelvic junction obstruction. J Urol 2001;165:1648e51.
18. Tsai JD, Huang FY, Lin CC, Tsai TC, Lee HC, Sheu JC, et al.
Intermittent hydronephrosis secondary to ureteropelvic junction obstruction: clinical and imaging features. Pediatrics
2006;117:139e46.
Pediatrics and Neonatology (2014) 55, 275e281
Available online at www.sciencedirect.com
ScienceDirect
journal homepage: http://www.pediatr-neonatol.com
ORIGINAL ARTICLE
Emergency Department Visits for Food
Allergy in Taiwan: A Retrospective Study
Chan-Fai Chan a,b, Po-Hon Chen b, Ching-Feng Huang c,d,
Tzee-Chung Wu b,e,*
a
Department of Pediatrics, National Yang-Ming University Hospital, Yilan, Taiwan
Division of Gastroenterology, Children’s Medical Center, Taipei Veterans General Hospital,
Taipei, Taiwan
c
Department of Pediatrics, Tri-service General Hospital, Taipei, Taiwan
d
National Defense University, Taiwan
e
School of Medicine, National Yang-Ming University, Taipei, Taiwan
b
Received Apr 15, 2013; received in revised form Nov 4, 2013; accepted Nov 26, 2013
Available online 17 March 2014
Key Words
anaphylaxis;
emergency
department;
epinephrine;
food allergens;
food allergy
Background: Little is known about the characteristics of patients who visit the emergency
department (ED) due to food allergy in Taiwan. This study aims to assess the triggers, clinical
presentations, and management of patients presenting to a tertiary ED for food allergy.
Methods: This is a retrospective study of 369 visits presenting to the ED of Taipei Veterans General Hospital, Taipei, Taiwan for food allergy over a 2 year period. Patients’ demographics,
food allergens, presenting features, and management were addressed and analyzed. Adult
and pediatric cases were also compared.
Results: The patients had an average age of 32.9 years [standard deviation (SD) 20.6]; the
cohort was 66.9% adult and 53.7% male. Seafood (67.5%), fish (6.2%), and fruits (4.3%) were
the major foods eliciting acute allergic reactions. Overall itchy mucocutaneous lesion was
the most common presentation (85.6%), followed by anaphylaxis (12.2%), respiratory distress
(1.4%), and anaphylactic shock (0.8%). Mucocutaneous involvement was more common in the
pediatric population (92.6% vs. 82.2%, p Z 0.007), whereas anaphylaxis was more prevalent
in adults (15.4% vs. 5.7%, p Z 0.0068). Antihistamines (98.6%) and systemic corticosteroids
(63.1%) were commonly used medications. Only 2.2% of patients with anaphylaxis received
epinephrine. The average duration in the ED was 1.6 hours (SD 1.8). No death was documented in the current study.
Conclusion: Seafood, fish, and fruits are common foods which cause acute allergic reactions in
Taiwan. Although most food allergies are mild, anaphylactic shock still presents in about 1% of
patients. Only a minority of patients with anaphylaxis receive epinephrine. As anaphylaxis may
* Corresponding author: Division of Gastroenterology, Children’s Medical Center, Taipei Veterans General Hospital, Number 201, Shih-Pai
Road, Section 2, Taipei 11217, Taiwan.
E-mail address: [email protected] (T.-C. Wu).
1875-9572/$36 Copyright ª 2014, Taiwan Pediatric Association. Published by Elsevier Taiwan LLC. All rights reserved.
http://dx.doi.org/10.1016/j.pedneo.2013.11.006
276
C.-F. Chan et al
be life-threatening, prompt education and use of an epinephrine auto-injector deserves
further concern.
Copyright ª 2014, Taiwan Pediatric Association. Published by Elsevier Taiwan LLC. All rights
reserved.
1. Introduction
During the past decades the prevalence of food allergy has
increased dramatically worldwide, making it an important
public health issue.1e6 Although most food allergies cause
minor symptoms, a small proportion of patients still suffer
from potentially fatal reactions. In contrast to most published
reports from Western countries, which demonstrated peanut
and tree nuts to be the most common foods causing severe
reactions,7 limited data revealed seafood to be the main
cause of food-induced anaphylaxis in Asia.8 Patients with
moderate-to-severe food allergies are often sent to a nearby
emergency room. Better outcome depends on prompt
recognition and timely management of these life-threatening
reactions. Despite universal recommendations for the use of
epinephrine in anaphylaxis, underuse of epinephrine has been
widely reported.9 Currently, little is known about the characteristics of patients visiting the emergency department
(ED) for food allergy in Taiwan. In order to improve the quality
of health care, it is therefore necessary to investigate the
common food allergens and current practice in the ED. The
aim of this study is to assess the food triggers, clinical presentations, and management in patients presenting to a tertiary ED for food allergy in Taiwan.
2. Methods
2.1. Study design and population
This is a retrospective study over a 2 year period, from
November 2009 to November 2011. Medical records of patients presenting to the ED of Taipei Veterans General Hospital, Taiwan, Taiwan with a clinical diagnosis of acute
allergic reactions to food were reviewed. Charts were
extracted using the International Classification of Disease,
ninth revision (ICD-9) codes: 995.0 (other anaphylactic
shock), 995.1 (angioneurotic edema), 995.3 (allergy, unspecified), 995.60 (allergy due to unspecified food),
995.61e995.69 (allergy due to specified food), 708.0 (allergic
urticaria), and 708.9 (urticaria, unspecified). All medical
records were reviewed thoroughly by two pediatricians who
were experts in both gastroenterology and immunology.
Special attention was paid to patients’ previous food allergy
history to help identify the cases of food allergy. Adverse
reactions caused by food intolerance, food poisoning, food
additive, drug, and insect bite were excluded from the study.
Cases of possible food allergies where the food consumed
was not mentioned were also excluded. For the patients who
visited the ED more than once because of treatment failure
or relapse symptoms, only one visit was counted to avoid
overestimating the specific food allergen. Patients’ demographics, eliciting foods, presenting features, treatment,
and disposition, were recorded in detail. All edible aquatic
animals, except fish, were collectively referred to as seafood. Food allergens were classified as “mixed food” if more
than one identifiable food was ingested. Clinical symptoms
were sorted by four categories which included itchy mucocutaneous lesion (e.g., reddening, itching, urticaria, or
angioedema), respiratory distress (e.g., wheezing, tachypnea, or dyspnea), anaphylaxis, and anaphylactic shock. The
definition of anaphylaxis was involvement of two or more
organ systems from the following: mucocutaneous, respiratory, cardiovascular, and gastrointestinal, as adopted from
the Second National Institute of Allergy and Infectious Disease/Food Allergy and Anaphylaxis Network Symposium.10
Anaphylactic shock was diagnosed when patients presented
with signs of poor perfusion, such as hypotension plus
delayed capillary refill, sudden collapse, syncope, or change
of mental status, and requiring prompt resuscitation. The
severity was further classified as mild (itchy mucocutaneous
lesion), moderate (respiratory distress or anaphylaxis), and
severe (anaphylactic shock). Eliciting foods, clinical features, and management of adult and pediatric cases were
compared. This study was approved by the Institutional Review Board of Taipei Veterans General Hospital.
2.2. Statistical analysis
The Chi-square test was used to compare between adults
and children, and to investigate their independence from
each other. A two-tailed p < 0.05 was considered a statistically significant difference.
3. Results
Over the 2 year study period, 369 visits to the ED (363
persons) were identified as having acute allergic reactions
to food. There were 247 adults (66.9%) and 122 children
aged < 19 years (33.1%). The median age of the patients
was 32.9 years [standard deviation (SD) 20.6]. Eighty-five
percent of the food allergies happened on the day when the
food was ingested. Most patients suffered from the allergic
reactions after lunch or dinner. Twenty-three percent of
the cases had experienced a similar food allergy due to the
same identified food in the past. Twenty-two percent of the
participants performed Multiple Allergen Simultaneous Test
(MAST) or ImmunoCAP at outpatient department follow up;
however, only 5% got positive results. The cohort’s information is shown in Table 1. Overall, seafood (67.5%), fish
(6.2%), and fruits (4.3%) were the major food allergens
encountered in our patients, accounting for nearly 80% of
all eliciting foods. The three most common food allergens in
the pediatric population were seafood (66.4%), fish (4.9%),
and fruits (3.3%), followed by egg, nuts, milk, wheat, vegetables, and meat. The three most common food triggers in
adults were seafood (68%), fish (6.9%), and fruits (4.9%),
Emergency department visits for food allergy in Taiwan
Table 1
277
Demographic characteristics, medical history, clinical presentations, and management of patients.
Visits
Demographics
Age (y), mean SD
Male (%)
Medical history
Previous food allergy (%)
Drug allergy (%)
Clinical presentations
Mucocutaneous lesion
Reddening
Itching
Urticaria
Urticaria þ angioedema
Respiratory distress
Dyspnea
Anaphylaxis
Mucocutaneous þ respiratory compromise
Mucocutaneous þ hypotension
Mucocutaneous þ persistent abdominal pain
Anaphylactic shock
Management
Antihistamine (oral/injection/both)
Antihistamine þ hydrocortisone injection
Epinephrine injection
Oral prednisolone
Overall
Adult
Children
p
369
247
122
32.9 20.6
53.7
42.8 16.9
50.2
11 5.4
60.9
23.0
10
24.3
12.1
20.5
6.6
0.43
0.11
316
1
5
305
5
5
5
45
28
8
9
3
203
1
3
195
4
4
4
38
21
8
9
2
113
0
2
110
1
1
1
7
7
0
0
1
0.007*
133 (43/21/69)
231
3
2
48 (5/1/42)
196
2
1
85 (38/20/27)
35
1
1
0.0001*
0.0001*
1
0.55
1
0.0068*
1
* p < 0.05.
followed by milk, meat, vegetables, egg, and nuts. Seventeen percent of adults and 14.8% of children ingested two
or more suspected food allergens prior to when the allergic
reactions happened.
Overall, itchy mucocutaneous lesion was the most
common presentation (85.6%), followed by anaphylaxis
(12.2%), respiratory distress (1.4%), and anaphylactic shock
(0.8%). Mucocutaneous involvement was more common in
the pediatric population (92.6% vs. 82.2%, p Z 0.007). Of
the 45 patients suffering from anaphylaxis, 38 out of 247
(15.4%) were adults and seven of out 122 (5.7%) were
children. Anaphylaxis was more prevalent in adults
compared with children (15.4% vs. 5.7%, p Z 0.0068;
Figure 1). The frequency of other clinical symptoms was not
significantly different between the two groups.
Moderate-to-severe reactions including respiratory
distress, anaphylaxis, and anaphylactic shock comprised
14.3% of all food allergies. Sixty-two percent of patients
with anaphylaxis presented with mucocutaneous lesion plus
respiratory compromise. Foods causing these serious reactions in children included unspecified seafood and fish.
Seafood (shrimp, crab), fish, and fruits (pineapple, kiwi,
and persimmon) were documented to be responsible for
these reactions in adults (Table 2). Although peanuts and
nuts are important food allergens which cause anaphylaxis
in Western countries, they only caused minor reactions in
our study and all cases of anaphylactic shock were elicited
by seafood.
Regarding treatment, most patients received antihistamines (98.6%), followed by systemic corticosteroids
(63.1%), bronchodilator inhalation (2.2%), and epinephrine injections (0.8%). No patient was given epinephrine
inhalation. Single usage of antihistamines was prescribed
to 85 pediatric patients and 48 adults, respectively,
(69.7% vs. 19.4%, p < 0.0001). Up to 196 adults received
antihistamines plus hydrocortisone injections, whereas
only 35 children received such management (79.4% vs.
28.7%, p <0.0001). All patients with anaphylactic shock
received epinephrine injections, whereas only one (2.2%)
out of the 45 patients with anaphylaxis received
epinephrine injection. Among the 363 patients, five were
admitted after initial management (1 persistent urticaria,
1 respiratory distress, 2 anaphylaxis, and 1 anaphylactic
shock).
The overall average duration in the ED was 1.6 hours
(SD 1.8). Patients who suffered from anaphylaxis were
observed in the ED for an average of 2.9 hours prior to
discharge (SD 2.8). No death was documented in our
patients. Of the 363 persons, five revisited the ED for
recurrent attack by the same food.
4. Discussion
To our knowledge this is the first study which was conducted in the ED to evaluate acute allergic reactions to
food in Taiwan. Our study not only describes the clinical
features of the patients, but also lists the possible food
triggers in detail, especially for patients with moderate-tosevere allergic reactions. Many reports state that food is
278
C.-F. Chan et al
Figure 1
Distribution of clinical presentations in adults and children.
the leading cause of anaphylaxis and common food allergens differ between areas. In 2007, Oren et al11 performed
a review of medical records to evaluate patients who presented to the ED with food-induced anaphylaxis in US. Their
study showed that among the 12 cases treated with
epinephrine, nine were caused by peanut and tree nuts. A
retrospective study conducted in New York, USA showed
that peanut was the most common food allergen to cause
anaphylaxis in hospitalized patients < 20 years old.12 Uguz
et al13 performed a questionnaire survey to assess allergic
reactions to food in the UK in 2005. According to their results, nearly half of the severe food allergies were elicited
by peanut and tree nuts. Data from Australia, Sweden, and
Germany revealed similar results.14e16
However, unlike those from most Western countries, our
study showed that seafood and fish were mainly responsible
for severe food allergies in Taiwan. Despite nuts being
important food allergens to cause anaphylaxis in Western
Table 2
countries, they caused minor reactions in our patients.
These results were consistent with those of other Asian
reports. One report from Hong Kong showed that nearly half
of the cases of anaphylaxis presenting to the ED were elicited by food, 70% of which was seafood.17 Two studies
from Thailand also found that seafood was the main cause
of anaphylaxis.18,19 Wu et al8 published a nationwide
questionnaire survey of food allergy in Taiwan in 2012.
According to their results, seafood, especially shrimp and
crab, was responsible for > 60% of all moderate-to-severe
reactions.
Some suppose that different cultural backgrounds, dietary habits, and cooking methods account for the regional
differences.20,21 As observed in two recent studies,
maternal consumption of peanut during pregnancy seems to
be associated with peanut sensitization.22,23 Although
peanut allergy is not common as compared with Western
countries, most Taiwanese women avoid peanuts during
Types of food allergens and their correlation with clinical severity.
Foods
Seafood
Fish
Fruit
Milk
Meat
Vegetable
Egg
Nut
Wheat
Mixed food
Total
Adult
Children
Overall, n (%)
Mild
Moderate/severe
Overall, n (%)
Mild
Moderate/severe
168
17
12
2
2
2
1
1
0
42
247
142
12
9
2
2
2
1
1
0
32
203
26
5
3
0
0
0
0
0
0
10
44
81
6
4
2
1
2
3
3
2
18
122
74
4
4
2
1
2
3
3
2
18
113
7
2
0
0
0
0
0
0
0
0
9
(68)
(6.9)
(4.9)
(0.8)
(0.8)
(0.8)
(0.4)
(0.4)
(0.4)
(17)
(100)
(66.4)
(4.9)
(3.3)
(1.6)
(0.8)
(1.6)
(2.5)
(2.5)
(1.6)
(14.8)
(100)
Emergency department visits for food allergy in Taiwan
pregnancy, wishing to protect their babies from peanut
allergy. Peanuts are also typically dry-roasted in Western
countries, whereas in Asia they are usually fried or boiled,
which reduces the allergenic property of peanut proteins.24,25 Traditionally, most Taiwanese women drank lots
of “peanut hoof soup” during lactation, as it is generally
believed that this enhances breast milk production. We
speculate that the avoidance of peanuts during pregnancy
and early exposure to peanuts with reduced allergenicity
through breast milk in young infants induce oral tolerance,
but this remains to be determined.
Nearly 90% of our patients presented with mucocutaneous
manifestation, which was the most common symptom in the
present study. Isolated gastrointestinal involvement was not
found in the study, which might be due to a lack of awareness
from the ED physicians. We believe parts of the pediatric
population present food allergy as gastrointestinal bleeding,
namely milk protein-induced proctitis, which is not seen in
adults. Despite most food allergies causing only minor reactions, 12% of our patients still suffered from anaphylaxis,
mainly mucocutaneous involvement plus dyspnea. The incidence was lower than other reports.26,27 We believed that
the true incidence was underestimated due to the underdiagnosis of anaphylaxis. Also, possible food allergies without
record of food allergens were excluded. Our study showed
that anaphylaxis was more prevalent in adults than in children, which may be due to repeated exposure to the causative food allergens in adults.
Current guidelines suggest prompt use of epinephrine in
anaphylaxis and that delayed use leads to poor
prognosis.28e30 Despite the universal recommendations, use
of epinephrine in anaphylaxis is not common. For example,
an 11 year retrospective review performed in Italy showed
that only 15% of patients with anaphylaxis received
epinephrine.31 Clark et al26 conducted a multicenter study of
food allergy with 678 ED patients. Half of the participants
were identified as having severe allergic reactions, but only
24% received epinephrine. An online questionnaire performed by Jacobs et al27 in the US showed that only 34% of
suspected cases of food-induced anaphylaxis were given
epinephrine. Another report by Russell et al32 revealed a
higher percentage of epinephrine use in which half of the
anaphylaxis patients received intramuscular epinephrine,
either in a pre-hospital setting or during ED visits. This trend
of epinephrine underuse was also illustrated in our study. Of
the 45 patients who fulfilled the criteria of anaphylaxis, only
2.2% received epinephrine. This underuse may be due to
physicians’ concern about the adverse effects of epinephrine. Also, there might be a lack of consensus among physicians and existing guidelines on how to appropriately manage
patients at risk for anaphylaxis.33 Most of our anaphylaxis
patients received antihistamines and corticosteroids. However, usage of these drugs, either alone or in combination,
cannot be a substitute for epinephrine’s life-saving role.34
A convenient device used to treat anaphylaxis is the
preload epinephrine auto-injector, namely the Epipen
(Mylan Inc.).35 Currently, there are two fixed doses available, 0.15 mg and 0.30 mg. The 0.15 mg device is recommended for children whose weight is 15e25 kg, and the
0.30 mg for those > 25 kg, as well as adults.36,37 The
optimal way to deliver epinephrine to children whose
weight is < 15 kg has not yet been established. Prescribing
279
the exact dosage by drawing from an ampule, or empirically
use the 0.15 mg device, are the two options for parents.
Intramuscular injection into the lateral tight (vastus lateralis muscle) is the preferred route in first-aid treatment.
For up to 20% of patients with anaphylaxis, two doses of
epinephrine are needed.38e40 Educating both patients at
risk of anaphylaxis and their parents on how and when to
use the epinephrine auto-injector is important.
Our patients stayed in the ED for an average of 1.6 hours
prior to discharge. Those who suffered from anaphylaxis
were observed for an average of 2.9 hours prior to being
sent home. Because of the risk of biphasic reactions in
anaphylaxis patients, prolonged observation of up to 12e24
hours should be considered.17,41
Twenty-two percent of the participants performed MAST/
CAP tests at OPD follow up; however, only 5% got positive
results. The low positive rate may be due to lost follow up of
the patients and longtime interval between the food allergy
and laboratory test. Another possible reason is that some of
the food allergens may not be included in the MAST/CAP tests
used. According to the 2013 update of the world allergy organization anaphylaxis guidelines, the Ara h components in
peanuts (especially Ara h 2), prebiotics galactooligosaccharides, orally ingested mites, gelatin, and Anisakis simplex were identified as new anaphylaxis triggers.42
More studies are needed to further determine the necessity
of including these new food allergens in current allergen
tests to help identify patients at risk of food allergy.
One possible limitation of our study is that this is a
retrospective chart review which depends on accurate and
complete documentation of the medical records. To overcome this, all charts were reviewed thoroughly by two
pediatricians with expertise in both gastroenterology and
immunology. Another limitation is the use of ICD-9 codes as
the method to trace patients, as symptoms and signs may
not be correctly coded. In order to include all food allergies, we reviewed not only the code for 995.60 (allergy
due to unspecified food), 995.61e995.69 (allergy due to
specified food), but also 995.0 (other anaphylactic shock),
995.1 (angioneurotic edema), 995.3 (allergy, unspecified),
708.0 (allergic urticaria), and 708.9 (urticaria, unspecified).
In conclusion, seafood, fish, and fruits are major food
allergens causing acute allergic reactions in Taiwan.
Although most food allergies are mild, anaphylactic shock
still presents in about 1% of patients. Educating both patients at risk of these serious allergic reactions and their
parents about how and when to use an epinephrine autoinjector is important.
Conflicts of interest
All authors declare no conflicts of interest.
Acknowledgments
This research was supported by grant V102C-110 from Taipei
Veterans General Hospital, Taiwan. The authors would like to
thank Li-Shu Wang, Children’s Medical Center, Taipei Veterans General Hospital, for field investigation and data statistical analysis.
280
C.-F. Chan et al
References
1. Rona RJ, Keil T, Summers C, Gislason D, Zuidmeer L,
Sodergren E, et al. The prevalence of food allergy: a metaanalysis. J Allergy Clin Immunol 2007;120:638e46.
2. Keil T. Epidemiology of food allergy: what’s new? A critical
appraisal of recent population-based studies. Curr Opin Allergy Clin Immunol 2007;7:259e63.
3. Branum AM, Lukacs SL. Food allergy among children in the
United States. Pediatrics 2009;124:1549e55.
4. Gupta RS, Springston EE, Warrier MR, Smith B, Kumar R,
Pongracic J, et al. The prevalence, severity, and distribution of
childhood food allergy in the United States. Pediatrics 2011;
128:e9e17.
5. Soller L, Ben-Shoshan M, Harrington DW, Fragapane J,
Joseph L, St Pierre Y, et al. Overall prevalence of self-reported
food allergy in Canada. J Allergy Clin Immunol 2012;130:
986e8.
6. Allen KJ, Koplin JJ. The epidemiology of IgE-mediated food
allergy and anaphylaxis. Immunol Allergy Clin North Am 2012;
32:35e50.
7. Ja
¨rvinen KM. Food-induced anaphylaxis. Curr Opin Allergy Clin
Immunol 2011;11:255e61.
8. Wu TC, Tsai TC, Huang CF, Chang FY, Lin CC, Huang IF, et al.
Prevalence of food allergy in Taiwan: a questionnaire-based
survey. Intern Med J 2012;42:1310e5.
9. Simons FE. Pharmacologic treatment of anaphylaxis: can the
evidence base be strengthened? Curr Opin Allergy Clin Immunol 2010;10:384e93.
10. Sampson HA, Mun
˜oz-Furlong A, Campbell RL, Adkinson Jr NF,
Bock SA, Branum A, et al. Second symposium on the definition
and management of anaphylaxis: summary report e Second
National Institute of Allergy and Infectious Disease/Food Allergy and Anaphylaxis Network symposium. J Allergy Clin
Immunol 2006;117:391e7.
11. Oren E, Banerji A, Clark S, Camargo Jr CA. Food-induced
anaphylaxis and repeated epinephrine treatments. Ann Allergy
Asthma Immunol 2007;99:429e32.
12. Lin RY, Anderson AS, Shah SN, Nurruzzaman F. Increasing
anaphylaxis hospitalizations in the first 2 decades of life: New
York state, 1990e2006. Ann Allergy Asthma Immunol 2008;
101:387e93.
13. Uguz A, Lack G, Pumphrey R, Ewan P, Warner J, Dick J, et al.
Allergic reactions in the community: a questionnaire survey of
members of the anaphylaxis campaign. Clin Exp Allergy 2005;
35:746e50.
14. de Silva IL, Mehr SS, Tey D, Tang ML. Paediatric anaphylaxis: a 5
year retrospective review. Allergy 2008;63:1071e6.
15. Vetander M, Helander D, Flodstro
¨m C, Ostblom E, Alfve
´n T,
Ly DH, et al. Anaphylaxis and reactions to foods in children e a
population-based case study of emergency department visits.
Clin Exp Allergy 2012;42:568e77.
16. Beyer K, Eckermann O, Hompes S, Grabenhenrich L, Worm M.
Anaphylaxis in an emergency setting - elicitors, therapy and
incidence of severe allergic reactions. Allergy 2012;67:
1451e6.
17. Smit DV, Cameron PA, Rainer TH. Anaphylaxis presentations to
an emergency department in Hong Kong: incidence and predictors of biphasic reactions. J Emerg Med 2005;28:381e8.
18. Piromrat K, Chinratanapisit S, Trathong S. Anaphylaxis in an
emergency department: a 2-year study in a tertiary-care hospital. Asian Pac J Allergy Immunol 2008;26:121e8.
19. Techapornroong M, Akrawinthawong K, Cheungpasitporn W,
Ruxrungtham K. Anaphylaxis: a ten years inpatient retrospective study. Asian Pac J Allergy Immunol 2010;28:262e9.
20. Woods RK, Abramson M, Bailey M, Walters EH. International
prevalences of reported food allergies and intolerances.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
39.
Comparisons arising from the European Community Respiratory
Health Survey (ECRHS) 1991e1994. Eur J Clin Nutr 2001;55:
298e304.
Shek LP, Lee BW. Food allergy in Asia. Curr Opin Allergy Clin
Immunol 2006;6:197e201.
DesRoches A, Infante-Rivard C, Paradis L, Paradis J, Haddad E.
Peanut allergy: is maternal transmission of antigens during
pregnancy and breastfeeding a risk factor? J Investig Allergol
Clin Immunol 2010;20:289e94.
Sicherer SH, Wood RA, Stablein D, Lindblad R, Burks AW,
Liu AH, et al. Maternal consumption of peanut during pregnancy is associated with peanut sensitization in atopic infants.
J Allergy Clin Immunol 2010;126:1191e7.
Beyer K, Morrow E, Li XM, Bardina L, Bannon GA, Burks AW,
et al. Effects of cooking methods on peanut allergenicity. J
Allergy Clin Immunol 2001;107:1077e81.
Maleki SJ, Viquez O, Jacks T, Dodo H, Champagne ET, Chung SY,
et al. The major peanut allergen, Ara h 2, functions as a trypsin
inhibitor, and roasting enhances this function. J Allergy Clin
Immunol 2003;112:190e5.
Clark S, Bock SA, Gaeta TJ, Brenner BE, Cydulka RK,
Camargo CA, et al. Multicenter study of emergency department visits for food allergies. J Allergy Clin Immunol 2004;113:
347e52.
Jacobs TS, Greenhawt MJ, Hauswirth D, Mitchell L, Green TD. A
survey study of index food-related allergic reactions and
anaphylaxis management. Pediatr Allergy Immunol 2012;23:
582e9.
Simons FE, Ardusso LR, Bilo
` MB, EI-Gamal YM, Ledford DK, Ring J,
et al. World Allergy Organization anaphylaxis guidelines: summary. J Allergy Clin Immunol 2011;127:587e93. e1e22.
Bock SA, Mun
˜oz-Furlong A, Sampson HA. Fatalities due to
anaphylactic reactions to foods. J Allergy Clin Immunol 2001;
107:191e3.
Bock SA, Mun
˜oz-Furlong A, Sampson HA. Further fatalities
caused by anaphylactic reactions to food, 2001-2006. J Allergy
Clin Immunol 2007;119:1016e8.
Cianferoni A, Novembre E, Mugnaini L, Lombardi E,
Bernardini R, Pucci N, et al. Clinical features of acute
anaphylaxis in patients admitted to a university hospital: an
11-year retrospective review (1985e1996). Ann Allergy Asthma
Immunol 2001;87:27e32.
Russell S, Monroe K, Losek JD. Anaphylaxis management in the
pediatric emergency department: opportunities for improvement. Pediatr Emerg Care 2010;26:71e6.
Kastner M, Harada L, Waserman S. Gaps in anaphylaxis management at the level of physicians, patients, and the community: a systematic review of the literature. Allergy 2010;65:
435e44.
Keet C. Recognition and management of food-induced
anaphylaxis. Pediatr Clin North Am 2011;58:377e88.
Simons FE, World Allergy Organization. Epinephrine autoinjectors: first-aid treatment still out of reach for many at
risk of anaphylaxis in the community. Ann Allergy Asthma
Immunol 2009;102:403e9.
Sicherer SH, Simons FE, Section on Allergy and Immunology,
American Academy of Pediatrics. Self-injectable epinephrine
for first-aid management of anaphylaxis. Pediatrics 2007;119:
638e46.
Simons FE. First-aid treatment of anaphylaxis to food: focus on
epinephrine. J Allergy Clin Immunol 2004;113:837e44.
Rudders SA, Banerji A, Corel B, Clark S, Camargo Jr CA.
Multicenter study of repeat epinephrine treatments for foodrelated anaphylaxis. Pediatrics 2010;125:e711e8.
Manivannan V, Campbell RL, Bellolio MF, Stead LG, Li JT,
Decker WW. Factors associated with repeated use of
epinephrine for the treatment of anaphylaxis. Ann Allergy
Asthma Immunol 2009;103:395e400.
Emergency department visits for food allergy in Taiwan
40. Ja
¨rvinen KM, Sicherer SH, Sampson HA, Nowak-Wegrzyn A. Use
of multiple doses of epinephrine in food-induced anaphylaxis in
children. J Allergy Clin Immunol 2008;122:133e8.
41. Tole JW, Lieberman P. Biphasic anaphylaxis: review of incidence, clinical predictors, and observation recommendations.
Immunol Allergy Clin North Am 2007;27:309e26.
281
42. Simons FE, Ardusso LR, Dimov V, Ebisawa M, EI-Gamal YM,
Lockey RF, et al. World allergy organization anaphylaxis
guidelines: 2013 update of the evidence base. Int Arch Allergy
Immunol 2013;162:193e204.
Pediatrics and Neonatology (2014) 55, 282e290
Available online at www.sciencedirect.com
ScienceDirect
journal homepage: http://www.pediatr-neonatol.com
ORIGINAL ARTICLE
Changes in Amplitude-integrated
Electroencephalograms in Piglets During
Selective Mild Head Cooling After
Hypoxia-ischemia
Ji-Mei Wang a,y, Guo-Fu Zhang b,y, Wen-Hao Zhou c,
Ze-Dong Jiang c,d, Xiao-Mei Shao c,*
a
Department of Neonatology, Gynecology, and Obstetrics, Hospital of Fudan University,
Shanghai 200011, China
b
Department of Radiology, Gynecology, and Obstetrics, Hospital of Fudan University, Shanghai 200011,
China
c
Department of Neonatology, Children’s Hospital of Fudan University, Shanghai 201102, China
d
Neonatal Unit, Department of Paediatrics, University of Oxford, John Radcliffe Hospital, Oxford,
United Kingdom
Received Jun 3, 2013; received in revised form Sep 26, 2013; accepted Sep 30, 2013
Available online 17 January 2014
Key Words
aEEG;
brain damage;
hypoxia-ischemia;
neonatology;
neuroprotection;
piglets
Background: Amplitude-integrated electroencephalogram (aEEG) is a simplified, alternative
means of monitoring cerebral function and may be more useful clinically in some situations
than conventional EEG. The aim of this study is to evaluate newborn piglets as an animal model
to examine the effect of selective mild head cooling (HC) on aEEG after hypoxia-ischemia (HI).
Methods: Thirty-four piglets were randomly allocated to the following treatment groups:
normothermic control group (NC, n Z 7), selective HC control group (HC, n Z 9), normothermic HI group (NHI, n Z 9), and selective HC HI group (SHC-HI, n Z 9). HI was induced
by temporary occlusion of both carotid arteries and simultaneous reduction of the concentration of inspired oxygen to 6% for 30 minutes. Mild hypothermia (35 C) was induced after HI using a HC cap and was maintained for 24 hours. Changes in aEEG were monitored for 6 days after
these treatments and the incidence of abnormalities analyzed. Physiological parameters were
also measured during this period.
Results: In the two HI groups, animals exhibited severely abnormal aEEGs [continuous low
voltage (CLV), burst-suppression, or flat tracing (FT)] 20 minutes after the beginning of HI.
* Corresponding author. Department of Neonatology, Children’s Hospital of Fudan University, 399 Wanyuan Road, Shanghai 201102, China.
E-mail address: [email protected] (X.-M. Shao).
y
As first authors who contributed equally to this work.
1875-9572/$36 Copyright ª 2013, Taiwan Pediatric Association. Published by Elsevier Taiwan LLC. All rights reserved.
http://dx.doi.org/10.1016/j.pedneo.2013.09.012
aEEG during HC after HI
283
At 2 hours, the aEEG returned to normal in most of these animals. From 12 hours to 6 days, all
animals in the NHI group exhibited severely abnormal aEEGs. Fewer animals in the SHC-HI
group exhibited severe abnormal aEEGs during this time period, and four out of nine (44.4%)
animals had continuous normal voltage (CNV) at 6 days.
Conclusions: Selective mild HC decreases the incidence of severe abnormal aEEGs at late times
after HI in newborn piglets.
Copyright ª 2013, Taiwan Pediatric Association. Published by Elsevier Taiwan LLC. All rights
reserved.
1. Introduction
Perinatal hypoxic-ischemic brain damage (HIBD) is a major
cause of perinatal mortality and long-term neurodisability.1
Evidence from both human and animal studies has shown
that hypothermia provides neuroprotection from hypoxiaischemia (HI) in adults and young/newborns,2,3 and it can
reduce the incidence of cerebral palsy.4e6 Clearly, hypothermia is a promising treatment for perinatal HIBD.
Monitoring changes in electrocortical brain activity during
HIBD can help assess the degree of brain damage and predict
neurological outcome. For many years, conventional electroencephalography (EEG) has been used for intermittent
and continuous assessment of brain function and for the
prediction of neurological outcomes in infants. However,
conventional EEG is of limited applicability in the neonatal
intensive care unit. The use of multichannel EEG to monitor
cerebral function in newborn infants who are in a critical
condition is impractical.7,8 Amplitude-integrated EEG (aEEG)
is a simplified, alternative means of monitoring cerebral
function.9 Advantages of aEEG are that it is easy to use and
analyze, shows less interference from artefacts such as
muscle contraction, and positions the electrodes over the
parietal zone, above an area known to be sensitive to
ischemia. This recording approach has been shown to be one
of the most accurate bedside methods for predicting neurological outcomes in term infants after HI.10,11 Indeed, this
non-invasive technique has been increasingly used to identify
infants suitable for hypothermic neuroprotection following
severe HI.12
To date, there is little information available regarding
the use of aEEG to monitor changes in brain electrical activity in newborn animals after induction of HI and hypothermia. Therefore, we carried out an experiment in
newborn piglets to examine aEEG and corresponding physiological data before and up to 6 days after the induction of HI
and subsequent 24-hour mild hypothermia. We hypothesized
that aEEG would be abnormal after the induction of HI and
that mild hypothermia would promote normalization.
2. Materials and methods
2.1. Animals
Thirty-four healthy newborn white piglets of either sex
were obtained from an experimental animal nursery in
Shanghai. They were aged between 5 days and 7 days (term
delivery) and weighed from 2.10 kg to 2.71 kg (mean:
2.35 0.18 kg).
Piglets were randomly allocated to one of the following
four groups: (A) normal temperature with rectal temperature maintained at approximately 39 C [normothermic
control group (NC), n Z 7]; (B) selective head cooling (HC)
without HI, with nasopharyngeal temperature maintained
at 35 C and rectal temperature maintained at 36 C [HC
without HI group (HC), n Z 7]; (C) HI insult without selective HC and rectal temperature maintained at 39 C
[normothermic HI group (NHI), n Z 9]; and (D) HC (mild
hypothermia) after HI insult, with nasopharyngeal temperature maintained at 35 C and rectal temperature
maintained at 36 C [selective HC after HI group (SHC-HI),
n Z 9].
This study was approved by the Animal Ethics Committee of
the Children’s Hospital of Fudan University, Shanghai, China.
2.2. Animal surgical preparation and postsurgical
treatment
In all groups, anesthesia was initiated with intramuscular
ketamine (10 mg/kg) and maintained with an infusion of
ketamine (10 mg/kg/hour). Maintenance fluid (Na 50 mEq/L,
K 20 mEq/L, Cl 50 mEq/L, 5% glucose) was infused at a rate of
5 mL/kg/hour. An ear vein was cannulated for continuous
infusion of maintenance fluids and administration of medication. Endotracheal intubation and mechanical ventilation
(Newport 200, Newport Medical Instruments Inc., Costa
Mesa, CA, USA) were then initiated. A Judkins number 4
catheter (Utal Medical Products, Inc USA) was inserted into
the left axillary artery to monitor blood pressure and for
repeated blood sampling. This artery is the continuation of
the left subclavian artery in the region of the shoulder joint
and left front leg. Bilateral carotid arteries were isolated for
cerebral ischemic insult. Ketamine administration was
terminated at the end of this surgery. Ventilation was
adjusted as necessary to maintain synchronized intermittent
mandatory ventilation: fraction of inspired oxygen, 30%;
peak inspiratory pressure, 12e15 cmH2O; respiratory rate,
20e30 breaths/minute; inspiratory time, 0.5 seconds; positive end-expiratory pressure, 3 cm H2O; flow, 8 L/minute;
and PaCO2, 40e50 mmHg. Piglets were then placed on a
radiant warmer bed in a prone position for 2 hours to stabilize
vital signs.
During the 6-day postsurgical period, animals were
maintained with milk containing chloral hydrate for mild
sedation (1 mL/kg of milk containing 10% chloral hydrate)
administered by gastric tube. Chloral hydrate at this dose
has no effect on aEEG. The protocol called for vasopressors
to
treat
hypotension
[mean
arterial
pressure
(MAP) < 40 mm Hg] and sodium bicarbonate to treat
284
acidosis, but in the actual experiment, these adjustments
were not needed for any animal. Animals were on mechanical ventilation with room air. No lung injury was
observed; that is, animals had no difficulty breathing, and
the results of the blood gas analyzer showed no CO2
retention. On Day 7, they were extubated and sacrificed
with an overdose of sodium pentobarbital.
2.3. Induction of acute cerebral HI
Cerebral HI was induced using a previously described
technique.13 After surgical preparation and the stabilization period on the warmer bed, HI was induced by simultaneous temporary occlusion of the bilateral carotid
arteries and abrupt reduction of the concentration of
inspired oxygen to 6% for 30 minutes.14
During HI, piglets received continuous physiological
monitoring of MAP, heart rate (HR), arterial pH and base
excess (BE), blood glucose, and electrolytes (Kþ, Naþ, and
Ca2þ). Physiological measurements were also taken during
the subsequent 2 hours to 6-day time period. At the end of
HI, the concentration of inspired oxygen was increased to
30%, and the occlusion of the bilateral carotid arteries was
terminated so that reperfusion could occur.
2.4. Selective HC and temperature monitoring
Prior to HC, rectal temperature was maintained at 39 C in
the two HC groups, as in the normal and HI groups. In animals
subjected to HI and HC, HC began 2 hours after the hypoxicischemic period. In animals given HC alone, it was given
immediately after the 2-hour postsurgical equilibration
period. For HC in both groups, the head of the piglet was
wrapped in a specially made thermostatically-controlled
cooling cap with a network of circulating water channels
(YJW608-04B Cooling Care System, Hengyang Radio Factory,
Hengyang, Hunan Province, China). Circulating cooled water
was automatically adjusted between 5 C and 24 C, as
required after cooling started, until the nasopharyngeal
temperature, a temperature that was shown to correlate
well with cerebral temperature in a preliminary study,
decreased to 35 0.2 C. This temperature was maintained
for 24 hours. Rectal temperature was continuously recorded
using a temperature probe (Datex-Engstrom CS/3 monitor)
(Hengyang Radio Factory) that was inserted 5 cm into the
rectum. The rectal temperature was maintained at
36.0 0.2 C using a servo-controlled radiant warmer bed.
Twenty-four hours after the start of selective HC, the cap
was removed and the animal was allowed to warm in room air
without application of additional heating. For piglets in the
two normothermic groups (NC and NHI), rectal temperature
was maintained at 39 C throughout the entire study period.
2.5. aEEG recording and analysis
A cerebral function monitor (CFM 5022, Lectromed Devices
Ltd, Letchworth Garden City, UK) was used for recording
aEEG. Recordings were made at time 0 (during the 2 hour
postsurgical equilibration period), 20 minutes (the middle
of the HI period for those that had this treatment), at 2
hours (the period between HI and HC in those that had both
J.-M. Wang et al
treatments), and 12 hours, 24 hours, 48 hours, 72 hours,
and 6 days after HI. Recordings at each time point were at
least 2 hours in duration. Because the hypothermic head
cap and the aEEG could not be used simultaneously, no
recording was made at the 12 hour time point in the two
selective hypothermia (HC and SHC-HI) groups. According
to a previously described method,15,16 aEEG traces were
assessed visually and classified into the following categories
(Figure 1): (A) continuous background activity with voltage
10e25 mV (continuous normal voltage, CNV); (B) discontinuous background activity with voltage predominantly
>10 mV (discontinuous normal voltage, DNV); (C) continuous background pattern of voltage below 10 mV (continuous low voltage, CLV); (D) discontinuous background
pattern with periods of very low cortical activity (<5 mV),
intermixed with bursts of higher amplitude (burst suppression, BS); and (E) mainly inactive (isoelectric tracing) of
extremely low voltage (<5 mV; flat tracing, FT). All aEEGs
were read by a single EEG specialist.
2.6. Measurements of physiological indices
Arterial blood samples were analyzed using a Nova analyzer
(Nova Biomedical, Waltham, MA, USA). HR and MAP were
monitored continuously using a critical care monitor
(Datex-Engstron CS/3).
2.7. Statistical analysis
Data are presented as mean and standard deviation and
were compared between groups at each time point by oneway analysis of variance. Post hoc comparisons were made
using Bonferroni corrections for pair-wise groups. The aEEG
results are presented as count (%) and were compared between groups by Fisher’s exact test. All statistical assessments were two-sided and evaluated at the 0.05 level of
significance. Statistical analyses were performed using SPSS
version 15.0 (SPSS Inc., Chicago, IL, USA).
3. Results
No deaths occurred during the HI insult or during the
following 6 days, and no convulsions occurred in any group.
One animal died at the beginning of the modeling process.
The baseline characteristics [sex, weight, HR, and MAP of
the 4 groups were comparable and no significant difference
in weight, sex, HR, or MAP was observed between the 4
groups (Table 1)].
3.1. Amplitude-integrated electroencephalograms
The aEEG results during the study period are summarized in
Table 2. All animals had normal aEEGs at 0 hours. For the
rest of the study period, as expected, all animals in the
normal control group exhibited CNV, as did all animals in
the hypothermic control group, except for five, which had
DNV at 24 hours, but not prior to or afterward. In the two
groups with HI (NHI and SHC-HI), all animals exhibited severe abnormal aEEGs at 20 minutes. In the NHI group, all
nine animals were CNV prior to HI; 20 minutes after HI, four
aEEG during HC after HI
285
Figure 1 Amplitude-integrated electroencephalogram waveforms in newborn piglets: (A) continuous normal voltage; (B)
discontinuous normal voltage; (C) continuous low voltage; (D) burst suppression; and (E) flat tracing.
BS (44.4%), two CLV (22.2%), and three (33.3%) FT were
observed, and no CNV or DNV; 2 hours after HI, six (66.7%)
animals returned to CNV, and two DNV and one BS were
observed. Similar results were observed during 12 hourse6
days after HI, with no CNV or DNV, but only BS, CLV, and FT
observed. In the SHC-HI group, all nine animals were CNV
prior to HI; 20 minutes after HI, six BS (66.7%) and three FT
(3.33%) were observed, and no CNV, CLV, or DNV; 2 hours
after HI, seven animals returned to CNV (77.8%), and two
DNV (22.2%) were observed; 24 hours and 48 hours after HI,
the majority of events were CLV (66.7%), followed by DNV
(22.2% and 33.3%); 72 hours after HI, all animals had DNV;
and finally 6 days after HI, four CNV (44.4%) and five DNV
(55.6%) were observed (Table 2).
286
Table 1
J.-M. Wang et al
Baseline comparisons between groups.
Sex
Female
Male
Weight (kg)
Heart rate (beat/
min)
MAP (mmHg)
NC (n Z 7)
HC (n Z 9)
NHI (n Z 9)
SHC-HI (n Z 9)
p
3
4
2.31
200.43
4
5
2.33
209.00
5
4
2.35
205.56
5
4
2.40
205.89
> 0.99
(42.9)
(57.1)
(0.23)
(12.53)
67.14 (6.01)
(44.4)
(55.6)
(0.17)
(12.69)
71.22 (4.82)
(55.6)
(44.4)
(0.08)
(16.77)
69.66 (4.65)
(55.6)
(44.4)
(0.24)
(16.74)
0.765
0.730
68.22 (6.83)
0.498
Weight, heart rate, and MAP are presented as mean (standard deviation); sex is presented as n (%).
HC Z head cooling; MAP Z mean arterial pressure; NC Z normothermic control; NHI Z normothermic hypoxia-ischemia;
SHC-HI Z selective head cooling hypoxia-ischemia.
3.2. pH, BE, and glucose concentrations
3.3. HR and MAP
The pH, BE, and glucose concentrations during the study
period are summarized in Figure 2. At 20 minutes (that is,
during the HI period) pH and BE were significantly lower and
glucose concentrations significantly higher in the two HI
groups compared with the two groups not subjected to this
procedure (all comparisons p < 0.05). pH levels were also
significantly lower in the SHC-HI group compared with the
NHI group (p < 0.05). At 20 minutes, pH levels were 7.23,
7.08, 7.46, and 7.47 in the NHI, SHC-HI, NC, and HC
groups, respectively. Corresponding BE concentrations
were 14.14 mmol/L, e14.68 mmol/L, 2.20 mmol/L, and
2.19 mmol/L, respectively. Corresponding glucose concentrations were 18.71 mmol/L, 16.19 mmol/L, 7.16 mmol/L, and
7.38 mmol/L, respectively. From 2 hours onwards, there were
no intergroup differences in pH or BE. From 2 hours to 36 hours,
there were no intergroup differences in glucose concentrations; however, from 48 hours onwards, glucose concentrations were significantly lower in the HC, NHI, and SHC-HI
groups compared with the NC group (all comparisons p < 0.05).
HR and MAP during the study period are summarized in
Figure 3. There were no significant intergroup differences
in HR at baseline 20 minutes or from 36 hours onwards. HR
was significantly lower in the HC group compared with the
NC and NHI groups at 2 hours, 12 hours, and 24 hours, and
compared with the SHC-HI group at 12 hours and 24 hours
(all comparisons p < 0.05). HR was also significantly lower
in the SHC-HI group compared with the NC and HC groups at
2 hours (both comparisons p < 0.05). There were no significant between group differences in MAP throughout the
study; MAP ranged from 67.0 mmHg to 72.4 mmHg.
Table 2
3.4. Electrolyte concentrations
The electrolyte (Naþ, Kþ, and Ca2þ) concentrations during
the study period are summarized in Figure 4. All electrolyte
concentrations remained stable in each of the four treatment groups throughout the study. There were no significant
Effects of mild hypothermia on amplitude-integrated electroencephalograms after hypoxia-ischemia.*,y
Time
NC (n Z 7)
HC (n Z 9)
NHI (n Z 9)
SHC-HI (n Z 9)
p
CNV
CNV
DNV
CNV
DNV
BS
CLV
FT
CNV
DNV
BS
CLV
FT
0h
20 min
2h
12 h
24 h
48 h
72 h
6d
7
9
0
9
0
0
0
0
9
0
0
0
0
NA
7
9
0
0
0
4
2
3
0
0
6
0
3
<0.0001z
7
9
0
6
2
1
0
0
7
2
0
0
0
0.2911
7
0
0
0
0
4
2
3
0
0
0
0
0
0.0002z
7
4
5
0
0
4
1
4
0
2
1
6
0
<0.0001z
7
9
0
0
0
5
1
3
0
3
0
6
0
<0.0001z
7
9
0
0
0
4
2
3
0
9
0
0
0
<0.0001z
7
9
0
0
0
4
3
2
4
5
0
0
0
<0.0001z
* Treatment groups: HC Z selective head cooling without hypoxia-ischemia þ nasopharyngeal temperature maintained at 35 C and rectal
temperature maintained at 36 C; NC Z normal temperature with rectal temperature maintained at approximately 39 C; NHI Z normal
temperature after hypoxic-ischemic insult þ rectal temperature maintained at 39 C; and SHC-HI Z head cooling (mild hypothermia) after
hypoxic-ischemic insult þ nasopharyngeal temperature maintained at 35 C and rectal temperature maintained at 36 C.
y
Amplitude-integrated electroencephalogram categories: BS Z burst suppression; CLV Z continuous low voltage; CNV Z continuous
normal voltage; DNV Z discontinuous normal voltage; FT Z flat tracing.
z
Indicates a significant difference among the groups (p < 0.05).
aEEG during HC after HI
287
Figure 2 (A) pH; (B) base excess (BE); and (C) blood glucose concentrations before and during the study period. Treatment
groups: HC Z selective head cooling without hypoxia-ischemia þ nasopharyngeal temperature maintained at 35 C and rectal
temperature maintained at 36 C; NC Z normal temperature with rectal temperature maintained at approximately 39 C;
NHI Z normal temperature after hypoxic-ischemic insult þ rectal temperature maintained at 39 C; SHC-HI Z head cooling (mild
hypothermia) after hypoxic-ischemic insult þ nasopharyngeal temperature maintained at 35 C and rectal temperature maintained
at 36 C. * Indicates a significant difference compared with the NC group (p < 0.05). y Indicates a significant difference compared
with the HC group (p < 0.05). z Indicates a significant difference compared with the NHI group (p < 0.05).
between group differences at any time point. Mean Naþ, Kþ,
and Ca2þ concentrations ranged from 136.22 mmol/L to
138.24 mmol/L, from 4.22 mmol/L to 4.62 mmol/L, and from
1.08 mmol/L to 1.14 mmol/L, respectively.
4. Discussion
In this study, we examined dynamic changes in aEEG after
the induction of HI and during subsequent mild hypothermia
in newborn piglets. We found that all animals exposed to HI
exhibited severe abnormal aEEGs at 20 minutes, during the
HI period. Thereafter, however, animals who also received
selective HC for the induction of hypothermia (SHC-HI)
exhibited fewer aEEG abnormalities after this treatment
was given. Indeed, by 72 hours, no animals in this group
exhibited severe abnormal aEEGs, whereas all animals who
had HI without subsequent hypothermia (NHI) exhibited
severe abnormal aEEGs from 12 hours onwards. These
findings indicate that selective mild HC decreases the
incidence of severe aEEG abnormalities.
We also studied physiological parameters during and up
to 6 days after hypoxia/ischemia and HC. HR was decreased
in the HC group for the first 2 days, and glucose was
increased during hypoxia/ischemia in the two hypoxia/
ischemia groups and very slightly in the last 2 days in the
normal control group. No differences in MAP or plasma
electrolytes were seen.
288
J.-M. Wang et al
Figure 3 (A) Heart rate (HR); and (B) mean arterial pressure (MAP) before and during the study period. Treatment groups:
HC Z selective head cooling without hypoxia-ischemia þ nasopharyngeal temperature maintained at 35 C and rectal temperature
maintained at 36 C; NC Z normal temperature with rectal temperature maintained at approximately 39 C; NHI Z normal temperature after hypoxic-ischemic insult þ rectal temperature maintained at 39 C; SHC-HI Z head cooling (mild hypothermia) after
hypoxic-ischemic insult þ nasopharyngeal temperature maintained at 35 C and rectal temperature maintained at 36 C. * Indicates
a significant difference compared with the NC group (p < 0.05). y Indicates a significant difference compared with the HC group
(p < 0.05). z Indicates a significant difference compared with the NHI group (p < 0.05).
In the current experiment, mild cooling (to 35 C) of the
head for 24 hours was used. The degree of cooling that
would produce optimum results depends on the species.
Experiments with rats have used cooling to a rectal temperature of 32e35 C for 24 hours.17 Research with humans
has used HC to 34 C for 72 hours. In a previous study, we
found that HC to 32 C in piglets after HI produced no
further improvement over cooling to 35 C.18 Thoresen19
also pointed out that within the clinical cooling range
(33.5e37.0 C) temperature did not affect aEEG in animal
experiments.
The severe aEEG abnormalities, CLV, BS, and FT, seen at
20 minutes in hypoxic-ischemic animals, are indicative of
severe depression of cerebral function and signify acute
hypoxic-ischemic damage to the brain. The second, later
period of abnormal aEEGs is suggestive of progressive
hypoxic-ischemic damage to the central nervous system
and consequent impaired function. These two clear periods
of abnormal aEEGs may reflect cerebral energy metabolism
failure in the cerebral cortex, and they are consistent with
findings from previous studies, suggesting that acute hypoxia can lead to initial and secondary failure of energy
metabolism in the cerebral cortex.20e23 Indeed, there is
evidence to suggest that brain cell energy metabolism recovers relatively quickly in newborn animals exposed to HI
after resuscitation.24,25 Then, after a relatively stable
period of 9e24 hours, oxidative phosphorylation becomes
impaired, leading to secondary energy failure.
Previous research has shown that cerebral lactate production is increased during the 8-hour period after HI.
Presumably, the increased metabolic requirements of the
recovering brain cause a shift from oxidative to glycolytic
metabolism. HC during this period significantly decreased
cerebral lactate production, presumably because hypothermia reduced metabolic requirements. Our experiment
covered a longer time period than 8 hours, however, and we
did not collect data on lactate. The increase in glucose seen
in our study during the HI period, in those that had this
treatment, was probably due to a general stress reaction.26,27 The decrease in blood glucose at late times in NHI
and SHC-HI rats was probably a result of the hypothermia,
as it has been shown that hypothermia reduces blood
glucose in patients with severe traumatic brain injury.28 It
is speculated that the lower blood glucose at late times in
the NHI group may be a long-term effect of the hypoxia/
ischemia. In rats studied 2 weeks and 6 weeks after HI with
or without subsequent hypothermia, a reduction in
neuronal loss was seen with hypothermia treatment.17,29
From our results, it may be speculated that an improved
matching of the brain’s metabolic needs with its oxygen
supply and a lessening of neuronal loss caused by the hypothermic period contributed to the observed improvement
in aEEG at 6 days in our experiment.
Findings from the present study in newborn piglets
suggest that brain function deteriorates with time after
the induction of HI; these are generally consistent with
aEEG during HC after HI
289
Figure 4 (A) Naþ concentrations; (B) Kþ concentrations; and (C) Ca2þ concentrations before and during the study period.
Treatment groups: HC Z selective head cooling without hypoxia-ischemia þ nasopharyngeal temperature maintained at 35 C and
rectal temperature maintained at 36 C; NC Z normal temperature with rectal temperature maintained at approximately 39 C;
NHI Z normal temperature after hypoxic-ischemic insult þ rectal temperature maintained at 39 C; SHC-HI Z head cooling (mild
hypothermia) after hypoxic-ischemic insult þ nasopharyngeal temperature maintained at 35 C and rectal temperature maintained
at 36 C.
the findings from our studies in human infants.30,31 Mild
hypothermia has been shown to be one of the most
promising treatments for HIBD.4e6 Findings from a previous study suggest that mild hypothermia protects brain
tissue by alleviating further damage to mitochondria after
ischemia reperfusion, improving energy metabolism,
shortening secondary energy failure time, and preventing
apoptosis.16 In this study, we found that HI led to marked
suppression of aEEG waveforms and a predominance of BS
and FT. This was particularly pronounced 24e48 hours
after HI. These findings lead us to suggest that mild HC
significantly increases the frequency of normal aEEGs, and
in turn, promotes recovery from neuronal damage associated with HI.
Thoresen16 has shown that in infants with neonatal
asphyxia, both aEEG recorded within 6 hours after birth and
recovery time to normal aEEG, were good predictors of
outcome at 18 months in both normothermic infants and
infants treated with HC. Our study was not conducted for
poor outcome prediction, so we cannot compare our results
with those of Thoresen16. In our animal study, we recorded
aEEG over a longer time period than Thoresen16 and
included physiological measurements. Instead of using
outcome as an endpoint, we wanted to begin to explore
whether aEEG measurements at later times might have
clinical utility in monitoring recovery from neonatal
asphyxia and for assessing the impact of different treatment procedures.
What is new in this research is recording the course of
aEEG patterns in cooled versus non-cooled animals. What is
also new, is the comparison of physiological data, including
HR, MA, pH, BE, blood glucose, and electrolytes, both prior
to and after hypoxia/hypothermia between NHI and SHC-HI
groups. Our results may provide a feasible piglet model
using aEEG to explore HI-related studies such as pharmaceutical applications.
290
Our study has a number of limitations. We only included
a relatively small number of animals in each treatment
group; hence, these studies should be repeated to confirm
the observed findings. We did not perform histological examinations of brain tissue. We also did not assess the longterm consequences of the treatments because, due to
budget constraints, we are unable to support and keep
animals for a long time period. Clearly, further studies are
warranted to assess the effects of mild hypothermia after
HI on later brain and overall functioning.
In summary, the findings from our study of newborn
piglets indicate that selective mild HC decreases the incidence of severe aEEG abnormalities after HI.
Conflicts of interest
The authors have no conflicts of interest relevant to this
article.
References
1. Pisani F, Orsini M, Braibanti S, Copioli C, Sisti L, Turco EC.
Development of epilepsy in newborns with moderate hypoxicischemic encephalopathy and neonatal seizures. Brain Dev
2009;31:64e8.
2. Blackmon LR, Stark AR; American Academy of Pediatrics
Committee on Fetus and Newborn. Hypothermia: a neuroprotective therapy for neonatal hypoxic-ischemic encephalopathy. Pediatrics 2006;117:942e8.
3. Rutherford MA, Azzopardi D, Whitelaw A, Cowan F,
Renowden S, Edwards AD, et al. Mild hypothermia and the
distribution of cerebral lesions in neonates with hypoxicischemic encephalopathy. Pediatrics 2005;116:1001e6.
4. Gluckman PD, Wyatt JS, Azzopardi D, Ballard R, Edwards AD,
Ferriero DM, et al. Selective HC with mild systemic hypothermia after neonatal encephalopathy: multicentre randomised
trial. Lancet 2005;365:663e70.
5. Shankaran S, Laptook AR, Ehrenkranz RA, Tyson JE,
McDonald SA, Donovan EF, et al. Whole-body hypothermia for
neonates with hypoxic-ischemic encephalopathy. N Engl J Med
2005;353:1574e84.
6. Zhou WH, Cheng GQ, Shao XM, Liu XZ, Shan RB, Zhuang DY,
et al. Selective HC with mild systemic hypothermia after
neonatal hypoxic-ischemic encephalopathy: a multicenter
randomized controlled trial in China. J Pediatr 2010;157:
367e72. 372.e1e3.
7. McBride MC, Laroia N, Guillet R. Electrographic seizures in
neonates correlate with poor neurodevelopmental outcome.
Neurology 2000;55:506e13.
8. Sinclair DB, Campbell M, Byrne P, Prasertsom W, Robertson CM.
EEG and long-term outcome of term infants with neonatal
hypoxic-ischemic encephalopathy. Clin Neurophysiol 1999;
110:655e9.
9. Al Naqeeb N, Edwards AD, Cowan FM, Azzopardi D. Assessment
of neonatal encephalopathy by amplitude-integrated electroencephalography. Pediatrics 1999;103:1263e71.
10. ter Horst HJ, Sommer C, Bergman KA, Fock JM, van
Weerden TW, Bos AF. Prognostic significance of amplitudeintegrated EEG during the first 72 hours after birth in
severely asphyxiated neonates. Pediatr Res 2004;55:1026e33.
11. Toet MC, Hellstro
¨m-Westas L, Groenendaal F, Eken P, de
Vries LS. Amplitude integrated EEG 3 and 6 hours after birth in
full term neonates with hypoxic-ischaemic encephalopathy.
Arch Dis Child Fetal Neonatal Ed 1999;81:F19e23.
J.-M. Wang et al
12. Pfister RH, Soll RF. Hypothermia for the treatment of infants with
hypoxic-ischemic encephalopathy. J Perinatol 2010;30:S82e7.
13. LeBlanc MH, Qian XB, Cai ZW. The effect of glucose during
ischemia on brain ATP, lactate, and glutamate in piglets. Biol
Neonate 1997;72:243e54.
14. Sola
˚s AB, Kutzsche S, Vinje M, Saugstad OD. Cerebral
hypoxemia-ischemia and reoxygenation with 21% or 100% oxygen in newborn piglets: effects on extracellular levels of
excitatory amino acids and microcirculation. Pediatr Crit Care
Med 2001;2:340e5.
15. Hellstro
¨m-Westas L, Rose
´n I, Svenningsen NW. Predictive value
of early continuous amplitude integrated EEG recordings on
outcome after severe birth asphyxia in full term infants. Arch
Dis Child Fetal Neonatal Ed 1995;72:F34e8.
16. Thoresen M, Hellstro
¨m-Westas L, Liu X, de Vries LS. Effect of
hypothermia on amplitude-integrated electroencephalogram
in infants with ashyxia. Pediatrics 2010;126:e131e9.
17. Xiong M, Cheng GQ, Ma SM, Yang Y, Shao XM, Zhou WH. Postischemic hypothermia promotes generation of neural cells and
reduces apoptosis by Bcl-2 in the striatum of neonatal rat
brain. Neurochem Int 2011;58:625e33.
18. Cheng G, Sun J, Wang L, Shao X, Zhou W. Effects of selective
HC on cerebral blood flow and metabolism in newborn piglets
after HI. Early Hum Dev 2011;87:109e14.
19. Thoresen M. Supportive care during neuroprotective hypothermia in the term newborn: adverse effects and their prevention. Clin Perinatol 2008;35:749e63.
20. Berger R, Jensen A, Krieglstein J, Steigelmann JP. Cerebral
energy metabolism in immature and mature guinea pig fetuses
during acute asphyxia. J Dev Physiol 1992;18:125e8.
21. Brooks KJ, Hargreaves I, Bhakoo K, Sellwood M, O’Brien F,
Noone M, et al. Delayed hypothermia prevents decreases in Nacetylaspartate and reduced glutathione in the cerebral cortex
of the neonatal pig following transient hypoxia-ischaemia.
Neurochem Res 2002;27:1599e604.
22. Chang YS, Park WS, Lee M, Kim KS, Shin SM, Choi JH. Near
infrared spectroscopic monitoring of secondary cerebral energy failure after transient global HI in the newborn piglet.
Neurol Res 1999;21:216e24.
23. Gunn AJ, Gunn TR, de Haan HH, Williams CE, Gluckman PD.
Dramatic neuronal rescue with prolonged selective HC after
ischemia in fetal lambs. J Clin Invest 1997;99:248e56.
24. Allen KA, Brandon DH. Hypoxic ischemic encephalopathy:
Pathophysiology and experimental treatments. Newborn Infant
Nurs Rev 2011;11:125e33.
25. Cotten CM, Shankaran S. Hypothermia for hypoxic-ischemic
encephalopathy. Expert Rev Obstet Gynecol 2010;5:227e39.
26. Cazzato G, Zorzon M, Mase
` G, Iona LG. Hyperglycemia at
ischemic stroke onset as prognostic factor. Ital J Neurol Sci
1991;12:283e8.
27. Woo E, Ma JT, Robinson JD, Yu YL. Hyperglycemia is a stress
response in acute stroke. Stroke 1988;19:1359e64.
28. Zhao QJ, Zhang XG, Wang LX. Mild hypothermia therapy reduces blood glucose and lactate and improves neurologic
outcomes in patients with severe traumatic brain injury. J Crit
Care 2011;26:311e5.
29. Xiong M, Ma SM, Shao XM, Yang Y, Zhou WH. Hypoxic ischaemic
hypothermia promotes neuronal differentiation and inhibits
glial differentiation from newly generated cells in the SGZ of
the neonatal rat brain. Neurosci Lett 2012;523:87e92.
30. Jiang ZD, Brosi DM, Wang J, Xu X, Chen GQ, Shao XM, et al.
Time course of brainstem pathophysiology during first month in
term infants after perinatal asphyxia, revealed by MLS BAER
latencies and intervals. Pediatr Res 2003;54:680e7.
31. Jiang ZD, Brosi DM, Shao XM, Wilkinson AR. Sustained depression of brainstem auditory electrophysiology during the first
months in term infants after perinatal asphyxia. Clin Neurophysiol 2008;119:1496e505.
Pediatrics and Neonatology (2014) 55, 291e296
Available online at www.sciencedirect.com
ScienceDirect
journal homepage: http://www.pediatr-neonatol.com
ORIGINAL ARTICLE
Changes in Outcome and Complication Rates
of Very-low-birth-weight Infants in One
Tertiary Center in Southern Taiwan Between
2003 and 2010
Shen-Dar Chen a, Yung-Chieh Lin b, Chin-Li Lu c,d,
Solomon Chih-Cheng Chen a,e,*
a
Department of Pediatrics, Ditmanson Medical Foundation Chiayi Christian Hospital, Chiayi City,
Taiwan
b
Department of Pediatrics, National Cheng Kung University Hospital, Tainan, Taiwan
c
Department of Medical Research, Ditmanson Medical Foundation Chiayi Christian Hospital,
Chiayi City, Taiwan
d
Department of Public Health, College of Medicine, National Cheng Kung University, Tainan, Taiwan
e
School of Medicine, Taipei Medical University, Taipei City, Taiwan
Received Jul 8, 2013; received in revised form Sep 23, 2013; accepted Oct 15, 2013
Available online 10 January 2014
Key Words
length of stay;
morbidity;
mortality;
preterm
Background: Neonatal intensive care has changed dramatically over the past few decades and
the survival of infants has generally improved in many countries. The purpose of this study was
to explore the recent evolution of mortality and morbidities among very-low-birth-weight
(VLBW) infants in southern Taiwan.
Methods: We retrospectively reviewed the medical records of VLBW (birth weight <1500 g) infants who were admitted to a neonatal intensive care unit at a tertiary medical center in
southern Taiwan from 2003 to 2010. The study period was divided into two cohorts: the first
cohort of 2003e2006 and the second cohort of 2007e2010. Demographic profiles and complications were recorded, including the following information: sex, birth body weight (BBW),
gestational age (GA), Apgar score, patent ductus arteriosus (PDA), necrotizing enterocolitis,
retinopathy, chronic lung disease (CLD), inguinal hernia, and sepsis. The length of stay (LOS)
in hospital was compared between the two cohorts.
Results: A total of 420 (212 male) VLBW infants were enrolled with 52 (12.4%) deaths.
Compared to surviving infants, deceased infants had significantly lower GA, Apgar scores,
* Corresponding author. Department of Pediatrics, Ditmanson Medical Foundation Chiayi Christian Hospital, Taiwan, Number 539,
Zhongxiao Road, East District, Chiayi City 60002, Taiwan.
E-mail address: [email protected] (S.C.-C. Chen).
1875-9572/$36 Copyright ª 2013, Taiwan Pediatric Association. Published by Elsevier Taiwan LLC. All rights reserved.
http://dx.doi.org/10.1016/j.pedneo.2013.10.010
292
S.-D. Chen et al
and BBW. The mortality of VLBW infants remained static between the two birth cohorts, but
the incidence of major morbidities generally decreased. The LOS for overall surviving infants
and the proportion of LOS > 60 days were both reduced in the period of 2007e2010. With
further stratification by BBW, the major reduction of long LOS was only found in the group
of BBW 1000 g. The multivariate logistic regression model found PDA, CLD, and
BBW < 1000 g were major complications to be associated with long LOS among surviving infants.
Conclusion: Periodic evaluation of the mortality and morbidity of preterm infants can help to
understand the changes and trends of our neonatal care. Further study using the national dataset to provide more representative information is warranted.
Copyright ª 2013, Taiwan Pediatric Association. Published by Elsevier Taiwan LLC. All rights
reserved.
1. Introduction
Neonatal intensive care has changed dramatically over the past
decades, with advances in technology such as high-frequency
oscillation,1,2 inhaled nitric oxide,3 antenatal corticosteroids,
and surfactant therapy.4,5 The survival rate of preterm infants
has improved,6e8 but morbidity of survivor infants remained
static or even increased,9,10 especially for those infants with
lower gestational age (GA) or low birth weight.
Taiwan, an island country in southeastern Asia with a
population of 23 million, has established a high-quality
medical care system during the past decades, and the general infant mortality in Taiwan has improved.11 There have
been several previous reports regarding the incidences and
survival rates of very-low-birth-weight (VLBW) infants in
Taiwan.12e18 The most recent paper was published in 2008
regarding the outcomes of VLBW between 1999 and 2006 at
one tertiary medical center in central Taiwan.18 However,
there was no report on VLBW for southern Taiwan in recent
years. We wondered how medical advances had affected the
outcomes for VLBW infants in the neonatal intensive care
unit (NICU) recently and if there were any geographic differences. Thus, the purpose of this study was to explore the
evolution of mortality and morbidities among VLBW infants
in southern Taiwan between 2003 and 2010.
2. Materials and Methods
2.1. Ethical approval
The Ethics Review Board of Ditmanson Medical Foundation
Chiayi Christian Hospital (Chiayi City, Taiwan) approved the
study protocol. The data was collected without identification. Because this was a retrospective data analysis, no
informed consent was necessary.
2.2. Cases and enrollment criteria
This study retrospectively reviewed the medical records of
VLBW (birth body weight <1500 g) infants who were
consecutively admitted to the NICU of a tertiary medical
center from 2003 to 2010. In order to assess the evolution of
the mortality and morbidity of VLBW infants across the
study period, we divided the cases into two birth cohorts:
the first cohort encompassing the years 2003e2006 and the
second cohort consisting of the years 2007e2010.
2.3. Demographic data and outcomes
Demographic information including sex, birth body weight
(BBW), GA, and 1st minute and 5th minute Apgar scores were
recorded. The main outcomes were death and any
morbidity, including patent ductus arteriosus (PDA),
necrotizing enterocolitis (NEC), retinopathy of prematurity
(ROP), intraventricular hemorrhage (IVH), chronic lung
disease (CLD), inguinal hernia (IH), and sepsis. PDA was
diagnosed by a pediatric cardiologist through heart echo
examination. Grades 3 and 4 IVH were defined as severe
IVH. NEC was diagnosed by the presence of intestinal
intramural gas on X-ray film, perforation, or the finding of
necrosis during an operation. Sepsis was confirmed by
positive blood culture. CLD was defined as dependence on
supplemental oxygen at 36 weeks’ postmenstrual age.19
Treated ROP was defined as treatment with pan-retinal
photocoagulation performed by ophthalmologists.
The secondary measured outcome was the length of stay
(LOS) in hospital. Because there was no definition for long LOS
in previous literature, we subjectively chose the cutoff of
LOS over 60 days, in order to compare the proportion change
of LOS between two time periods. One reason for choosing 60
days as the cutoff was that the median LOS of all live births in
this study was 53 days. Deceased infants were excluded from
the analysis of LOS, because these infants usually had a
shorter LOS, reducing the possibility of developing some
specific complications due to their premature death.
2.4. Statistical analyses
Data were analyzed using Microsoft Excel and IBM SPSS Statistics for Windows (Version 21.0. Armonk, NY: IBM Corp.).
Continuous variables were expressed as mean standard
deviation (SD), but the LOS of each complication was
expressed as median and interquartile range (IQR). Student t
test was used for continuous variables and the nonparametric median test was used to test the median (IQR)
of LOS for infants with/without each complication because
of the skewness of the data. Chi-square test was used for
categorical data. A p value <0.05 was considered significant.
Multivariate logistic regression analysis adjusting for confounders was performed by using LOS over 60 days as the
Outcomes of VLBW infants in Taiwan
293
Table 1 Characteristics of alive or dead very-low-birthweight infants.
Gestational age,
wk
Apgar 1
Apgar 5
Birth body
weight, g
Length of hospital
stay, d
Sex, male (%)
PDA, n (%)
Sepsis, n (%)
ROP, n (%)
CLD, n (%)
Severe IVH, n (%)
NEC, n (%)
Hernia, n (%)
Alive
(n Z 368)
Dead
(n Z 52)
p value
28.9 3.0
26.4 2.9
<0.001
5.0 2.3
7.2 2.4
1074 260
3.2 2.0
5.6 2.7
797 305
<0.001
<0.001
<0.001
67.0 42.8
18.7 28.4
<0.001
182 (49.5%)
125 (34.0%)
126 (34.2%)
102 (27.7%)
113 (30.7%)
20 (5.4%)
23 (6.3%)
34 (9.2%)
30 (57.7%)
22 (42.3%)
7 (13.5%)
3 (5.8%)
7 (13.5%)
4 (7.7%)
1 (1.9%)
2 (3.8%)
0.266
0.238
0.003
0.001
0.010
0.512
0.208
0.193
PDA Z patent ductus arteriosus; ROP Z retinopathy of prematurity; CLD Z chronic lung disease; IVH Z intraventricular
hemorrhage; NEC Z necrotizing enterocolitis.
dependent variable. In the regression model, we used BBW
but not GA, because these two factors have a high correlation, and GA has a higher possibility of error due to memory
bias or poor estimates by the obstetrician.
3. Results
3.1. Basic information of surviving and deceased
infants
This study analyzed the basic data and clinical information of
VLBW infants admitted to the NICU at a tertiary medical
center in southern Taiwan during the years 2003e2010. A
Table 2
total of 420 (212 male, 50.5%) VLBW infants were enrolled.
There were 52 deaths, giving a mortality rate of 12.4%. The
characteristics of surviving and deceased VLBW infants are
compared in Table 1. Compared to surviving infants,
deceased infants had significantly lower GA, lower Apgar
scores, lower BBW, shorter LOS, and lower incidence rates of
all complications except for severe IVH and PDA (Table 1).
3.2. Comparison of complications in two birth
cohorts
The GA, BBW, sex, and clinical outcomes of VLBW infants
stratified by BBW < 1000 g and BBW 1000 g in two birth
cohorts (2003e2006 and 2007e2010), were compared in
Table 2. All complications occurred less frequently in the
latter period except for death, severe IVH, and IH among
the group of BBW < 1000 g, although no significant statistical difference was observed in the frequency of these
three complications.
3.3. Comparison of LOS in two birth cohorts
The LOS of 368 surviving infants with/without each
complication, stratified by two birth cohorts, is compared
in Table 3. Infants with any complication or BBW < 1000 g
usually had a longer LOS than infants without any
complication.
The overall median (IQR) of LOS for surviving infants was
reduced from 62.0 days (38.5e 83.0 days) in 2003e2006 to 48.0
days (37.0e72.3 days) in 2007e2010, with a significant
p Z 0.036. The proportion of long LOS, i.e., >60 days, was
significantly higher among the 2003e2006 birth cohort than
the 2007e2010 birth cohort (57.4% vs. 38.4%, p Z 0.001). We
further performed a stratified analysis by BBW 1000 g and
BBW < 1000 g: a higher proportion of long LOS was only found
among the groups with BBW 1000 g (32.3% for the 2003e2006
birth cohort vs. 19.5% for the 2007e2010 birth cohort,
p Z 0.035), but it was not found among the group with
BBW < 1000 g (85.6% for the 2003e2006 birth cohort vs. 84.8%
Mortality and morbidity of low-birth-weight infants stratified by birth body weight less than or 1000 g or more.
BBW < 1000 g (n Z 183)
Subtotal
N
%
2003e2006
N
%
2007e2010
N
BBW 1000 g (n Z 237)
p value
%
Sex, male (%)
88
48.1 46
43.4 42
54.5
Death
39
21.3 20
18.9 19
24.7
PDA
87
47.5 58
54.7 29
37.7
Sepsis
78
42.6 52
49.1 26
33.8
ROP
74
40.4 57
53.8 17
22.1
CLD
85
46.4 53
50.0 32
41.6
Severe IVH
13
7.1
6
5.7
7
9.1
NEC
15
8.2 14
13.2
1
1.3
Hernia
24
13.1 12
11.3 12
15.6
Gestational age,
26.4
2.5 26.5
2.5 26.3
2.7
week (mean, SD)
BBW,
772
155
773
154
770
156
g (mean, SD)
Subtotal
N
0.136
0.344
0.023
0.039
0.000
0.258
0.372
0.004
0.399
0.556
0.895
124
13
60
55
31
35
11
9
12
30.2
1246
%
52.3
5.5
25.3
23.2
13.1
14.8
4.6
3.8
5.1
2.4
152
2003e2006
N
%
44
6
37
28
20
20
9
8
6
30.2
1253
41.9
5.7
35.2
26.7
19.0
19.0
8.6
7.6
5.7
2.5
158
2007e2010
N
p value
%
80
7
23
27
11
15
2
1
6
30.1
1240
60.6
5.3
17.4
20.5
8.3
11.4
1.5
0.8
4.5
2.3
149
0.004
0.890
0.002
0.260
0.015
0.098
0.010
0.006
0.684
0.771
0.516
BBW Z birth body weight; PDA Z patent ductus arteriosus; ROP Z retinopathy of prematurity; CLD Z chronic lung disease; IVH Z
intraventricular hemorrhage; NEC Z necrotizing enterocolitis; SD Z standard deviation.
294
S.-D. Chen et al
Table 3 Univariate analysis for the mean and standard deviation of length of stay in hospital of 335 survivors with/without
each complication, stratified by two birth cohorts.
Year 2003e2006
Without
Year 2007e2010
With
p value
Complication
N
Mean
SD
N
Mean
SD
Sex (male vs. female)
PDA
Sepsis
ROP
CLD
Severe IVH
NEC
Hernia
BBW (1000 g vs. <1000 g)
71
97
100
105
108
162
155
159
93
71.9
62.7
57.3
57.8
54.3
68.8
67.6
66.9
54.0
34.9
35.1
25.8
37.3
22.5
31.8
37.4
34.1
32.4
105
79
76
71
68
14
21
17
83
68.9
79.3
87.0
88.4
95.2
85.7
88.4
100.4
88.1
38.3
37.2
42.1
27.6
41.3
73.9
26.3
48.2
33.2
0.610
0.003
0.000
0.000
0.000
0.409
0.015
0.000
0.000
Without
With
p value
N
Mean
SD
N
Mean
SD
94
114
110
133
116
154
158
142
113
64.7
60.2
54.9
61.9
54.2
62.8
63.4
58.3
48.1
43.8
54.7
29.6
51.3
36.9
48.5
48.6
41.4
19.0
65
45
49
26
43
5
1
17
46
61.9
72.1
83.1
71.9
88.8
87.6
88
107.7
101.5
54.7
25.1
72.0
28.7
64.8
44.2
e
75.7
72.6
0.726
0.161
0.011
0.341
0.000
0.261
0.614
0.017
0.000
SD Z standard deviation; PDA Z patent ductus arteriosus; ROP Z retinopathy of prematurity; CLD Z chronic lung disease; IVH Z
intraventricular hemorrhage; NEC Z necrotizing enterocolitis; BBW Z birth body weight.
for the 2007e2010 birth cohort, p Z 0.907) as shown in
Figure 1. Nevertheless, as the LOS was treated as a continuous
variable, the differences of median were not statistically significant irrespective of BBW 1000 g or BBW < 1000 g. Among
the group of BBW 1000 g, the median LOS was 47.0 days (IQR:
36.5e66.5) for the 2003e2006 birth cohort and 46.0 days (IQR:
36.0e55.0) for the 2007e2010 birth cohort (p Z 0.229); among
the group of BBW < 1000 g, LOS was also similar between these
two birth cohorts [median (IQR) of LOS for the 2003e2006 birth
cohort was 84.0 days (72.0e 104.0 days) and for the 2007e2010
birth cohort it was 86.0 days (70.8e100.5 days), p Z 0.937].
3.4. Risk factors associated with long LOS in
survivors
Table 4 shows the odds ratio (OR) and 95% confidence interval (95% CI) of each complication to be associated with
long LOS > 60 days in the multivariate logistic regression
modeling of the two birth cohorts. PDA, CLD, and
BBW < 1000 g were the major complications to be associated with long LOS > 60 days. Sepsis and hernia were factors which were only significant in the second cohort, but
the relatively wide 95% CI suggested an unstable statistic
probably due to a small case number.
Fortunately, our results did not show an increase of incidences of morbidity despite the stationary mortality between two birth cohorts. Some previous studies found that
morbidities were mostly shifted to respiratory problems.6,7
However, this study found that the impact of CLD on long
LOS has decreased, as its OR of being associated with long
LOS > 60 days has decreased from 4.57 to 2.92 (Table 4). The
median LOS of infants with CLD also decreased from 95 days in
the first cohort to 88 days in the second cohort (Table 3). The
evidence above suggests a remarkable improvement in the
quality of respiratory care. As described in other papers but
not investigated in this study, the increased use of nasal
continuous positive airway pressure, early surfactant treatment, and rapid extubation20 may have contributed to
reducing the CLD-related LOS and minimizing the clinical
burden of respiratory complications of VLBW infants in Taiwan.
Septicemia is a major antecedent of morbidity and
mortality in VLBW infants and can prolong hospital stay
among VLBW survivors.21,22 The present study found that
4. Discussion
This study summarized the mortality and morbidity of VLBW
infants at a tertiary medical center in southern Taiwan from
2003 until 2010. The mortality remained static, but the
incidences of morbidities and the median of LOS of surviving infants decreased significantly from the first cohort
(2003e2006) to the second cohort (2007e2010). The reasons for such improvement, although not clear, may be due
to a combination of several factors including the advancement of medical intervention, improved prenatal care, and
more skilled medical professionals in Taiwan.
Previous studies showed that mortality of preterm infants
may decline due to improvement of neonatal care,6e8 but
morbidity may, by contrast, increase because infants with
morbidities that might previously have died now survive.9,10
Figure 1 Stacked bar charts representing proportions of <30
days, 31e60 days, 61e90 days, and >90 days for hospital
length of stay during two study periods (2003e2006,
2007e2010) are depicted for live births with birth body weight
(BBW) 1000 g and BBW < 1000 g, respectively.
Outcomes of VLBW infants in Taiwan
295
Table 4 The odds ratio and 95% confidence interval of each complication associated with long length of stay >60 days in
multivariate logistic regression modeling, stratified by two birth cohorts of years 2003e2006 and 2007e2010.
Year 2003e2006 (N Z 185)
Sex
PDA
Sepsis
ROP
CLD
Severe IVH
NEC
Hernia
BBW
Male vs.
With vs.
With vs.
With vs.
With vs.
With vs.
With vs.
With vs.
<1000 g
female
without
without
without
without
without
without
without
vs. 1000 g
Year 2007e2010 (N Z 183)
OR
95% CI
p value
OR
95% CI
p value
1.182
2.912
1.977
2.014
4.565
1.082
4.378
5.020
5.120
0.49e2.87
1.28e6.62
0.85e4.60
0.82e4.96
1.76e11.83
0.24e4.84
0.97e19.77
0.76e33.05
2.19e12.00
0.712
0.011
0.114
0.128
0.002
0.918
0.055
0.093
<0.001
0.789
4.413
3.763
2.740
2.920
5.176
1.189
6.048
5.661
0.32e1.97
1.63e11.96
1.50e9.41
0.78e9.66
1.07e7.96
0.23e116.72
0.25e5.69
1.42e25.72
2.26e14.21
0.611
0.004
0.005
0.117
0.036
0.301
0.828
0.015
<0.001
OR Z odds ratio; CI Z confidence interval; PDA Z patent ductus arteriosus; ROP Z retinopathy of prematurity; CLD Z chronic lung
disease; IVH Z intraventricular hemorrhage; NEC Z necrotizing enterocolitis; BBW Z birth body weight.
infants with sepsis were more likely to have a significantly
longer LOS (Table 3). Sepsis was a significant factor to be
associated with long LOS > 60 days in the second cohort,
but not the first cohort. The reason for its resurgence in the
second cohort was not clear. We guess it might be due to
the redistribution of other competing causes. Whether the
advancement of medical intervention, such as the administration of central lines, contributed to its occurrence requires further studies for clarification.
Screening for ROP was highly recommended for VLBW
infants or GA 31 weeks.23 Surviving infants with ROP were
associated with a longer LOS, although this difference was
only significant in the first cohort (Table 3). The incidence of
ROP was reduced from 36.5% in the first cohort to 13.5% in the
second cohort (Table 2). A previous study suggested septicemia and respiratory problems were two major risk factors
for developing ROP.24 There was a much lower incidence of
sepsis and CLD in the second cohort of our study, which may
explain the reduction of ROP incidence among VLBW infants.
As found in previous studies, symptomatic PDA may
occur in up to 50% of VLBW infants.25 The incidence rate of
PDA was 35% among all cases in this study and was slightly
higher in deceased infants than surviving infants (Table 1).
The incidence of PDA seemed to reduce over time, as it
occurred in 45% of the first cohort but only in 25% of the
second cohort (Table 2). One possible reason could be that
the heart echo examination was routinely performed for
every VLBW infant in the first period, but it was only performed for those VLBW infants with symptoms/signs in the
second period. Therefore, the incidence of PDA could be
severely underscored in the second birth cohort. Nevertheless, PDA was a significant risk factor associated with
long LOS > 60 days in both cohorts (Table 4). Surviving infants with PDA had longer LOS than those without (Table 3).
A diagnosis of NEC in a VLBW infant may impose a significant additional LOS and medical burden on the neonatal
community as a whole.26 Our study found that infants with
NEC had much longer LOS compared to those infants
without NEC (Table 3). The case number of NEC dramatically dropped from 21 in the first cohort to only one in the
second cohort (Table 3). We considered that the great
improvement was due to early detection of potential NEC
cases. Attention to vulnerable cases and early detection
could have reduced the chance of NEC, leading to a subsequent decrease of medical burden.26
IH is a common disease seen in pediatric practice and may
cause complications such as abdominal distension, feeding
intolerance, and incarceration. Prematurity is the single
most important predisposing factor for the development of
IH.27 Compared to full-term newborns, its incidence is relatively high in VLBW infants.28 We found that IH was more
common in male infants (13.7%) than female infants (3.4%),
which was in agreement with previous studies that male sex
was an important factor significantly associated with IH.29
Although IH was not a significant factor associated with
mortality in this study (Table 2), infants with IH had much
longer LOS than those infants without IH (Table 3), which was
consistent with one previous study.29 Finally, in the multivariate logistic regression, infants with IH had 5e6 times the
OR of having long LOS > 60 days, although the OR had a wide
range of 95% CI (Table 4).
Some limitations of this study should be mentioned. First,
it was based on only one neonatal care center and the sample
size was not large, although we followed for a long period
spanning 8 years. A relatively small size may adversely affect
the statistical power; thus, the 95% CI of some risk factors
were relatively wide. However, we consider that this study
still provides valuable information about the evolution of
neonatal care in Taiwan. Second, this study was conducted at
a tertiary medical center in southern Taiwan. Because the
approaches to neonatal care and outcomes may vary among
centers,30,31 the representativeness of the outcomes of this
study may be limited. Thus, if possible, we would like to
analyze the national registry dataset of preterm babies in
order to get a whole picture of neonatal care in Taiwan.
In conclusion, deceased infants have much lower GA,
Apgar scores, and BBW than surviving infants. The mortality
of VLBW infants remained static between two birth cohorts,
but the incidence rates of major morbidities were generally
reduced. The median LOS for overall surviving infants and
the proportion of long LOS > 60 days were both reduced in
the second birth cohort 2007e2010. However, if stratified
by BBW 1000 g and BBW < 1000 g, the significant
reduction of proportion of long LOS was only found among
the groups with BBW 1000 g. PDA, CLD, and BBW < 1000 g
were three major complications associated with long LOS
296
among surviving infants. Evaluation of the mortality and
morbidity of preterm infants across a long period can help
to understand the changes and trends of neonatal care.
Further study using the national registry dataset to provide
more representative information is warranted.
Conflicts of interest
All authors declare no conflicts of interest.
Acknowledgments
We highly appreciate the great help of Mr. Darren Wu for
English editing. No financial support was received.
References
1. Tang JR, Yau KI, Shih HH. High-frequency oscillatory ventilation for infants and children with adult respiratory distress
syndrome. Zhonghua Min Guo Xiao Er Ke Yi Xue Hui Za Zhi
1997;38:137e44.
2. Rimensberger PC, Beghetti M, Hanquinet S, Berner M. First intention high-frequency oscillation with early lung volume optimization
improves pulmonary outcome in very low birth weight infants with
respiratory distress syndrome. Pediatrics 2000;105:1202e8.
3. Hallman M, Aikio O. Nitric oxide in critical respiratory failure of very
low birth weight infants. Paediatr Respir Rev 2004;5:S249e52.
4. Wang YC, Tseng HI, Yang SN, Lu CC, Wu JR, Dai ZK, et al. Effects
of antenatal corticosteroids on neonatal outcomes in very-lowbirth-weight preterm newborns: a 10-year retrospective study
in a medical center. Pediatr Neonatol 2012;53:178e83.
5. Modanlou HD, Beharry K, Padilla G, Iriye B. Combined effects
of antenatal corticosteroids and surfactant supplementation
on the outcome of very low birth weight infants. J Perinatol
1996;16:422e8; quiz 429e30.
6. Stoelhorst GM, Rijken M, Martens SE, Brand R, den Ouden AL,
Wit JM, et al. Changes in neonatology: comparison of two cohorts of very preterm infants (gestational age <32 weeks): the
Project On Preterm and Small for Gestational Age Infants 1983
and the Leiden Follow-Up Project on Prematurity 1996e997.
Pediatrics 2005;115:396e405.
7. Young TE, Kruyer LS, Marshall DD, Bose CL. Population-based
study of chronic lung disease in very low birth weight infants in
North Carolina in 1994 with comparisons with 1984. The North
Carolina Neonatologists Association. Pediatrics 1999;104:e17.
8. Zeitlin J, Ancel PY, Delmas D, Bre
´art G, Papiernik E, EPIPAGE
and MOSAIC Ile-de-France Groups. Changes in care and outcome
of very preterm babies in the Parisian region between 1998 and
2003. Arch Dis Child Fetal Neonatal Ed 2010;95:F188e93.
9. Lemons JA, Bauer CR, Oh W, Korones SB, Papile LA, Stoll BJ,
et al. Very low birth weight outcomes of the National Institute
of Child health and human development neonatal research
network, January 1995 through December 1996. NICHD
Neonatal Research Network. Pediatrics 2001;107:E1.
10. Fanaroff AA, Stoll BJ, Wright LL, Carlo WA, Ehrenkranz RA,
Stark AR, et al. Trends in neonatal morbidity and mortality for very
low birthweight infants. Am J Obstet Gynecol 2007;196:147. e1e8.
11. Life statistics: Infant mortality. Taiwan National Statistical
Bureau.
Available
at:
http://www.stat.gov.tw/ct.asp?
xItemZ15409&CtNodeZ3622&mpZ4; 2012. Accessed May
23, 2013.
12. Chen SJ, Guo NW, Wang PF, Hwang BT. Outcome for very low
birth-weight infants. Zhonghua Min Guo Xiao Er Ke Yi Xue Hui
Za Zhi 1989;30:30e9.
S.-D. Chen et al
13. Teng RJ, Tsou Yau KI, Lu CP, Lee CY. The neonatal morbidity
and mortality of low birth weight neonates. Zhonghua Min Guo
Xiao Er Ke Yi Xue Hui Za Zhi 1992;33:167e75.
14. Chen TJ, Lin CH, Wang CJ, Wang ST, Yeh TF. Vital statistics of
premature and low birthweight infants in Tainan area. Zhonghua Min Guo Xiao Er Ke Yi Xue Hui Za Zhi 1996;37:439e43.
15. Tsao PN, Wu TJ, Teng RJ, Tang JR, Yau KI. Comparison of the
outcome of extremely-low-birth-weight infants between two
periods. Zhonghua Min Guo Xiao Er Ke Yi Xue Hui Za Zhi 1998;
39:233e6.
16. Tsou KI, Tsao PN, Taiwan Infant Development Collaborative
Study Group. The morbidity and survival of very-low-birthweight infants in Taiwan. Acta Paediatr Taiwan 2003;44:349e55.
17. Lin YC, Lin YJ, Lin CH. Growth and neurodevelopmental outcomes of extremely low birth weight infants: a single center’s
experience. Pediatr Neonatol 2011;52:342e8.
18. Hsu CF, Chen CH, Wang TM, Hsu CC. Changes in outcome among
very low birth weight infants at one NICU. Clin Neonatol 2008;
15:36e42.
19. Tai PY, Hsu CH, Kao HA, Hung HY, Chang JH, Jim WT, et al. Risk
factors for chronic lung disease in very low birth weight infants: a five-year multicenter study in Taiwan. Clin Neonatol
2005;12:13e8.
20. Friedman CA, Menchaca RC, Baker MC, Rivas CK, Laberge RN,
Rios EH, et al. Bubble nasal CPAP, early surfactant treatment, and
rapid extubation are associated with decreased incidence of
bronchopulmonary dysplasia in very-low-birth-weight newborns:
efficacy and safety considerations. Respir Care 2013;58:1134e42.
21. Fanaroff AA, Korones SB, Wright LL, Verter J, Poland RL,
Bauer CR, et al. Incidence, presenting features, risk factors
and significance of late onset septicemia in very low birth
weight infants. The National Institute of Child Health and
Human Development Neonatal Research Network. Pediatr
Infect Dis J 1998;17:593e8.
22. Stoll BJ, Hansen N, Fanaroff AA, Wright LL, Carlo WA,
Ehrenkranz RA, et al. Late-onset sepsis in very low birth weight
neonates: the experience of the NICHD Neonatal Research
Network. Pediatrics 2002;110:285e91.
23. Chiang MC, Tang JR, Yau KI, Yang CM. A proposal of screening
guideline for retinopathy of prematurity in Taiwan. Acta Paediatr Taiwan 2002;43:204e7.
24. Teng RJ, Wu TJ, Yau KI. Retinopathy of prematurity in verylow-birthweight neonates: epidemiology and risk factors.
Zhonghua Min Guo Xiao Er Ke Yi Xue Hui Za Zhi 1997;38:370e4.
25. Mouzinho AI, Rosenfeld CR, Risser R. Symptomatic patent
ductus arteriosus in very-low-birth-weight infants: 1987e1989.
Early Hum Dev 1991;27:65e77.
26. Bisquera JA, Cooper TR, Berseth CL. Impact of necrotizing
enterocolitis on length of stay and hospital charges in very low
birth weight infants. Pediatrics 2002;109:423e8.
27. Wang H, Gao X, Liu C, Yan C, Lin X, Yang C, et al. Morbidity and
mortality of neonatal respiratory failure in China: surfactant
treatment in very immature infants. Pediatrics 2012;129:
e731e40.
28. Wen S, Guo Y, Xie R, Dy J, Walker M. Secular trends in trial of
labor and associated neonatal mortality and morbidity in the
United States, 1995 to 2002. Zhong Nan Da Xue Xue Bao Yi Xue
Ban 2012;37:1088e96.
29. Isayama T, Lee SK, Mori R, Kusuda S, Fujimura M, Ye XY, et al.
Comparison of mortality and morbidity of very low birth weight
infants between Canada and Japan. Pediatrics 2012;130:e957e65.
30. Stoll BJ, Hansen NI, Bell EF, Shankaran S, Laptook AR, Walsh MC,
et al. Neonatal outcomes of extremely preterm infants from the
NICHD Neonatal Research Network. Pediatrics 2010;126:443e56.
31. Marshall G, Luque MJ, Gonzalez A, D’Apremont I, Musante G,
Tapia JL. Center variability in risk of adjusted length of stay for
very low birth weight infants in the Neocosur South American
Network. J Pediatr (Rio J) 2012;88:524e30.
Pediatrics and Neonatology (2014) 55, 297e305
Available online at www.sciencedirect.com
ScienceDirect
journal homepage: http://www.pediatr-neonatol.com
ORIGINAL ARTICLE
Histopathological Evaluation of Horse
Serum-induced Immune Complex Vasculitis
in Swine: Implication to Coronary Artery
Lesions in Kawasaki Diseasey
Saji Philip a,b,c,*, Wen-Chuan Lee c, Mei-Hwan Wu d,
Cherian Kotturathu Mammen a,b, Hung-Chi Lue d
a
Department of Pediatric Cardiology, St. Gregorios Cardio-Vascular Center, Parumala, Kerala, India
Department of Cardiothoracic Surgery, Fontier Lifeline Hospital, Dr. K. M. Cherian Heart Foundation,
Ambattur, Chennai, India
c
Division of Biotechnology, Cardiovascular Research Center, Animal Technology Institute, Miaoli 350, Taiwan
d
Division of Pediatric Cardiology, Department of Pediatrics, National Taiwan University Children’s
Hospital, National Taiwan University, Number 7, Chung-Shan South Road, Taipei 100, Taiwan
b
Received Jun 17, 2013; received in revised form Oct 6, 2013; accepted Oct 24, 2013
Available online 1 February 2014
Key Words
coronary artery
lesions;
histopathology;
immune complex
vasculitis;
Kawasaki disease;
two-dimensional
echocardiography
Background: Immune complex (IC) vasculitis can be experimentally induced in animal models
by intravenous injection of horse serum (HS), and the findings of HS-induced IC vasculitis in
swine were very similar to that of Kawasaki disease (KD). The IC mechanism may be involved
in the pathogenesis of vasculitis in KD. Here, we studied the two-dimensional (2D) echocardiographic and histopathological findings of acute, subacute, and healing phases of vasculitis
induced by two different types of HS, and the reproducibility of IC vasculitis in swine.
Methods and results: Our study group consisted of 24 pure-bred landrace male piglets of 1.5e3
months of age. They were divided into three HS groups (n Z 17), namely, Group A (n Z 8)
receiving gamma globulin-free HS, and Group B (n Z 6) receiving donor herd HS, three doses
at 5-day intervals, and Group C (n Z 3) that received only one dose of donor herd HS on Day 1,
and the saline group (n Z 7) that received three doses of intravenous normal saline (NS) at 5day intervals. The 2D echocardiography was performed every 3e4 days, and all piglets were
killed for histopathological studies at different dates from Days 2 to Day 60. All the HS groups
developed rashes and demonstrated significant dilation (54e150%) of coronary arteries in
Groups A and B; when compared (p < 0.02) with 9e53% dilation in Group C and the saline
group. Histopathological changes of test groups were asymmetric coronary vasculitis in various
stages, whereas none of the piglets in the control group developed vasculitis. No significant
y
This research was conducted at the Animal Technology Institute of Taiwan with the assistance of the Cardiac Children’s Foundation.
* Corresponding author. Dr. K M. Cherian Heart Foundation, Parumala, Pathanamthitta District, Kerala 689626, India.
E-mail addresses: [email protected], [email protected] (S. Philip).
1875-9572/$36 Copyright ª 2014, Taiwan Pediatric Association. Published by Elsevier Taiwan LLC. All rights reserved.
http://dx.doi.org/10.1016/j.pedneo.2013.10.012
298
S. Philip et al
difference in the echocardiographic and histopathological findings was observed among the
piglets that received two types of HS.
Conclusion: HS can induce IC vasculitis in swine. The rashes and 2D echocardiographic and histopathological studies of the acute to healing phases showed close similarities with KD, and it
is concluded that swine may serve as a unique experimental model for IC vasculitis and for
various therapeutic trials.
Copyright ª 2014, Taiwan Pediatric Association. Published by Elsevier Taiwan LLC. All rights
reserved.
1. Introduction
Kawasaki disease (KD), an acute febrile disease with a
systemic vasculitis, has become a leading cause of acquired
heart disease other than rheumatic heart disease in many
developed countries.1 Coronary artery lesions (CALs) with
aneurismal dilation, thrombosis, and/or stenosis, leading to
myocardial infarction and death have been recognized as
the most severe complication.2 Circulating immune complexes (ICs), triggered by infectious agents, bacteria, viral,
or other unknown causes, have been detected in the early
phase of KD patients, suggesting that immunopathological
mechanisms might be involved in the pathogenesis of
Table 1
vasculitis in KD (Table 1).3e22 Attempts to produce coronary
vasculitis in animal models have been tried in mice,
weanling rabbits, and guinea pigs by injecting infectious
agents, foreign proteins, Lactobacillus casei cell walls, and
horse serum (HS).23e26 Swine is a unique and promising
animal for biomedical research, especially in the field of
cardiovascular diseases.27 IC vasculitis induced in swine
showed rashes, and significant dilations of echocardiographic CA. In addition, the histopathological changes in
the subacute stage of vasculitis were closely related to the
lesions in KD, and thus we postulate that IC-mediated
mechanisms may play a significant role in the pathogenesis of CALs in KD.26e28 Here, we further evaluated the HS-
Summary of studies on detection of immune complexes in KD patients.
No. of
studies
Authors
Year
No. of
KD cases
Positive for IC
Methods of detection
1.
2.
Fossard and Thompson3
Sawa4
1977
1979
1
15
Stronglyþ
6 (41%)
3.
4.
5.
Weindling et al5
Eluthesen et al6
Furuse and Matsuda7
1979
1981
1983
1
81
16
6.
7.
8.
9.
Yanase et al8
Miyata et al9
Takiguchi et al10
Mason et al11
1984
1984
1984
1985
30
32
35
42
1 (100%)
48 (59%)
50% and 26.6%
(25 samples)
100%
11 (34.4%)
26%
29 (69%)
10.
11.
Ono et al12
Levin et al13
1985
1985
32
19
11 (34%)
13 (68%)
12.
Pachman et al14
1987
6
13.
14.
Levin et al15
Lin and Hwang16
1987
1987
19
20
6 (100%)
5 (83%)
19 (100%)
70%
15.
16.
Fujimoto et al17
Ohshio et al18
1987
1987
67
43
50 (75%)
22 (51%)
17.
18.
19.
20.
Salcedo et al19
Salo et al20
Li et al21
Koike22
1988
1988
1990
1991
1
27
17
11
1 (100%)
99%
8 (47%)
11 (100%)
Platelet aggregation test
Raji cell method
Inhibition latex agglutination
Inhibition latex agglutination
C1q solid phase array
In acute and remission phase
C1q binding assay
No significant positive titer
C1q binding assay
Antibody inhibition test
Raji cell method
C1q solid phase array
C1q binding assay
Polyethylene glycol/
Precipitation method
Deposition in coronary artery
in the myocardium (IHS)
(Platelet interaction study)
Polyethylene glycol method
60% Raji cell method
C1q enzyme immune assay
ELISA solid phase
Anti-C3 assays
IC deposition in kidney (IF)
C1q binding assay
Polyethylene glycol method
Sodium dodecyl sulfate
polyacrylamide gel
electrophoresis
ELISA Z enzyme-linked immunosorbent assay; IC Z immune complex; IF Z immunofluorescent sections; IHS Z immunohistochemical
staining; KD Z Kawasaki disease; þ Z positive.
Horse Serum-Induced Immune Complex Vasculitis in Swine
299
induced vasculitis in swine during the acute, subacute, and
healing phases, from 2 days to 60 days of follow up, by twodimensional (2D) echocardiography and histopathological
studies. In addition, we also attempted to establish the
reproducibility of IC coronary vasculitis in swine with two
different kinds of HS infusions. The implications to CALs in
KD are discussed in this study.
virus, mycoplasma, and cytopathogenic agents with a total
protein content of 5.2 g/dL and <5% gamma globulin content (catalog number 16270-035, Lot no. 1026238; Gibco
BRL Life Technologies, Gaithersburg, MD, USA); and the HS2
was donor herd heat-inactivated, virus- and mycoplasmafree HS, with a total protein content of 6.6 g/dL and 1.49 g/
dL gamma globulin content (catalog number 29211Z0, Lot
no. R15383 ICN; MP Biomedicals, Santa Ana, California,
USA). The serum was slowly infused at around 10 drops/
minute and then increased to 20e40 drops/minute for
subsequent doses to avoid severe immediate reactions such
as chills and cyanosis, respiratory distress, convulsions,
shock.
In Group A (n Z 8), 10 mL (0.52 g protein)/kg of HS1 was
intravenously infused and in Group B (n Z 6), 10 mL (0.7 g
protein)/kg of HS2 was infused intravenously three times at
5-day intervals. In Group C (n Z 3) only one dose of 10 mL
(0.7 g protein)/kg HS2 was infused intravenously. The piglets were killed on Day 2 and Day 3, respectively, to study
the histopathological changes during the early phase of HS
infusion. In seven piglets in the saline group, 10 mL/kg of NS
was administered intravenously three times at 5-day
intervals.
2. Methods
2.1. Experimental animals
A total of 24 pure-bred castrated piglets, weighing
20e39 kg, aged 1.5e3 months, which were randomly
selected from a certified farm of the national nuclear herd
of the Animal Technology Institute of Taiwan were induced
in this study. They were equivalent to human age of 4
months to 1 year according to the percentage of maturation
and metabolic age chart.29 The Institutional Review Board
of the Animal Technology Institute of Taiwan approved the
study design; the care and handling of piglets followed the
guidelines of the Animal Protection Law, Council of Agriculture.30 The HS group consisted of 17 piglets, aged 1.5
months (n Z 3), 2.5 months (n Z 8), and 3 months (n Z 6),
and the saline (NS) group consisted of seven piglets, aged
1.5 months (n Z 1), 2.5 months (n Z 4), and 3 months
(n Z 2), respectively (Table 2).
2.2. Procedures
Piglets were anesthetized by either azaperone, 8e10 mg/kg
intramuscularly, or thiamylal sodium, 5e8 mg/kg intravenously or in combination. We used two types of HS, namely,
HS1 and HS2 to detect any differences in the induction of
vasculitis in swine model. The HS1 type was negative for
Table 2
2.3. Echocardiography
The 2D echocardiographic examinations were performed
using Hewlett Packard Sonos 100CF, color ultrasound system, using 3.5 mechanical cardiac probe, CA, USA. The
diameters of the left CA (LCA) and right CA (RCA) were
checked and measured at 4e5-day intervals prior to and
after the HS or NS infusion up to 14 days and then weekly
until the autopsy. For the comparative study of the changes
in diameter, measurements of the diameter were taken
5 mm from the orifice of the RCA and LCA. All measurements were taken on the modified parasternal long axis and
Flow chart on experimental study design on immune complex vasculitis in piglets treated with horse serum and saline.
Total no. of piglets studied (n = 24)
Horse serum group (n = 17)
Saline group (n = 7)
Group A
GG-free HS
Three doses (10
mL/kg)
(n = 8)
Results
Rashes ++
Persisted for 4–5 days
(n = 2)
CA changes
40–74% (n = 3)
>75–100 (n = 5)
Group C
DH HS
One dose (10 mL/kg)
(n = 3)
Group B
DH HS
Three doses (10
mL/kg)
(n = 6)
Results
Rashes ++
Persisted for 4–5 days
(n = 3)
CA changes
50–74% (n = 1)
>75–100 (n = 5)
Results
Few rashes
disappeared after
3–4 h (n = 3)
CA changes
12–55% (n = 3)
Saline group
Saline
Three doses (10
mL/kg)
(n = 7)
Results
No rashes
CA changes
12–53%
CA = coronary artery; DH = donor herd; GG = gamma globulin; HS = horse serum;
300
short axis, and on the modified apical four-chamber views
in both right and left lateral positions with the probe at the
opposite side, close to the posterior axillary fold. All piglets
were carefully observed prior to and after infusions until
the day of autopsy. Intraobserver and interobserver measurements were tested.
2.4. Tissue collection and histopathology
Autopsies were performed at 2 days, 3 days, 4 days, 10
days, 14 days, 24 days, 34 days, 41 days, or 60 days after
the first dose of HS and at 10 days, 14 days, 24 days, or 41
days after the first dose of NS infusion, respectively. Gross
appearance and histopathology of the LCA, RCA, left
anterior descending, and left circumflex coronary arteries,
and of the myocardium and systemic arteries, such as the
aorta, and subclavian, iliac, and femoral arteries, were
checked and studied. The liver, kidney, spleen, ear, and
biopsy of skin lesions during infusions were also studied. All
tissue specimens were perfused and put in 10% phosphatebuffered formaldehyde. All materials were serially
sectioned into segments of 2e3 mm thickness and slides
were prepared in hematoxylin and eosin stain. Other special stains, such as Masson’s trichrome and Van Gieson’s
stains, for collagen and ground substances, and Verhoeff’s
stain for internal elastic membrane, were also obtained.
For the convenience of recording histopathological changes
of specimens in the acute, subacute, and healing phases,
the samples were further grouped as follows: 2e4-, 5e13-,
14e24-, and 25e60-day-old samples.
2.5. Statistical analysis
Statistical analysis was carried out using the SPSS version
7.5 (SPSS Inc., Chicago, IL, USA) for Windows. Mean values
were used for comparison. Results are expressed as
mean standard deviation (SD). A comparison between HS
and NS groups was performed using paired and two sample t
tests. A p value < 0.05 was taken to be significant.
3. Results
Within 20e45 minutes following the first, second, or/and
third HS infusion, all piglets developed rashes and showed
immediate severe systemic reactions. Cyanosis, chills,
S. Philip et al
respiratory distress, and convulsions were seen in four
piglets, shock in four piglets, which were revived by
resuscitation; another five piglets showed flushing and mild
chills.
3.1. Erythematous rashes
Rashes appeared during or immediately after the HS infusion at the perineal and perianal regions in 12/14 (86%),
over the legs in 9/14 (64%), over the chest in 8/14 (57%),
over the ears in 8/14 (57%), and on the mouth, lips, and
perioral areas in 8/17 (35%) piglets as seen in Figure 1AeC.
No significant changes were noted between Groups A and B
in terms of rashes, which were less frequent/severe in
Group C that received only one dose of HS. Piglets between
2.5 months and 3 months of age developed more skin rashes
than the younger ones. The rashes faded and disappeared
in 3e6 hours from all piglets, except for two in Group A and
three in Group B that had rashes which persisted for 4e5
days after the second dose of HS. None of the piglets in the
NS group developed the skin rashes or systemic reactions.
3.2. CA changes
Results of 2D echocardiography demonstrated 12e53% increase in the diameter of coronary arteries in the NS group
and in Group C (Table 3). The HS group showed a more
significant dilation of the LCA and RCA (Table 3). The CA
dilation was noted from Day 4 to Day 10, which gradually
resolved to normal size from Day 14 to Day 20; however,
one piglet showed thickening and irregularity of the CA
wall. Of the 14 piglets in Groups A and B, eight (50%)
showed severe dilation (>100%), three (21%) showed moderate dilation (75e99%), and three (21%) showed mild
dilation (54e74%). However, there was no significant dilation of CA in Group C. Groups A and B showed moderate to
severe dilation of CA in 11 (78%) piglets. Changes in the
diameter of the LCA (p < 0.003 in Groups A and B) and RCA
(p < 0.009 in Group A and p < 0.003 in Group B) in the HS
group were highly significant when compared with the saline group. There were no significant differences in the
changes in the CA diameter in Groups A and B (p > 0.6).
Mild pericardial effusions were seen in two piglets in HS2.
The mean 1 SD of intraobserver measurements of the CA
diameter was 0.5 0.05 mm and that of interobserver
Figure 1 Erythematous rashes in test groups after a horse serum (HS) infusion. After 1 hour of HS: (A) rashes on buttocks, back,
and perianal areas in piglet number 7 of test Group A; (B) redness and rashes at the ear, chin, and lips in piglet number 6 of Group B.
After the second dose of HS: (C) rashes on the back, buttocks, and perianal areas in piglet number 4 of Group A.
Horse Serum-Induced Immune Complex Vasculitis in Swine
Table 3
Case
no.
301
Data summary of coronary artery enlargement in horse serum and control saline groups.
Age
(mo)
Weight
(kg)
Left coronary artery
Base (mm)
Horse serum group
Group A (HS1; n Z 8)
1.
2.5
27e36
2.2
2.
2.5
25e30
2.2
3.
2.5
13e24
2.0
4.
1.5
10e12
1.6
5.
3.0
26e34
2.0
6.
2.5
24e34
2.8
7.
2.5
24e35
2.6
8.
1.5
14e35
2.3
Group B (HS2; n Z 6)
1.
2.5
20e24
2.0
2.
2.5
26e28
2.2
3.
3.0
32e39
2.7
4.
2.5
23e26
2.5
5.
2.5
15e24
1.7
6.
2.5
25e32
1.6
Group C (HS2; n Z 3)
1.
2.5
13e14
2.6
2.
2.5
21e22
3.2
3.
2.5
10e11
1.7
Normal saline group (n Z 7)
1.
3.0
24e28
2.8
2.
2.5
16e22
1.8
3.
2.5
22e27
2.8
4.
3.0
27e37
3.0
5.
2.5
20e31
2.8
6.
1.5
09e10
1.3
7.
2.5
26e32
2.4
Right coronary artery
Max (mm)
Enlargement (%)
Base (mm)
Max (mm)
Enlargement (%)
Day at
autopsy
4.4
3.7
2.8
2.4
4.5
6.2
5.8
4.2
100
68
50
50
125
121
123
82
2.0
2.3
2.1
2.0
1.6
2.0
2.0
1.7
3.1
3.0
3.2
3.3
3.5
2.6
3.3
3.0
55
30
52
65
118
30
65
76
D10
D10
D14
D24
D24
D34
D41
D60
3.0
4.0
5.4
5.0
3.4
4.6
50
81
106
100
100
187
1.5
1.4
2.2
2.3
2.0
2.0
2.7
2.3
3.2
3.4
3.6
3.8
80
64
45
47
80
90
D10
D14
D24
D34
D41
D60
2.8
3.4
2.0
7.6
6.2
17
3.1
2.6
1.6
3.4
2.9
1.8
9.6
11.5
12
D02
D03
D04
3.4
2.4
2.9
3.6
3.4
2.0
2.8
33
33
3.5
20
21
53
20
2.0
1.9
2.2
2.4
2.0
1.6
1.8
2.9
2.7
2.6
3.3
2.6
1.8
2.2
45
42
18
37
30
12
22
D10
D14
D24
D24
D34
D41
D60
HS Z horse serum; Max Z maximum.
measurements was 0.6 0.08 mm, indicating that the
measurements were reproducible.
Histopathological examinations of coronary and systemic
arteries of the NS group showed no significant changes
(Figure 2AeC). In the HS group, there were many changes
of varying intensities, such as cellular infiltrates, internal
elastic membrane disruption, mild to severe intimal
proliferative changes, and subintimal change such as
coagulation of the cytoplasm, as well as disorientation,
separation, cytolysis, vacuolization, degranulation,
collagen deposition, and total dissociation and fibrosis of
the smooth muscle cells. The histopathological findings of
vasculitis in the test group from 2e60 days after the first
dose of HS infusion were grouped as follows: 2e4 days
(Figure 2DeH; leucocytic and lymphocytic cellular infiltrates in the myocardium, perivenular, and periarterial
infiltrates in the heart were seen. Cellular infiltrates were
evident in the smooth muscle cells and also around the vasa
vasorum of the aorta and in the distal tubular areas of the
kidney. There were no significant changes in other vessels
and organs); from 5e13 days (Figure 3AeC; intimal thickening, inner smooth muscle cells proliferation, patchy
edematous changes, and early smooth muscle cells (SMC)
disorganization were noted in coronary arteries. There
were a few cellular infiltrates. The iliac artery showed mild
intimal thickening. There were no significant changes in
other vessels and organs); from 14e24 days (Figure 4AeE;
there were intimal and inner SMC proliferations, moderate
to severe disorientation of SMC, edematous separation of
SMC (moth-eaten appearance), subintimal changes, such as
coagulation of the cytoplasm, and disorientation, separation, cytolysis, vacuolization, degranulation, and collagen
deposition in coronary arteries. Intimal proliferation was
also noted in the intramural artery. No significant changes
were observed in other vessels and organs); and from 25e60
days (Figure 4F; patchy areas of fibrosis existed within the
SMC with resolving stages and no further progression of
proliferation of SMC in the tunica media and intima in
piglets that received HS2 infusions on Day 10. No significant
changes were observed in other vessels and organs). A
morphological examination of the heart showed adhesions
and thickening of pericardium (Figure 3D) in two piglets.
Arteritis changes of varying degrees were noted in 79% of
the LCA and left anterior descending artery (LAD), and 64% of
RCA. Arteritis changes of mild degree, such as disruption of
internal elastic membrane, or patchy edematous areas and/or
smooth muscle cell proliferation were also noted in systemic
arteries with varying percentages: femoral artery, 21%;
ascending aorta, 21%; renal artery, 14%; iliac artery, 14%; and
subclavian artery, 14%. In the acute stage, diffuse cellular
302
S. Philip et al
Figure 2 H&E staining of coronary arterial walls of piglets in the saline and HS groups. H&E staining of coronary arterial walls of
piglets in the saline group (AeC) showing normal-looking walls after three doses of NS infusions. (A) Left coronary artery (400) in
piglet number 2 at Day 41 showing normal intima (I), internal elastic membrane (IM), tunica media (M), and adventitia (A). (B) Left
coronary artery (magnification 40) in piglet number 4 at Day 14. (C) Left anterior descending artery (40) in piglet number 1 at
Day 24. (DeH) H&E staining of coronary arterial walls of piglets in the HS group. (D, E; 2e4 days) Perivenular cellular infiltrates of
the coronary vein and vasa vasorum (<) (200) and diffuse cellular infiltrates in the tunica media (400) in piglet number 2 of
Group C at Day 3. (F) Cellular infiltrates in the tunica media (200) of the ascending aorta at Day 2 in piglet number 1 of Group C.
(G) Diffuse cellular infiltrates in the myocardium (400) and (H) in the distal tubular areas of the right kidney (200) at Day 2 in
piglet number 1. H&E Z hematoxylin and eosin stain; HS Z horse serum; NS Z normal saline.
Figure 3 H&E staining of arterial walls of piglets after HS infusions in Groups A and B (5e13 days). (A) Verhoeff’s staining for
internal elastic membrane ( ) and intimal thickening (*) of the right iliac arterial wall (200) of piglet number 2 of Group A; H&E staining
^
of arterial walls of the HS group. (B) Proliferation and thickening (light yellow arrowhead) of the intima and inner SMC of the tunica media of
the left anterior descending artery (40) of piglet number 6 of Group A. (C) Patchy edematous areas (*) in SMC of the LCA (100) in piglet
number 3 of Group B. (D) Pericardial thickening (*) and adhesions in the heart of piglet number 3 of Group B at the right side with normal
heart of the saline group number 6 on the left side. H&E Z hematoxylin and eosin stain; HS Z horse serum; LCA Z left coronary artery.
Horse Serum-Induced Immune Complex Vasculitis in Swine
303
Figure 4 H&E staining of arterial walls of piglets after HS infusions in Groups A and B (14e24 days). (A) Intima and inner SMC
proliferation in RCA (200) of piglet number 4 of Group B. (B) Severe disorientation of SMC in the tunica media of the LCA (400) in
piglet number 5 of Group A. (C) Edematous separation (moth-eaten appearance) of the left descending artery (200) in piglet
number 6 of Group A. (D) Collagen deposition at the subintimal area of the LCA (200) in piglet number 1 of Group A. (E) Intimal
and inner media thickening of the intramural artery (100) (*) in piglet number 5 of Group A; (25e60 days). (F) H&E staining of the
LCA arterial wall (400) showing patchy areas of fibrosis. H&E Z hematoxylin and eosin stain; HS Z horse serum; LCA Z left
coronary artery; RCA Z right coronary artery.
infiltrates, both neutrophils and lymphocytes, were noted in
the myocardium, in the tunica media and peri vasa vasorum
areas of the aorta, and also at the distal tubular areas of kidney, which rapidly resolved within 10 days of HS infusions
(Figure 2EeH). The histopathology of the skin biopsy taken
from the site of rashes showed perivasculitis. Histopathology
of other organs and vessels showed no significant changes
except for some congestive areas and increased lymphoid
follicles in the spleen.
4. Discussion
The pathology of KD has been extensively studied. Although
the pathogenesis of the lesions is not well understood,
immunopathological mechanisms may play an important role
in the genesis of vasculitis in KD (Table 1).3e22 Circulating ICs
in patients with early phase KD have been detected.5,6
Onouchi et al26 reported that HS-induced IC vasculitis in
rabbits showed similar pathophysiology to CAL in KD. Swine
has been used for the study of cardiovascular diseases.27,28
They are large, omnivorous, and convenient for therapeutic
trials.31,32 The heart and vessels are easy to examine with 2D
echocardiography. The CA system of swine is similar to that of
humans, and it is applicable for interventional cardiology,
cardiac xenotransplantation, and even heart lung transplantation.33,34 We used piglets for the experimental study,
weighing 20e28 kg, and which were equivalent to human age
from 4 months to 1 year because more than 80% of the patients with KD are infants and children aged < 5 years.29
304
The IC coronary vasculitis has been elicited by various
agents in mice, guinea pigs, and weanling rabbits, with or
without aneurysm of arterial walls.23e26 The proteins present in the HS can induce acute serum sickness and vasculitis. The pathogenesis of vasculitis postulated is the
fixation of compliments by ICs, activation of compliment
cascade, and the release of biologically active fragments,
notably the anaphylatoxins (C3a and C5a), which increase
vascular permeability and yield chemotactic factors for
polymorphonuclear leukocytes.35 Tissue damage may also
be mediated by free radicals, which are produced by activated neutrophils.
All the piglets receiving HS infusion in our study developed
varying degrees of exanthemas, starting mostly from the
perineal regions, and then spreading to the trunk, legs, ears,
and mouth. The perianal appearance and spread of the
rashes we observed were somewhat similar to those of KD
described by Friter and Lucky.36 Indurative edema and
peeling of the skin were not observed in our study. Multiple
infusions of HS may be better than a single dose because
prolonged and continuous exposure to the sensitizing agent
may lead to excess antigen and the formation of small to
intermediate IC aggregates, which are not easily phagocytosed by the macrophages and circulate widely. These
circulating ICs tend to get deposited in the walls of blood
vessel. In areas where there is low exposure to the sensitizing
antigen, larger IC aggregates are formed, which are easily
phagocytosed by the macrophages.35 There was no significant difference in 2D echocardiographic or histopathological
changes between Groups A and B with two types of HS, but
the initial reactions such as cyanosis, respiratory distress,
convulsions, and shock were more severe with donor herd
HS2, which could be due to the higher concentration of total
proteins in HS2 when compared with HS1.
To the best of our knowledge, 2D echocardiographic
studies on the normal CA diameter and its changes in
weanling piglets have not been reported. We interpreted
the CA dimension as abnormal when the increase was larger
than 9e53% of the baseline diameter, which was observed
in Group C and the control group. Our study showed that CA
dilations started to occur 4e10 days after the first infusion
of HS. All piglets in Group C were sacrificed prior to 5 days
for studying the early changes, and therefore, no changes
in CA were detected by 2D echocardiography. The echocardiography findings of CALs that were observed in the
piglets of this study were similar to those observed in our
clinical KD patients.37
The histopathological changes of coronary arteries that
we induced in piglets by HS infusions were similar to the
acute, subacute, and convalescent phases and to the four
pathological stages related to the duration of illness of
KD.37e39 In all piglets, the changes were most significant in
the tunica media. Similarly, the initial changes in the diameter of coronary arteries in KD occurred in the tunica media
at about 7e9 days after the onset of the disease, as reported
by Naoe.40 The pathological stages of IC vasculitis induced by
large doses of HS infusions in piglets were shortened to 0e4
days in Stage I, 5e14 days in Stage II, 15e24 days in Stage III,
and >25 days in Stage IV when compared with the pathological staging in KD as 0e11 days in Stage I, 12e25 days in
Stage II, and >30e40 days in Stages III and IV. Vasculitis
changes in piglets were resolved from Day 14 onward.
S. Philip et al
Cellular infiltrations such as mononuclear cells were fewer in
the 5e13-day autopsy group. The time span may differ in the
human pathology as the swine model would have a shorter
course for each stage of vasculitis and the presence of
cellular infiltrates and their composition would vary
accordingly.29 Hence, the presence of cellular infiltrations in
each stage may change accordingly. The courses and the
severity of vasculitis in piglets could be different from the
natural course of KD, as it was induced by the large multiple
doses of HS. Arteritis changes of varying degrees in piglets
were noted more in the LCA than in the RCA, which were
similar to the study in KD by Takahashi.41
Type III hypersensitivity reaction, induced by antigeneantibody complexes, activates a variety of serum mediators, mainly the compliment system. Both ICs and platelets
may have some role to play in the pathogenesis of vasculitis.14
ICs were identified in the autopsy specimens of KD suggesting
that IC might have played a role in producing the coronary
arteries changes in KD patients.14 ICs were also identified in
the circulation of the experimental rabbit models with serum
sickness.26 This study documented that systemic type III hypersensitivity reactions and histopathological changes in
various stages produced in the HS group were similar to the
rashes and histopathological changes of vasculitis in KD; and
we suggest, therefore, that a similar mechanism may be
involved in the pathogenesis of coronary arteritis in KD.
Type III hypersensitivity reaction in serum sickness is a
prototype of IC vasculitis. Induction and reproducibility of
IC vasculitis with two different types of HS were possible in
piglets weighing 9e39 kg. The rashes and the findings of
CALs detected by 2D echocardiography, and histopathological studies in the acute to healed phases of vasculitis
showed close similarities to KD. We postulate that ICmediated mechanisms may play a role in the pathogenesis
of CALs in KD and that swine may serve as an experimental
model for various therapeutic trials.
Conflicts of interest
The authors declare that they have no conflicts of interest.
Acknowledgments
The authors wish to thank Dr J.H. Lin and Dr M.T Chiou for
pathology discussions, Mrs P.H. Lin and Ms Lilly Ho for
expert technical assistance and autopsy, Dr C.C. Hsu (for
statistical advice), and Dr T.S. Yang (for further thoughts
and advice).
References
1. Kawasaki T, Kosaki F, Okawa S, Shigematsu I, Yanagawa H. A
new infantile acute febrile mucocutaneous lymph node syndrome (MLNS) prevailing in Japan. Pediatrics 1974;54:271e6.
2. Naoe S, Shibuya K, Takahashi K, Wakayama M, Masuda H,
Tanaka M. Pathological observations concerning the cardiovascular lesions in Kawasaki disease. Cardiol Young 1991;1:
212e20.
3. Fossard C, Thompson RA. Mucocutaneous lymph-node syndrome (Kawasaki disease): probable soluble-complex disorder.
Br Med J 1977;1:883.
Horse Serum-Induced Immune Complex Vasculitis in Swine
305
4. Sawa F. Circulating immune complexes in MCLS. Acta Paediatr
Jpn 1979;83:493e8.
5. Weindling AM, Levinsky RJ, Marshall WC, Hood J. Circulating
immune complexes in mucocutaneous lymph-node syndrome
(Kawasaki disease). Arch Dis Child 1979;54:241e2.
6. Eluthesen K, Marchette N, Melish M, et al. Circulating immune
complexes in Kawasaki’s disease: detection of C1q binding
assay. Presented at 21st inter science conference on Antimicrobial Agents and chemotherapy. November 4 to 6, 1981.
7. Furuse A, Matsuda I. Circulating immune complex in the mucocutaneous lymph node syndrome. Eur J Pediatr 1983;141:50e1.
8. Yanase Y, Kawasaki T, Yoshinoya S, Aikawa T, Hashimoto Y,
Mitamura T, et al. A study of immune complexes in Kawasaki
disease. Arerugi 1984;33:59e65 [Article in Japanese].
9. Miyata K, Kawakami K, Onimaru T, Baba Y, Ono S,
Hokonohara M, et al. Circulating immune complexes and
granulocytes chemotaxis in Kawasaki disease. Jpn Circ J 1984;
48:1350e3.
10. Takiguchi M, Tamura T, Goto M, Kusakawa S, Milgrom F, Kano K.
Immunological studies on Kawasaki disease. I. Appearance of
Hanganutziu-Deicher antibodies. Clin Exp Immunol 1984;56:
345e52.
11. Mason WH, Jordan SC, Sakai R, Takahashi M, Bernstein B.
Circulating immune complexes in Kawasaki syndrome. Pediatr
Infect Dis 1985;4:48e51.
12. Ono S, Onimaru T, Kawakami K, Hokonohara M, Miyata K.
Impaired granulocyte chemotaxis and increased circulating
immune complexes in Kawasaki disease. J Pediatr 1985;106:
567e70.
13. Levin M, Holland PC, Nokes TJ, Novelli V, Mola M, Levinsky RJ,
et al. Platelet immune complex interaction in pathogenesis of
Kawasaki disease and childhood polyarteritis. Br Med J (Clin
Res Ed) 1985;290:1456e60.
14. Pachman LM, Herold BC, Davis AT, Hang LM, Schaller JG,
Beckwith B, et al. Immune complexes in Kawasaki syndrome: a
review. Prog Clin Biol Res 1987;250:193e207.
15. Levin M, Holland PC, Novelli V. Platelet immune complex
interaction in the pathogenesis of Kawasaki disease. Prog Clin
Biol Res 1987;250:227e37.
16. Lin CY, Hwang B. Serial immunologic studies in patients with
mucocutaneous lymph node syndrome (Kawasaki disease). Ann
Allergy 1987;59:291e7.
17. Fujimoto T, Kato H, Inoue O, Tomita S, Koga Y. Immune complex study of biopsy specimens from Kawasaki disease patients. Prog Clin Biol Res 1987;250:209e17.
18. Ohshio G, Furukawa F, Khine M, Yoshioka H, Kudo H,
Hamashima Y. High levels of IgA-containing circulating immune
complex and secretory IgA in Kawasaki disease. Microbiol
Immunol 1987;31:891e8.
19. Salcedo JR, Greenberg L, Kapur S. Renal histology of mucocutaneous lymph node syndrome (Kawasaki disease). Clin
Nephrol 1988;29:47e51.
20. Salo E, Kekoma
¨ki R, Pelkonen P, Ruuskanen O, Viander M,
Wagner O. Kawasaki disease: monitoring of circulating immune
complexes. Eur J Pediatr 1988;147:377e80.
21. Li CR, Yang XQ, Shen J, Li YB, Jiang LP. Immunoglobulin G subclasses in serum and circulating immune complexes in patients
with Kawasaki syndrome. Pediatr Infect Dis J 1990;9:544e7.
22. Koike R. The effect of immunoglobulin on immune complexes
in patients with Kawasaki disease (MCLS). Acta Paediatr Jpn
1991;33:300e9.
23. Murata H. Experimental Candida-induced arteritis in mice.
Relation to arteritis in the mucocutaneous lymph node syndrome. Microbiol Immunol 1979;23:825e31.
24. Lehman TJ, Walker SM, Mahnovski V, McCurdy D. Coronary
arteritis in mice following the systemic injection of group B
Lactobacillus casei cell walls in aqueous suspension. Arthritis
Rheum 1985;28:652e9.
25. Rich AR, Gregory JE. The experimental demonstration that
periarteritis nodosa is manifestation of hypersensitivity. Johns
Hopkins Hosp 1943;72:65e88.
26. Onouchi Z, Ikuta K, Nagamatsu K, Tamiya H, Sakakibara Y,
Ando M. Coronary artery aneurysms develop in weanling rabbits with serum sickness but not in mature rabbits. An experimental model for Kawasaki disease in humans. Angiology
1995;46:679e87.
27. Brown DR, Terris JM. Swine in physiological and pathophysiological research. In: Tumbleson ME, Schook LB, editors. Advances in swine biomedical research, Vol. 1. New York: Plenum
Press; 1995. pp. 5e15.
28. Philip S, Lee WC, Liu SK, Wu MH, Lue HC. A swine model of
horse serum-induced coronary vasculitis: an implication for
Kawasaki disease. Pediatr Res 2004;55:211e9.
29. Kirkwood JK, Webster AJF. Energy budget strategies for growth
in mammals and birds. Anim Prod 1984;38:147e55.
30. Animal Protection Law. Council of Agriculture Executive Yuan,
Taiwan, amended, 2001. Chapters IeIII. Taipei, Taiwan: HuaZong, Yi-Tzi. Enforcement Rules of Animal Protection; 1998.
Available at: http://www.coa.gov.tw/coa/eng/index.html.
Accessed April 19, 2013.
31. Lee KT. Swine as animal models in cardiovascular research. In:
Tumbleson ME, editor. Swine in biomedical research, Vol. 3.
New York: Plenum Press; 1986. pp. 1481e96.
32. Hall TS, Rosengrad BR, Stone CD, Baumgartner WA, Reitz BA.
Pig models for heart-lung transplantation research. Proceedings of the Second International Symposium on Pig Model
for Biomedical Research; 1990. pp. 55e65.
33. Sachs DH, Leight G, Cone J, Schwarz S, Stuart L,
Rosenberg S. Transplantation in miniature swine. I. Fixation
of the major histocompatibility complex. Transplantation
1976;22:559e67.
34. Allan JS, Rose GA, Choo JK, Arn JS, Vesga L, Mawulawde K,
et al. Morphometric analyses to predict appropriate donor size
for swine-to-human cardiac xenotransplantation. Transplant
Proc 1999;31:975e7.
35. Janeway CA, Travers P, Walport M, Capra JD. Immune biology:
immune system in health and disease. 4th ed. New York:
Garland Publishing; 1999. pp. 479e81.
36. Friter BS, Lucky AW. The perineal eruption of Kawasaki syndrome. Arch Dermatol 1988;124:1805e10.
37. Yang CC, Lue HC, Wang JK, Wu MH, Wu YN. A detection and
follow up study of coronary arterial lesions in Kawasaki disease
by two-dimensional echocardiography. Acta Cardiol Sin 1990;
6:262e75.
38. Tanaka N, Naoe S, Masuda H, Ueno T. Pathological study of
sequelae of Kawasaki disease (MCLS). With special reference to
the heart and coronary arterial lesions. Acta Pathol Jpn 1986;
36:1513e27.
39. Suzuki A, Miyagawa-Tomita S, Komatsu K, Nishikawa T,
Sakomura Y, Horie T, et al. Active remodeling of the coronary
arterial lesions in the late phase of Kawasaki disease: immunohistochemical study. Circulation 2000;101:2935e41.
40. Naoe S. Pathology of coronary aneurysms in the young.
Abstract (CS12) presented at the 10th Asian Congress of
Pediatrics. Taipei: The Chinese Taipei Pediatric Association;
2000.
41. Takahashi M. The endothelium in Kawasaki disease: the next
frontier. J Pediatr 1998;133:1771e9.
Pediatrics and Neonatology (2014) 55, 306e311
Available online at www.sciencedirect.com
ScienceDirect
journal homepage: http://www.pediatr-neonatol.com
ORIGINAL ARTICLE
Cyclic Pamidronate Infusion for Neonatalonset Osteogenesis Imperfecta
Chia-Hsuan Lin a, Yin-Hsiu Chien a,b, Shinn-Forng Peng c,
Wen-Yu Tsai a, Yi-Ching Tung a, Cheng-Ting Lee a,
Chun-Ching Chien b, Wuh-Liang Hwu a,b, Ni-Chung Lee a,b,*
a
Department of Pediatrics, National Taiwan University Hospital and National Taiwan University
College of Medicine, Taiwan
b
Department of Medical Genetics, National Taiwan University Hospital and National Taiwan University
College of Medicine, Taiwan
c
Department of Radiology, National Taiwan University Hospital and National Taiwan University
College of Medicine, Taiwan
Received Aug 15, 2013; received in revised form Dec 9, 2013; accepted Dec 13, 2013
Available online 31 January 2014
Key Words
bisphosphonate;
bone mineral density;
neonate;
osteogenesis
imperfecta
Background: Patients with severe osteogenesis imperfecta (OI; MIM number 259420) suffer
from low bone mass, fractures, and bone pain since birth, and have poor prognosis. This study
assessed the outcome of patients with severe OI who were treated with cyclic pamidronate
prior to the age of 1 year.
Methods: The six patients, who had bone fractures either in utero or in their 1st month of life,
were treated with cyclic pamidronate from a mean age of 2.8 months.
Results: All the patients tolerated the infusion, except for having transient hypocalcemia at
the first infusion. Decreases in irritability and improvements in feeding were observed 2e3
months after the first infusion. All patients showed a rapid increase in bone mineral density
over the first 2 years. Fractures occurred at a rate of 0.6/year. At a mean age of 6.4 years, five
patients with no interruption in treatment had normal ambulatory function, but they were
short in height.
Conclusion: Patients with neonatal OI can have a favorable outcome when treated with cyclic
pamidronate infusions early in life.
Copyright ª 2014, Taiwan Pediatric Association. Published by Elsevier Taiwan LLC. All rights
reserved.
* Corresponding author. Department of Pediatrics and Medical Genetics, Room 19005, 19F, Children’s Hospital Building, National Taiwan
University Hospital, 8 Chung-Shan South Road, Taipei 10041, Taiwan. Tel.: þ886 2 23123456x71938; fax: þ886 2 2331 4518.
E-mail address: [email protected] (N.-C. Lee).
1875-9572/$36 Copyright ª 2014, Taiwan Pediatric Association. Published by Elsevier Taiwan LLC. All rights reserved.
http://dx.doi.org/10.1016/j.pedneo.2013.12.001
Neonatal treatment of pamidronate in osteogenesis imperfecta
1. Introduction
Osteogenesis imperfecta (OI; MIM number 259420) is characterized by decreased bone mass, increased bone
fragility, bone pain, and frequent fractures because of
defects in the collagen genes.1 Clinical manifestations of
patients may vary from nondeforming OI with blue sclera to
lethal perinatal complications.1,2 At present, OI is classified
into eight different types and 10 disease-causing genes
have been shown to cause OI.3 The incidence rate is six to
seven/100,000 when including all types of OI.1 Most patients have an autosomal dominant inheritance with 100%
penetrance but variable clinical expression.1 In general,
patients with earlier onset of fracture have poorer
prognosis.
OI has been treated with intravenous administration of
bisphosphonates since 19874; pamidronate has been the
first and still most commonly used drug.5 The bisphosphonate compounds are potent inhibitors of bone resorption,
and histomorphometric studies of bone from OI patients
suggest that this disorder is associated with an increase in
osteoclastic activity and a reduction in the formation of
new bone. The therapy has been shown to decrease bone
pain, increase bone mineral density (BMD) and vertebral
body size, accelerate growth rate, and improve mobility in
patients with OI.6e12
Because bisphosphonate treatment does not correct the
defects in collagen fibers, its effect on severe patients who
have fractures at or shortly after birth has not been
confirmed. Without proper treatment, these patients either
die early in life or survive with severe bone fractures and
extremely short stature. Munns et al13 treated 29 patients
starting from a median age of 6 months. After 3 years, the
treated patients showed improved bone strength and better gross motor function, but there was no comparison with
normal controls in either mobility or fracture. They were
also concerned about the suppression of bone turnover by
pamidronate treatment. Antoniazzi et al14 treated 10 patients diagnosed at a mean age of 33 days, starting treatment either immediately or after 6 months. One year later,
those treated at birth showed improvements in biochemical
profiles and in vertebral body morphologies. Astro
¨m et al15
gave monthly infusions of pamidronate to 11 patients at a
median age of 3.6 months. At a median age of 4.8 years, all
children walked but all required femoral intramedullary
rods for fractures, and five needed tibial rodding for
extreme curvatures that prevented functional standing and
walking.
We began treating newborn infants with OI with pamidronate since 2001.16 In this study, we report the outcome of six
patients treated with pamidronate. The mean age of the
patients while starting the treatment was 2.8 months. Five
patients with no interruption in treatment obtained normal
motor, while the other one walks with walker.
2. Methods
2.1. Patients
From January 2001 to March 2013, there were six patients
diagnosed with OI who were either born with fractures or
307
had fractures within the 1st month of life and received
cyclic pamidronate infusions in our hospital. The diagnosis
of OI depended on clinical features including bone fractures, blue sclera, family history, radiologic findings, and
BMD. Collagen Type I [COL1A1 (MIM number 120150) and
COL1A2 genes (MIM number 120160)] mutation analyses
were performed in six patients, and COL1A1 and COL1A2
mutations were found in five patients (Table 1). The study
protocol was approved by the Institutional Review Board of
our hospital (No. 201307031RINC).
2.2. Treatment protocol
The dosage and frequency of pamidronate infusion depended on age. In younger populations, the dosage was lower,
but the frequency was higher than that in older patients.
The dosing and frequency schedule was as follows: 0.5 mg/
kg every 2 months prior to 1 year of age; 1 mg/kg every 2
months between 1 year and 2 years of age; 1.5 mg/kg every
3 months between 2 years and 4 years of age; and 2 mg/kg
every 3e4 months after 4 years of age. The drug was
diluted in a 0.9% saline solution and infused over 4 hours.
Oral calcium carbonate (125 mg two times a day) and
vitamin D (cholecalciferol 400 international units daily)
were also supplied. Patient 5 was initially treated at
another hospital by a monthly infusion protocol prior to
being transferred to our hospital at 3 years of age; at that
time he still could not sit.
2.3. Data collection
Our study retrospectively collected the anthropometric,
radiological, and biochemical results from patient medical
records. We also collected data related to their symptomatic presentations, sex, treatment starting dates, lumbar
spine BMD results, functional assessments, biochemical
profiles, bone fracture rates, and disease complication
outcomes. The standard score (standard deviation scores or
SDS) for height was calculated based on the World Health
Organization international growth references if the patient
was aged <5 years; the Taiwanese growth curve was used if
the patient was aged >5 years.17,18 Lumbar spine (L2eL4)
BMD examinations were performed in the anteroposterior
direction with dual-energy X-ray absorptiometry (Norland
XR-26 Mark II; Norland, Fort Atkinson, WI, USA). The BMD
results were transformed into age-specific Z scores based
on reference data from Zanchetta et al19 and from normal
young Chinese in the Taipei region.20,21
3. Results
3.1. Symptom relief after pamidronate infusion
Cyclic pamidronate infusion was started prior to the age 1
year in all six patients. These patients were either born
with fractures or had fractures within the 1st month of life,
and they were all irritable and poorly fed prior to treatment. Symptom relief was seen 1 month after the first
infusion. Their feeding normalized after the second
pamidronate infusion; specifically, 2e3 months after the
308
Table 1
C.-H. Lin et al
Growth, fractures, and ambulation status of patients after cyclic pamidronate infusions.
Patient no.
1
2
3
4
5
6
Sex
COL1A1 mutation
F
c.2236-47_2273
del 85
In utero
0.5
2.37
12.2
4.12
12.13
1.23
M
c.1273G>A
F
c.2497G>A
F
*c.901G>A
M
Not found
F
c.769G>A
In utero
0.7
5.39
0.8
7.13
0.78
0
In utero
1.2
8.26
1
6.61
0.9
1.27
1 mo
4.3
5.12
5.4
4.53
5.06
0.82
3d
6.5
2.46
9.3
4.22
8.71
0.24
At birth
3.3
0.8
9.5
4.53
9.23
0
Rolling over
Sit with
support
Walk and
run fast
Walk with
walker
Walk and run,
leg length
discrepancy
Time of first fracture
Age at start (mo)
Starting height (SDS)
Current age (y)
Latest height (SDS)
Years of treatment
Fracture rate
(per year)
Mobility
Walk and run,
leg length
discrepancy
Mean SD
2.8
3.8
6.4
5.2
6.1
0.6
2.4
3.1
4.8
1.3
4.7
0.6
F Z female; M Z male; SD Z standard deviation; SDS Z standard deviation scores. *COL1A2 mutation.
first infusion. At that time, the babies could be handled
easier during baths or diaper changes, without triggering
irritable crying.
Their average lumbar spine (L2eL4) BMD increased quickly
after pamidronate infusion (Figure 1). The BMD was usually
stabilized 2e3 years after the start of treatment.
3.2. Hypocalcemia after cyclical pamidronate
treatment
3.4. Changes in radiological features
Hypocalcemia after pamidronate infusion was observed in
all three patients receiving pamidronate during their 1st
month of life, although oral calcium and vitamin D supplies
were prescribed. Patient 1 suffered from a seizure attack 5
days after the first infusion when she was discharged from
the hospital. Her blood calcium level was 1.225 mmol/L
(normal range: 2.02e2.60 mmol/L) at the time of seizure
occurrence.16 After this experience, we monitored blood
calcium levels more closely for 3e5 days in these patients,
especially after the first infusion, and then gave aggressive
intravenous calcium supplementation (0.54 mEq/kg calcium chloride by slow intravenous push every 8 hours) when
we discovered blood calcium levels lower than 2 mmol/L.
Patients 2 and 3 were noted to have respective blood calcium levels of 1.97 mmol/L and 1.72 mmol/L 5 days after
the first infusion. As a result of timely management of their
hypocalcemia, no seizure occurred in these patients. As a
whole, the calcium levels of the six patients were
2.34 0.28 mmol/L prior to the first infusion and
2.19 0.29 mmol/L 3 days after the infusion, and the
phosphate levels were 5.44 0.27 mg/dL prior to the first
infusion and 4.50 0.66 mg/dL 3 days after the infusion
(both were statistically nonsignificant). Fever, bone pain
and common adverse effects of pamidronate infusion were
not seen in these patients.
All patients showed significant improvements in skeletal
radiological features. Patient 1 was treated and followed
for more than 12 years. Her four extremities were curved
when she was born. Radiological studies of her lower extremities performed shortly after birth revealed poor
mineralization of both the long bones and the pelvis
(Figure 2A). The cupping of the metaphysis was prominent,
and her femur and tibial bones were curved. Her skull
bones, facial bones, and vertebrae were also poorly
mineralized (Figure 2B). Radiological studies performed at
3 years of age revealed much improved bone mineralization
over the long bones and vertebrae (Figure 2C and D). Most
of her long bones were straight, and her femur bones were
only slightly curved. Radiological studies at 12 years of age
revealed straight and normally developed bones (Figure 2E
and F). However, metal plates were placed after fractures
on both femurs to increase bone strength.
3.3. Growth after cyclical pamidronate infusion
The mean age of the patients at the start of treatment was
0.2 years (median: 0.2 years), and the mean height SDS at
the start of treatment was 3.89 (median: 3.14). Their
mean height SDS after 6.1 years (median: 6.9 years) of
cyclical pamidronate treatment was 5.2 (median: 4.5).
Figure 1 Changes in bone mineral density (BMD) in the six
patients from Group 1 after they received cyclic pamidronate
infusion. BMD is expressed as a Z score. Patients 2 and 3 had
only one BMD measurement (arrows).
Neonatal treatment of pamidronate in osteogenesis imperfecta
Patient 2 had fractures prior to 1 month of age. Radiological studies at that time revealed curved femur bones
with prominent fractures (arrow) and metaphyseal flaring
(Figure 3A). His vertebral bone development was poor. A
repeat study at 10 months of age revealed that his femurs
were straighter than previously, although both tibial bones
were slightly curved (Figure 3B). Vertebral development
was also better. Radiological studies of Patient 3 at 5 weeks
of age revealed curved bones and multiple fractures (arrows) over all extremities (Figure 3A). At 10 months of age,
all fractures healed and their bone morphologies were also
much improved (Figure 3D). The radiological features of
Patient 4 were less dramatic than those seen in Patients
1e3; however, curving, fracture (arrow), and metaphyseal
flaring were all present (Figure 3E). Follow-up pictures at
both 3 years and 5 years of age (Figure 3F) showed straight
bones except for the right femur. The tibial bones still
experienced fractures, and one intramedullary rod was
placed there to strengthen the right tibial bone. In
Figure 3F, parallel high-density lines were seen in the distal
femoral and tibial metaphysis (white arrows). Each line
represents a course of pamidronate infusion; these lines are
also referred to as zebra lines. Patient 5 had more bonerelated complications because of inappropriate treatment
prior to 3 years of age, and sclerotic changes were observed
over the metaphysis at 9 years of age. Patient 6 had a
fracture over the right arm at birth.
309
3.5. Fracture rates and ambulation status
The mean annual fracture rate of the patients was 0.6 0.5
per year (Table 1). Internal fixation after fracture by either
a rod or plate was a common practice in these patients. The
fractures usually occurred after minor injuries; although
these fractures bothered the patients, healing was normal
and internal fixation increased their bone strength. Their
ambulation status was close to normal. All the four patients
who were older than 1 year of age walked; three of them
could run. Patient 5 was treated by a monthly cyclic infusion protocol of pamidronate until 3 years of age, which
was withdrawn 9 months prior to when we treated him. Leg
length discrepancy and scoliosis occurred in Patients 1
and 6.
4. Discussion
4.1. Pamidronate infusion improves the outcome of
severe OI
Patients who have curved bones at birth or have fractures
during the 1st month of life are usually classified as Type III
OI. These patients can have severe complications and a
limited life span. In this study, cyclic pamidronate infusions
were started at 1e2 months of age, earlier than most other
Figure 2 Radiological studies of Patient 1. (A) Curved leg bones that are poorly mineralized at 12 days of age. (B) Poorly
mineralized skull bones at 12 days of age. (C and D) Improvements in bone mineralization over the long bones and vertebrae are
seen at 3 years of age. (E) Straight long bones with metal plates on both femurs at 12 years of age. (F) Better mineralization of the
vertebral bodies at 12 years of age.
310
C.-H. Lin et al
Figure 3 Radiological studies of Patients 2e4. (A) Patient 2 showed curved femur bones with metaphyseal flaring and fractures
(arrow) at 1 month of age, and (B) improvement at 10 months of age. (C) Patient 3 showed curved bones and fractures (arrow) at 5
weeks of age, and (D) improvements at 10 months of age. (E) Patient 4 showed curving, fracture (arrow), and metaphyseal flaring at
10 months of age, and (F) straight tibial bones but with fractures at both mid shafts at 5 years of age. Zebra lines were prominent on
both the distal femur and tibia (white arrows).
studies; the oldest patients have been treated for more
than 12 years. We saw that those patients functioned normally. Bisphosphonate infusion in neonates carries a higher
risk of hypocalcemia, but can be managed by calcium
supplementation. Although patients still had fractures
(mean 0.6 fractures/year), the fractures did not markedly
disturb their life. As a whole, early cyclic pamidronate infusions changed the natural history of severe OI.
4.2. Mechanisms and regimens of cyclic
bisphosphonate infusion
The OI pathological process results from defective collagen
fibers that cause excessive bone material destruction and
absorption.22 Bisphosphonates decrease the bone material
absorption and increase the bone density in OI patients.13,23
The 2e4-month cyclic infusion protocol design is based on
the long biological half-life of pamidronate and allows bone
growth between infusions.5,7,13,14 Monthly infusion of
bisphosphonates6,15 may not allow enough time for bone
remodeling, which could be one of the reasons of bowling
tibias in the study by Astro
¨m et al.15 However, in view of
the faster bone metabolism in infants, we and others13 used
a 2-month infusion regimen for infants.
Most studies used a pamidronate dosage of 4e9 mg/kg/
year,5,7,13,24 although a dose up to 12 mg/kg/year or
monthly injections have also been described9,25; however,
Senthilnathan et al reported that infants receiving 12 as
opposed to 6 mg/kg/year pamidronate had increased spine
bone density. We used a smaller total annual dose (3 mg/
kg/day) than in other protocols for patients younger than 1
year, but still obtained a result similar to that in other
studies. Because the three consecutive daily infusions
treatment protocol is not convenient for patients, we
pooled the three daily doses into a single infusion. We
found that patients tolerated this regimen very well, and
most infusions could be executed in the day clinics.
Neonatal treatment of pamidronate in osteogenesis imperfecta
4.3. Body height in neonatally treated OI patients
Untreated patients with severe OI are usually very short,
due to a combination of poor bone growth, fractures and
curving of long bones, and vertebral body compression
fractures. Zeitlin et al8 have shown that pamidronate
therapy increases the height SDS in Type III OI. Unfortunately, in this study, height SDS decreased during the
treatment periods. However, the four older patients were
all with a height SDS of around 4, better than the patients
described in the study by Zeitlin et al.8 Our patients had
straight legs and there was no vertebral body collapse. The
dense zebra lines were evenly spaced and might increase
bone strength. Nevertheless, we will need to further
improve treatment, and a final solution will depend on new
treatments specific to the disease etiology.
4.4. Limitations of the study
Our study has several limitations, such as lack of controls,
small number of cases, and the presence of clinical heterogeneity, for example, some patients had in utero fractures
whereas others had postnatal fractures. Therefore, the results from this study need to be compared with historical
cohorts. Subsequent studies may require further subgrouping
of disease severity to elucidate the treatment effect.
5. Conclusion
In conclusion, we demonstrated that patients with severe
OI can have a favorable outcome when treated with cyclic
pamidronate infusions in early infancy.
Conflicts of interest
The authors declare no conflicts of interest.
Acknowledgments
The authors thank the patients and their families for their
cooperation.
References
1. Steiner RD, Adsit J, Basel D. COL1A1/2-Related Osteogenesis
Imperfecta. In: Pagon RA, Adam MP, Bird TD, Dolan CR, Fong
CT, Stephens K, editors. GeneReviews [Internet]. Seattle
(WA): University of Washington, Seattle; 1993e2013.
2. Byers PH. Disorders of collagen biosynthesis and structure. In:
Scriver CR, Beaudet AL, Sly WS, Valle D, Vogelstein B, editors.
The online metabolic and molecular bases of inherited disease
(OMMBID). New York: McGraw-Hill; 2013. Available at: http://
www.ommbid.com/. Chap 205. [Accessed June 15, 2013].
3. van Dijk FS, Byers PH, Dalgleish R, Malfait F, Maugeri A,
Rohrbach M, et al. EMQN best practice guidelines for the laboratory diagnosis of osteogenesis imperfecta. Eur J Hum Genet
2012;20:11e9.
4. Devogelaer JP, Malghem J, Maldague B, Nagant de
Deuxchaisnes C. Radiological manifestations of bisphosphonate
treatment with APD in a child suffering from osteogenesis
imperfecta. Skeletal Radiol 1987;16:360e3.
311
5. Glorieux FH, Bishop NJ, Plotkin H, Chabot G, Lanoue G,
Travers R. Cyclic administration of pamidronate in children
with severe osteogenesis imperfecta. N Engl J Med 1998;339:
947e52.
6. Astro
¨m E, So
¨derha
¨ll S. Beneficial effect of long term intravenous bisphosphonate treatment of osteogenesis imperfecta.
Arch Dis Child 2002;86:356e64.
7. Falk MJ, Heeger S, Lynch KA, DeCaro KR, Bohach D, Gibson KS,
et al. Intravenous bisphosphonate therapy in children with
osteogenesis imperfecta. Pediatrics 2003;111:573e8.
8. Zeitlin L, Rauch F, Plotkin H, Glorieux FH. Height and weight
development during four years of therapy with cyclical intravenous pamidronate in children and adolescents with osteogenesis imperfecta types I, III, and IV. Pediatrics 2003;111:
1030e6.
9. Lin HY, Lin SP, Chuang CK, Chen MR, Chang CY. Intravenous
pamidronate therapy in Taiwanese patients with osteogenesis
imperfecta. Pediatr Neonatol 2008;49:161e5.
10. Gatti D, Viapiana O, Lippolis I, Braga V, Prizzi R, Rossini M,
et al. Intravenous bisphosphonate therapy increases radial
width in adults with osteogenesis imperfecta. J Bone Miner Res
2005;20:1323e6.
11. Glorieux FH. Experience with bisphosphonates in osteogenesis
imperfecta. Pediatrics 2007;119:S163e5.
12. Fujiwara I, Ogawa E, Igarashi Y, Ohba M, Asanuma A. Intravenous pamidronate treatment in osteogenesis imperfecta. Eur J
Pediatr 1998;157:261e2.
13. Munns CF, Rauch F, Travers R, Glorieux FH. Effects of intravenous pamidronate treatment in infants with osteogenesis
imperfecta: clinical and histomorphometric outcome. J Bone
Miner Res 2005;20:1235e43.
14. Antoniazzi F, Zamboni G, Lauriola S, Donadi L, Adami S, Tato
` L.
Early bisphosphonate treatment in infants with severe osteogenesis imperfecta. J Pediatr 2006;149:174e9.
15. Astro
¨m E, Jorulf H, So
¨derha
¨ll S. Intravenous pamidronate
treatment of infants with severe osteogenesis imperfecta.
Arch Dis Child 2007;92:332e8.
16. Chien YH, Chu SY, Hsu CC, Hwu WL. Pamidronate treatment of
severe osteogenesis imperfecta in a newborn infant. J Inherit
Metab Dis 2002;25:593e5.
17. WHO Multicentre Growth Reference Study Group. WHO Child
Growth Standards based on length/height, weight and age.
Acta Paediatr Suppl 2006;450:76e85.
18. Chen W, Chang MH. New growth charts for Taiwanese children
and adolescents based on World Health Organization standards
and health-related physical fitness. Pediatr Neonatol 2010;51:
69e79.
19. Zanchetta JR, Plotkin H, Alvarez Filgueira ML. Bone mass in
children: normative values for the 2e20-year-old population.
Bone 1995;16:393Se9S.
20. Tsai KS, Cheng WC, Chen CK, Sanchez TV, Su CT, Chieng PU,
et al. Effect of bone area on spine density in Chinese men and
women in Taiwan. Bone 1997;21:547e51.
21. Gallo S, Vanstone CA, Weiler HA. Normative data for bone mass
in healthy term infants from birth to 1 year of age. J Osteoporos 2012;2012:672403.
22. Braga V, Gatti D, Rossini M, Colapietro F, Battaglia E,
Viapiana O, et al. Bone turnover markers in patients with
osteogenesis imperfecta. Bone 2004;34:1013e6.
23. Russell RG. Bisphosphonates: mode of action and pharmacology. Pediatrics 2007;119:S150e62.
24. Plotkin H, Rauch F, Bishop NJ, Montpetit K, Ruck-Gibis J,
Travers R, et al. Pamidronate treatment of severe osteogenesis
imperfecta in children under 3 years of age. J Clin Endocrinol
Metab 2000;85:1846e50.
25. Senthilnathan S, Walker E, Bishop NJ. Two doses of pamidronate in infants with osteogenesis imperfecta. Arch Dis Child
2008;93:398e400.
Pediatrics and Neonatology (2014) 55, 312e315
Available online at www.sciencedirect.com
journal homepage: http://www.pediatr-neonatol.com
CASE REPORT
Normal Uricemia in LescheNyhan Syndrome
and the Association with Pulmonary
Embolism in a Young ChilddA Case Report
and Literature Review
Jeng-Dau Tsai a,b, Shan-Ming Chen a,b, Chien-Heng Lin c,
Min-Sho Ku a,b, Teng-Fu Tsao d, Ji-Nan Sheu a,b,*
a
Department of Pediatrics, Chung Shan Medical University Hospital, Taichung, Taiwan
Institute of Medicine, Chung Shan Medical University, Taichung, Taiwan
c
Department of Pediatrics, Jen-Ai Hospital, Taichung, Taiwan
d
Department of Medical Imaging, Chung Shan Medical University Hospital, Taichung, Taiwan
b
Received Jan 11, 2012; received in revised form Mar 28, 2012; accepted Nov 22, 2012
Available online 4 February 2013
Key Words
HPRT;
hyperuricemia;
LescheNyhan
syndrome;
pulmonary embolism
Deficiency of hypoxanthine phosphoribosyltransferase activity is a rare inborn error of purine
metabolism with subsequent uric acid overproduction and neurologic presentations. The diagnosis of LescheNyhan syndrome (LNS) is frequently delayed until self-mutilation becomes evident. We report the case of a boy aged 1 year and 10 months who was diagnosed with profound
global developmental delay, persistent chorea, and compulsive self-mutilation since the age of
1 year. Serial serum uric acid levels showed normal uric acid level, and the spot urine uric acid/
creatinine ratio was >2. The hypoxanthine phosphoribosyltransferase cDNA showed the deletion of exon 6, and the boy was subsequently diagnosed to have LNS. He also had respiratory
distress due to pulmonary embolism documented by chest computed tomography scan. This
report highlights the need to determine the uric acid/creatinine ratio caused by increased
renal clearance in LNS in young children. The presence of pulmonary embolism is unusual
and may be the consequence of prolonged immobilization.
Copyright ª 2013, Taiwan Pediatric Association. Published by Elsevier Taiwan LLC. All rights
reserved.
* Corresponding author. Department of Pediatrics, Chung Shan Medical University Hospital, #110, Section 1, Jianguo North Road, Taichung
402, Taiwan.
E-mail address: [email protected] (J.-N. Sheu).
1875-9572/$36 Copyright ª 2013, Taiwan Pediatric Association. Published by Elsevier Taiwan LLC. All rights reserved.
http://dx.doi.org/10.1016/j.pedneo.2012.12.016
Lesch-Nyhan syndrome and pulmonary embolism
1. Introduction
LescheNyhan syndrome (LNS) is a genetic disorder of purine
salvage caused by a complete deficiency of hypoxanthine
phosphoribosyltransferase (HPRT) activity resulting from a
mutation in the corresponding gene on the long arm of the X
chromosome.1 Affected individuals usually suffer from
overproduction of uric acid that may lead to hyperuricemia
or nephrolithiasis.2 High serum uric acid level is usually the
biochemical finding that prompts special testing for the
specific diagnosis. However, infants with LNS may have
borderline serum uric acid levels because of an increased
renal clearance of uric acid.3 Diagnosis can be made by
clinical features and further confirmed by direct sequencing of all exons of the HPRT gene.4,5
Pulmonary embolism is more commonly found in adults
with underlying predisposing factors and conditions that
interfere with normal venous blood flow. In children, pulmonary embolism is rare. The few reports about pediatric
pulmonary embolism reflect its relative rarity and difficult
evaluation.6 Imaging evaluation of pulmonary embolism is
crucial for diagnosis. The use of a multidetector computed
tomography (MDCT) shortens the time of the study and
greatly improves its resolution.7
2. Case Report
A boy aged 1 year and 10 months presented with developmental delay. He was born to nonconsanguineous healthy
parents, and had been diagnosed with nephrolithiasis at the
age of 5 months. When the patient was 6 months old,
313
recurrent airway infection and respiratory distress complicated by laryngomalacia required repeated hospital
admission and central catheterization during hospitalization. He was diagnosed with profound psychomotor
retardation and was bed-ridden since 1 year of age. He also
had chorea when awake, and compulsive self-destructive
behavior such as biting fingers, lips, and buccal mucosa
were noted for about 6 months.
Laboratory tests showed normal complete blood count,
electrolytes, blood gas, ammonia, and lactate. Although
serial serum uric acid level showed normal uric acid level
(7.6 mg/dL), the spot urine uric acid/creatinine ratio was
116/50 or >2.0, which subsequently pointed to a diagnosis
of LNS. There were no particular findings on brain magnetic
resonance image and electroencephalography.
The patient and his family, as well as control individuals
were analyzed using multiplex quantitative polymerase
chain reaction (PCR) to amplify the FGFR2 gene, the KRIT1
gene, and the HPRT gene simultaneously. The HPRT cDNA,
amplified from the total RNA of the patient’s peripheral
blood by multiple specific primers for multiplex PCR,
showed the deletion of exon 6. The family study revealed
that the patient’s mother was a heterozygous carrier
(Figure 1). He was put on allopurinol for his uric acid
overproduction and nephrolithiasis.
Because of his persistent respiratory distress, the study
of chest MDCT with intravenous contrast injection was
performed. The images of chest MDCT showed pulmonary
embolism with nonobstructive lobar and segmental arterial
thrombi in the basal segments of the left lower lobe. This
may explain the persistent respiratory distress of the
patient (Figure 2).
Figure 1 The pedigree and sequence amplified cDNA of the HPRT gene in the patient (black). The patient’s mother is a heterozygous carrier.
314
J.-D. Tsai et al
Figure 2 Chest images using a 320-MDCT (multidetector computed tomography) scanner with a nonhelical one-volume scan
mode. (A) Oblique coronal thin-slab maximum-intensity projection image shows a tubular filling defect (arrow) in the branch of the
left pulmonary artery in the left lower lobe. (B) 3D image using volume rendering techniques shows eccentric thinning of the
embolismic artery (arrow). The distal branches (arrowheads) of the left pulmonary artery are still patent.
3. Discussion
The diagnosis of LNS is frequently delayed until selfmutilation becomes evident.8 It is the most striking feature of LNS and is only present in patients with the complete
enzyme defect, although some patients never show this type
of behavior. The self-mutilation associated with LNS typically first appears with the emergence of teeth, and steadily
worsens with increasing age.9 In the current patient, such
behavior prompted the investigation of his uric acid under
the suspicion of LNS. Patients with HPRT deficiency must be
confirmed by clinical, biochemical, enzymatic, and molecular analyses. In this case, despite fulfilling the diagnosis
criteria of LNS, the patient’s serial serum uric acid only
showed normal uric acid. Thus, the uric acid/creatinine
ratio was used to corroborate the clinical diagnosis. In young
children, renal function is quite efficient in eliminating uric
acid into the bladder, and hence, they may have borderline
hyperuricemia due to increased renal clearance. As such,
the urinary uric acid/creatinine ratio can be used as a
screening test for inherited disorders of purine metabolism
based on the age of the patient. Values for the urinary ratio
should be <1.0 after the age of 3 years.10
The HPRT gene is localized to the Xq26 region, and the
complete amino acid sequence for HPRT is known to be
44 kb, which consists of nine exons with a coding for a
219-amino acid protein that converts hypoxanthine into
inosinic acid and guanine into guanylic acid. To date, more
than 300 disease-associated mutations in the HPRT gene
have been identified,5 but reports of LNS in Taiwan remains
limited. Mak et al11 reported a case series of four patients
with LNS from three families. Three patients in two families
were revealed to have novel missense mutation in exons 3
and 8, and in one patient, a splicing region of intron 4 of the
HPRT encoding region was reported. Hou12 reported a
9-year-old boy with LNS complicated with atlantoaxial
subluxation, and the direct genomic DNA sequencing of the
HPRT gene revealed a single nucleotide substitution in
intron 5. To the best of our knowledge, this is the first
case of LNS with exon 6 deletion reported in Taiwan.
Treatment of LNS remains limited for self-mutilation and
motor syndrome. However, allopurinol should be started as
soon as the enzyme deficiency is diagnosed, although it has
no reported effect on behavioral and neurological symptoms. It should be adjusted to reduce hyperuricemia and
achieve a urinary uric acid/creatinine ratio lower than 1.0.
Allopurinol is efficacious and generally safe as treatment
for uric acid overproduction in patients with HPRT
deficiency.13
Pediatric patients rarely present with conditions that
place them at high risk for thrombus formation. When it
occurs, pulmonary embolism has been shown to cause
serious illness and even death. Traditional risk factors in
adults may be a consequence of continuous immobilization,14 because the pathophysiology of thrombus formation
is blood flow stasis.15 Nonetheless, there is no article on LNS
that suggests the complication of pulmonary embolism, and
as such, it can be considered a consequence of long-term
immobilization stemming from his profound retardation
and recurrent central catheterization.
In conclusion, this report highlights the value of the uric
acid/creatinine ratio as a screening tool for young male
children highly suspected to have LNS. To the best of our
knowledge, this is the first case report of LNS associated
with pulmonary embolism, complicated by the patient’s
long-term immobilization.
Conflicts of interest
The authors have no conflicts of interest relevant to this
article.
References
1. Nyhan WL. Inherited hyperuricemic disorders. Contrib Nephrol
2005;147:22e34.
2. Lesch M, Nyhan WL. A familial disorder of uric acid metabolism
and central nervous system function. Am J Med 1964;36:
561e70.
Lesch-Nyhan syndrome and pulmonary embolism
3. Rinat C, Zoref-Shani E, Ben-Neriah Z, Bromberg Y, BeckerCohen R, Feinstein S, et al. Molecular, biochemical, and
genetic characterization of a female patient with
LescheNyhan disease. Mol Genet Metab 2006;87:249e52.
4. Jinnah HA, Harris JC, Nyhan WL, O’Neill JP. The spectrum of
mutations causing HPRT deficiency: an update. Nucleosides
Nucleotides Nucleic Acids 2004;23:1153e60.
5. Yamada Y, Nomura N, Yamada K, Wakamatsu N. Molecular
analysis of HPRT deficiencies: an update of the spectrum of Asian
mutations with novel mutations. Mol Genet Metab 2007;90:
70e6.
6. Monagle P, Adams M, Mahoney M, Ali K, Barnard D, Bernstein M,
et al. Outcome of pediatric thromboembolic disease: a report
from the Canadian Childhood Thrombophilia Registry. Pediatr
Res 2000;47:763e6.
7. Schoepf UJ. Diagnosing pulmonary embolism: time to rewrite
the textbooks. Int J Cardiovasc Imaging 2005;21:155e63.
8. Nyhan WL. LescheNyhan disease. Nucleosides Nucleotides
Nucleic Acids 2008;27:559e63.
9. Anderson LT, Ernst M. Self-injury in LescheNyhan disease. J
Autism Dev Disord 1994;24:67e81.
315
10. Kaufman JM, Greene ML, Seegmiller JE. Urine uric acid to
creatinine rtio e a screening test for inherited disorders
of purine metabolism. Phosphoribosyltransferase (PRT) deficiency in X-linked cerebral palsy and in a variant of gout. J
Pediatr 1968;73:583e92.
11. Mak BS, Chi CS, Tsai CR, Lee WJ, Lin HY. New mutations of the
HPRT gene in LescheNyhan syndrome. Pediatr Neurol 2000;23:
332e5.
12. Hou JW. Atlantoaxial subluxation with recurrent consciousness
disturbance in a boy with LescheNyhan syndrome. Acta Paediatr 2006;95:1500e4.
13. Torres RJ, Prior C, Puig JG. Efficacy and safety of allopurinol in
patients with hypoxanthineeguanine phosphoribosyltransferase deficiency. Metabolism 2007;56:1179e86.
14. Lee LC, Shah K. Clinical manifestation of pulmonary embolism.
Emerg Med Clin North Am 2001;19:925e42.
15. Dalen JE. Pulmonary embolism: what have we learned since
Virchow? Natural history, pathophysiology, and diagnosis.
Chest 2002;122:1440e56.
Pediatrics and Neonatology (2014) 55, 316e319
Available online at www.sciencedirect.com
journal homepage: http://www.pediatr-neonatol.com
CASE REPORT
Vertebral Artery Dissection Complicated by
Basilar Artery Occlusion
Chia-Yin Kuan a, Kun-Long Hung a,b,c,*
a
Department of Pediatrics, Cathay General Hospital, Taipei, Taiwan
Department of Pediatrics, Cathay General Hospital-Sijhih, New Taipei, Taiwan
c
School of Medicine, Fu-Jen Catholic University, Shinchuang, New Taipei, Taiwan
b
Received Mar 21, 2012; received in revised form Jul 16, 2012; accepted Nov 20, 2012
Available online 23 January 2013
Key Words
basilar artery
occlusion;
pediatric stroke;
vertebral artery
dissection
Acute basilar artery occlusion (ABAO) is an infrequent but potentially fatal complication that
can cause strokes in both adults and children. Traumatic vertebral artery dissection (VAD) is
one of the most common causes of ABAO in young patients. We present a case of an
11-year-old boy with VAD complicated by basilar artery occlusion 2 days after a fight with classmates that caused severe neurological deficits. He did not have any direct head trauma or concomitant risk factors. Clinical symptoms included nausea, vomiting, and rapid alteration of
consciousness. Magnetic resonance imaging showed total occlusion of the basilar artery, and
angiography confirmed VAD from the third to the fourth segments. A history of such subtle precipitating events should be noted when diagnosing young patients with brainstem strokes.
A delay in the diagnosis of ABAO is frequently due to misleading symptoms and signs and the
lack of awareness of this rare condition.
Copyright ª 2013, Taiwan Pediatric Association. Published by Elsevier Taiwan LLC. All rights
reserved.
1. Introduction
Acute basilar artery occlusion (ABAO) is an infrequent but
potentially fatal cause of strokes in both adults and children.1 Traumatic vertebral artery dissection (VAD) is one of
* Corresponding author. Department of Pediatrics, Cathay General Hospital, Taipei, Taiwan.
E-mail address: [email protected] (K.-L. Hung).
the most common causes of ABAO in young patients, and
must be strongly suspected in patients presenting with
cervical pain preceding a dramatic neurological
deterioration.2
Symptomatic occlusion of the basilar artery leads to
various neurological symptoms and signs3 depending on the
mechanism of the stroke,4,5 occlusion localization, and
development of collateral circulation.6,7 The prognosis of
ABAO is generally very poor; however, some patients survive with minor or no neurological deficits.8e10 Herein, we
1875-9572/$36 Copyright ª 2013, Taiwan Pediatric Association. Published by Elsevier Taiwan LLC. All rights reserved.
http://dx.doi.org/10.1016/j.pedneo.2012.12.014
Acute basilar artery occlusion in children
317
An 11-year-old boy presented with acute onset of nausea,
vomiting, and consciousness disturbance. The first assessment took place at a regional hospital, where he underwent
urgent computed tomography, and the result was interpreted as normal. A few hours later, he was transferred to
our intensive care unit owing to a progressive alteration of
consciousness. The boy’s condition continued to deteriorate and he became comatose. The patient had no history
of surgery or major trauma or vehicular accidents. There
were no cases of early-onset or young adulthood stroke in
the family. A neurological reevaluation revealed bilateral
positive Babinski signs and an absence of Doll’s eye sign.
The patient also experienced one episode of generalized
toniceclonic seizure on the 1st day. His Glasgow Coma
Scale score was 6 (E1V1M4) upon admission with apneustic
breathing, fixed dilated pupils, and decerebrate rigidity.
Magnetic resonance imaging, and specifically diffusionweighted and T2 fluid attenuated inversion recovery
images confirmed parenchymal lesions in the pons, bilateral
cerebellar hemisphere, and superior vermis, suggesting
ischemic changes (Figure 1) consistent with the recent
infarction without any signs of hemorrhage. Magnetic resonance angiography disclosed total occlusion of the basilar
artery with proximal bilateral posterior cerebral arteries
reconstituted by bilateral posterior communicating arteries
(Figure 2). A lumbar puncture showed normal findings.
Angiography performed 36 hours after the onset of symptoms disclosed mural filling defects at the left vertebral
artery (VA), V3 and V4 segment, and a suspected smaller
one in the V2 segment, suggesting a recent dissection
(Figure 3). Contrast injections into both VAs showed opacification of the left posterior inferior cerebellar artery and
a common trunk for the right posterior inferior cerebellar
artery and anterior inferior cerebellar artery. The basilar
artery (BA) was completely occluded. Injections into the
carotid arteries (Figure 3) showed patent posterior communicating artery filling both posterior cerebral arteries.
Transcranial Doppler showed no significant stenosis of the
carotid arteries, but did show left VA stenosis with borderline VA flow, and no blood flow in the BA beyond the
anterior inferior cerebellar artery.
When questioned for possible causes of the vertebral
dissection, the patient’s mother could not recall any recent
head trauma, but mentioned that the boy had had a fight
with four classmates 2 days before admission. Complete
blood count, bleeding parameters, and erythrocyte sedimentation rate were all within normal limits. Antiphospholipid antibodies, antinuclear antibodies, and
anticardiolipin antibodies were also normal, excluding
vasculitis and collagen vascular diseases. The venereal
disease research laboratory test result was negative. Cardiac ultrasonography showed no vegetative cardiac valve or
thrombus formation, and electrocardiography excluded
atrial fibrillation. Intra-arterial thrombolysis was not contemplated in our patient, because it was beyond the
accepted 3-hour optimum time window. Throughout
admission, the patient was treated with antiplatelet therapy
(combined
salicylate
and
extended-release
Figure 1 Magnetic resonance imaging of the brain showed
diffusion restriction at the pons, bilateral cerebellar hemisphere, and superior vermis suggesting ischemic changes.
Figure 2 Magnetic resonance angiography showed total
occlusion of the basilar artery with proximal bilateral posterior
cerebral arteries reconstituted by the bilateral patent posterior communicating arteries.
present a case of an 11-year-old boy with acute symptomatic occlusion of the basilar artery distal segment
resulting from VAD, manifesting as severe neurological
symptoms and deficits. Events leading up to the dissection
are examined, as well as related signs and symptoms,
diagnostic examinations, treatment, and follow-up. In
addition, we discuss how lesions of different vessels can
present with distinct outcomes.
2. Case Report
318
Figure 3 Vertebral angiogram showed total occlusion of the
basilar artery beyond the anterior inferior cerebellar artery,
and mural filling defects were noted in the left vertebral artery
(VA), V3 and V4 segment, and a suspected small defect in the
V2 segment.
dipyridamole) to prevent recurrent strokes. Dexamethasone and glycerol were given to suppress brain
swelling, and phenobarbital was added for seizure control.
Somatosensory evoked potentials revealed bilateral
cortical dysfunction, and auditory brainstem evoked
potentials revealed mild, prolonged wave I to wave V
latency in both ears (right > left), indicative of bilateral
central auditory pathway dysfunction, whereas visual
evoked potentials were indicative of bilateral visual pathway dysfunction. Electroencephalography disclosed diffuse
cortical dysfunction with suspected diffuse encephalopathy. There were some complications during the remainder
of his stay, including drug hypersensitivity syndrome, which
we suspected was related to phenobarbital, which was then
managed accordingly.
Twenty days later, the patient was transferred to our
general ward. A neurological examination revealed a
Glasgow Coma Scale score of 7 (E1V2M4), and the patient
remained comatose and quadriplegic. He shed tears and
uttered unintelligible sounds when his mother spoke to him.
Thirty-five days later, the patient was discharged with
subsequent follow-up in the outpatient clinic. At 6 months
follow-up, the patient was still quadriplegic and bedridden.
However, he was more responsive to simple commands and
the functions of swallowing and phonation were restored.
3. Discussion
In general, the annual incidence of pediatric strokes is
estimated to be 2.5 cases per 100,000 children.10 Furthermore, occlusion of the BA in children is even more
rare.2 One of the leading causes of BA occlusion in children
C.-Y. Kuan, K.-L. Hung
is VAD.2 A history of trivial neck or head trauma is often
identified, and this was the case in our patient. The difficulty in diagnosing a BA occlusion in children stems from
the clinical variability, and often its similarity to a seizure
disorder as myoclonic and tonic jerking movements are
frequently seen.11 A delay in the diagnosis of ABAO happens
frequently, owing mainly to misleading symptoms and signs
and the lack of awareness of this rare condition.
In most reported studies in the literature, the V2eV3
segment of the VA is most commonly involved in dissections. The dissection of VA in our case was located at the
V3eV4 segment.
The overall prognosis of this clinical entity depends on the
clinical state at presentation, the length and location of the
occlusion, the degree of recanalization, and the time to
treatment. When the occlusion is located at the proximal part
of the BA, there is a better chance for a good outcome mainly
because of the preservation of blood flow into the brainstem.
A possible mechanism for this may be the collateral supply by
the anterior circulation via the posterior communicating
arteries. More distal occlusions have less favorable outcomes.12 In our case, the patient had total occlusion of the BA
beyond the anterior inferior cerebellar artery.
A successful thrombolytic reversal of ABAO depends on
the early initiation of treatment. Intravenous fibrinolytic
therapy at a cerebral circulation dose within the 1st 3 hours
of ischemic stroke onset offers substantial net benefits for
virtually all patients with potentially disabling deficits.13
Patients treated in the 3e4.5-hour window show a modest, but still clinically worthwhile, therapeutic yield.
Beyond 4.5 hours after onset, no net benefit of therapy has
been demonstrated. Current U.S. and international consensus guidelines accordingly recommend intravenous
thrombolysis when treatment can be initiated within 3
hours from stroke onset, the most well-established treatment time frame.14
No intra-arterial thrombolysis was contemplated in our
patient because it was beyond the accepted therapeutic
time window. We hypothesize that acute anticoagulation
and antiaggregation is the best treatment option in preventing secondary events in cases of embolic strokes. If a
secondary stroke occurs in patients being treated conservatively, and imaging shows expansion of the intramural
hematoma, we suggest parent vessel occlusion to stop the
blood from entering the dissection cavity.
In conclusion, it is worth keeping in mind that VAD is an
important cause of posterior circulation strokes in children
and young adults. We emphasize that a history of precipitating events such as trivial neck torsion should be taken
into consideration when diagnosing young patients with
brainstem strokes.
Conflicts of interest
The authors have no conflicts of interest relevant to this
article.
References
1. Ezaki Y, Tsutsumi K, Onizuka M, Kawakubo J, Yagi N,
Shibayama A, et al. Retrospective analysis of neurological
Acute basilar artery occlusion in children
2.
3.
4.
5.
6.
7.
outcome after intra-arterial thrombolysis in basilar artery
occlusion. Surg Neurol 2003;60:423e30.
Hasan I, Wapnick S, Tenner MS, Couldwell W. Vertebral artery
dissection in children: a comprehensive review. Pediatr Neurosurg 2002;37:168e77.
Kompanje EJ, Walgaard C, de Groot YJ, Stevens M. Historical
sources of basilar artery occlusion. Neurology 2011;76:
1520e3.
Voetsch B, DeWitt LD, Pessin MS, Caplan LR. Basilar artery
occlusive disease in the New England Medical Centre Posterior
Circulation Registry. Arch Neurol 2004;61:496e504.
Ferbert A, Bruckmann H, Drummen R. Clinical features of
proven basilar artery occlusion. Stroke 1990;21:1135e42.
Devuyst G, Bogousslavsky J, Meuli R, Moncayo J, de Freitas G,
van Melle G. Stroke or transient ischemic attacks with basilar
artery stenosis or occlusion: clinical patterns and outcome.
Arch Neurol 2002;59:567e73.
Schonewille WJ, Algra A, Serena J, Molina CA, Kappelle LJ.
Outcome in patients with basilar artery occlusion treated conventionally. J Neurol Neurosurg Psychiatry 2005;76:1238e41.
319
8. Brandt T, Pessin MS, Kwan ES, Caplan LR. Survival with basilar
artery occlusion. Cerebrovasc Dis 1995;5:182e7.
9. Arnold M, Fischer U, Comter A, Gralla J, Findling O, Mattle HP,
et al. Acute basilar artery occlusion in the Basilar Artery
International Cooperation Study: does gender matter? Stroke
2010;41:2693e6.
10. Kirkham FJ. Stroke in childhood. Arch Dis Child 1999;81:85e9.
11. Ropper AH. “Convulsions” in basilar artery occlusion. Neurology 1988;38:1500e1.
12. Cross DT 3rd, Moran CJ, Akins PT, Angtuaco EE, Derdeyn CP,
Diringer MN. Collateral circulation and outcome after basilar
artery thrombolysis. AJNR Am J Neuroradiol 1998;19:1557e63.
13. Hacke W, Donnan G, Fieschi C, Kaste M, von Kummer R,
Broderick JP, et al. Association of outcome with early stroke
treatment: pooled analysis of ATLANTIS, ECASS, and NINDS rtPA stroke trials. Lancet 2004;363:768e74.
14. Chalela JA, Katzan I, Liebeskind DS, Rasmussen P, Zaidat O,
Suarez JI, et al. Safety of intra-arterial thrombolysis in the
postoperative period. Stroke 2001;32:1365e9.
Pediatrics and Neonatology (2014) 55, 320e322
Available online at www.sciencedirect.com
ScienceDirect
journal homepage: http://www.pediatr-neonatol.com
BRIEF COMMUNICATION
Severe Anaphylaxis in Children:
A Single-center Experience
Erdem Topal*, Arzu Bakirtas, Ozlem Yilmaz,
¨l, Mustafa Arga,
Ilbilge H. Ertoy Karago
Mehmet S. Demirsoy, Ipek Turktas
Department of Pediatric Asthma and Allergy, Gazi University Faculty of Medicine, Ankara, Turkey
Received Jan 15, 2013; received in revised form Aug 20, 2013; accepted Oct 7, 2013
Available online 20 November 2013
1. Introduction
Anaphylaxis involves several organ systems and evolves
dynamically, that is, it may spontaneously stop with mild
symptoms or rapidly progress to life-threatening laryngeal
edema or shock. It is not possible to determine the severity of
the attack in advance. Severe anaphylaxis deserves special
attention because of its increased risk of morbidity and
mortality. Although not as much as anaphylaxis in general,1e3
reports of severe anaphylaxis have been increasing in number with time. For this reason our study aimed to investigate
the demographics, atopic status, etiological factors, clinical
properties, treatment, and follow-up of patients who presented with severe anaphylaxis.
2. Methods
We conducted a retrospective medical chart review for a
10-year period between January 2002 and September 2012.
The patients’ files were separately reviewed by two
* Corresponding author. Department of Pediatric Allergy and
Asthma, Gazi University Faculty of Medicine, 06510 Bes‚evler,
Ankara, Turkey.
E-mail address: [email protected] (E. Topal).
pediatric allergists. Children whose anaphylaxis diagnosis
was confirmed by the two allergists and who fulfilled the
criteria for severe anaphylaxis according to the position
paper of the European Academy of Allergology and Clinical
Immunology4 were included in the final analysis. Demographics, triggers, number of previous anaphylaxis episodes, clinical manifestations and involved systems, the
interval between exposure and the onset of anaphylaxis,
treatment of the acute episode, comorbid diseases, and
long-term management were recorded. Etiologic factors
were determined by history and skin prick and/or intradermal tests with the suspected allergens mentioned in the
history.
The study was approved by the Ethics Committee of Gazi
University, Ankara, Turkey.
3. Results
During a period of 10 years in the Gazi University Pediatric
Allergy and Asthma Department, 34 (25%) of 136 cases were
diagnosed as severe anaphylaxis. Twenty-one of these
patients were male (61.8%), and the mean age was
79.0 61.7 months. Drugs were the most common etiological agents (61.8%; Table 1) . In six cases (17.6%) there
was a comorbidity of atopic disease (3 asthma cases, 2
allergic rhinitis cases, and 2 atopic eczema cases). Five
1875-9572/$36 Copyright ª 2013, Taiwan Pediatric Association. Published by Elsevier Taiwan LLC. All rights reserved.
http://dx.doi.org/10.1016/j.pedneo.2013.10.002
Severe anaphylaxis in children
321
cases had a history of type 1 food allergy, four cases had a
history of type 1 drug allergy, and four cases had an
anaphylaxis attack previously (foods were the etiological
agents in all previous attacks of anaphylaxis). The analysis
of the reported interval between exposure to the causative
agent and the appearance of the first symptom showed that
in 20 cases (58.8%) the reaction was immediate
(1e5 minutes), in eight cases (23.5%) the reaction within
5e30 minutes, and in six cases (17.6%) the reaction was
within 30e120 minutes. Respiratory symptoms (25 cases,
73.5%) were the most common symptoms observed, followed by cardiovascular and cutaneous symptoms (23
cases, 67.6%; Table 1). Hypotension was detected in 21 of
25 patients whose blood pressure was measured. Five cases
Table 1 Causes, symptoms, and signs of severe anaphylaxis (n Z 34).
(14.7%) had a cardiac arrest. Twelve cases (35.2%) had
stridor, 17 cases (50%) had cyanosis, and 17 cases (50%) had
syncope. The youngest patient who had a cardiac arrest
was 30 months old and the oldest patient was 72 months
old. All were male, and the reaction occurred within the
first 5 minutes after exposure. The etiological agent was
ceftriaxone in three cases, penicillin in one case, and a
general anesthetic agent in one case. Detailed information
was available in three cases that had cardiac arrest. Two
patients responded to cardiopulmonary resuscitation within
5 minutes (etiological agents were ceftriaxone in one case
and general anesthetic agent in the other case). The other
case was treated with cardiopulmonary resuscitation for
15 minutes, had been intubated, and taken to the intensive
care unit (etiological drug was ceftriaxone). After this
event, the child recovered but had neurological problems.
Treatment of anaphylaxis in the medical facility for those
30 cases whose medical records have been accessed
included antihistamines in 26 patients (86.6%), corticosteroids in 27 patients (90%), epinephrine in 20 patients
(66.6%), oxygen in 28 patients (93.3%), IV fluid in 23 patients (76.6%), and a b-2 agonist in five patients (16.6%).
Four patients (11.7%) needed IM adrenaline twice. Six patients (17.6%) were observed in the emergency room for
more than 24 hours, five patients (14.7%) were followed up
in the intensive care unit, and two patients were hospitalized for underlying systemic disease. Thirteen of 34
cases were prescribed an epinephrine autoinjector: six for
food (46.1%), six for hymenoptera venom (46.1%), and one
for exercise-induced (7.8%) anaphylaxis. Thirty-two patients could be contacted for follow-up. Two had experienced a second episode after diagnosis (6.2%). Although an
epinephrine autoinjector was prescribed and patients/
parents were trained in their use previously, only one of
them had used the epinephrine self-administration device
(Epipen) at the time of the reaction.
n (%)
Specific agent (n)
Causative agent
Drugs
21 (61.8)
Food
6 (17.6)
Venom
6 (17.6)
Antibiotics (11)
Cephalosporins (6)
Penicillin (4)
Vancomycin (1)
NSAIDs (3)
Enzyme replacement
therapy (3)
a-L-iduronate (1)
Idursulfase (1)
Galsulfase (1)
General anesthetic
agents (3)
Insulin (1)
Hen’s egg (3)
Cow’s milk (2)
Peach (1)
Vespids (4)
Aphids (2)
Exercise-induced
Symptom or sign
Respiratory
Dyspnea
Stridor
Wheezing
Cough
Rhinorrhea
Hoarseness
Cyanosis
Cardiovascular
Hypotension
Syncope
Arrest
Cutaneous
Angioedema
Urticaria
Flushing
Pruritus
Gastrointestinal
Persistent vomiting
1 (2.9)
4. Discussion
25 (73.5)
6 (17.6)
12 (35.3)
2 (5.9)
6 (17.6)
2 (5.9)
2 (5.9)
17 (50)
23 (67.6)
21/25 (84)*
17 (50)
5 (14.7)
23 (67.6)
17 (50)
18 (52.9)
6 (17.6)
4 (11.8)
4 (11.8)
4 (11.8)
This study showed that drugs were the major etiological
factor in severe anaphylaxis cases, and symptoms and signs
occurred immediately after contact with the suspicious
agent. Although the most frequent symptoms/signs involved
were respiratory, hypotension was detected in a remarkable
percentage of patients with severe ana-phylaxis.
In retrospective studies of anaphylaxis cases, cutaneous
symptoms/signs were the most common during the reaction. Respiratory and cardiovascular symptoms were
generally observed less often than cutaneous symptoms/
signs.1e3 However, Stoevesandt et al5 reported that the
absence of urticaria and angioedema was frequently associated with venom-induced severe anaphylaxis. In our
study, respiratory symptoms/signs were most prominent in
cases of severe anaphylaxis. Cardiovascular and cutaneous
symptoms/signs followed respiratory symptoms/signs. Also,
more than 80% of our patients were reported to be hypotensive, which is an important sign. It shows the severity of
the reaction and is an alarming signal for morbidity and
mortality from anaphylaxis. Renaudin et al6 analyzed 333
cases who were diagnosed with drug-induced severe
anaphylaxis. Among those, 76.6% had anaphylactic shock.
NSAIDs Z non-steroidal anti-inflammatory drugs.
* Only 25 of the cases whose blood pressure was measured.
322
These results illustrate that hypotension is a common
finding in severe anaphylaxis.
Pumphrey7 studied the average time from contact with
the trigger to onset of cardiopulmonary arrest in severe
anaphylaxis that resulted in death. This time was reported
to be 30 minutes in food-induced, 15 minutes in venominduced, and 5 minutes in drug-induced anaphylaxis.
Stoevesandt et al5 found a significant association between
the onset of the reaction and the severity of venominduced anaphylaxis, that is, if the reaction starts within
the first 5 minutes, the episode is more severe. In our study,
two-thirds of patients with severe anaphylaxis had reactions that developed within the first 5 minutes.
Failure to measure serum tryptase level in patients
diagnosed with severe anaphylaxis is a limitation of this
study. It has been shown that a high serum tryptase level is
a risk factor in severe anaphylaxis.5 Moreover, the small
number of patients and the retrospective nature of the
study are additional limitations.
In conclusion, drugs are the major triggers of severe
anaphylaxis. The reaction occurs immediately after contact
with the etiological factor. The most frequent symptoms/
signs involve the respiratory system followed by the cardiovascular system and cutaneous.
Conflicts of interest
The authors have no conflicts of interest relevant to this
article.
E. Topal et al
References
1. Orhan F, Canitez Y, Bakirtas A, Yilmaz O, Boz AB, Can D, et al.
Anaphylaxis in Turkish children: a multi-centre, retrospective,
case study. Clin Exp Allergy 2011;41:1767e76.
2. De Swert LF, Bullens D, Raes M, Dermaux AM. Anaphylaxis in
referred pediatric patients: demographic and clinical features,
triggers, and therapeutic approach. Eur J Pediatr 2008;167:
1251e61.
3. Rudders SA, Banerji A, Clark S, Camargo Jr CA. Age-related
differences in the clinical presentation of food-induced
anaphylaxis. J Pediatr 2011;158:326e8.
4. Muraro A, Roberts G, Clark A, Eigenmann PA, Halken S, Lack G,
et al. The management of anaphylaxis in childhood: position
paper of the European Academy of Allergology and Clinical
Immunology. Allergy 2007;62:857e71.
5. Stoevesandt J, Hain J, Kerstan A, Trautmann A. Over- and underestimated parameters in severe hymenoptera venom-induced
anaphylaxis: cardiovascular medication and absence of urticaria/angioedema. J Allergy Clin Immunol 2012;130:698e704.
6. Renaudin JM, Beaudouin E, Ponvert C, Demoly P, MoneretVautrin DA. Severe drug-induced anaphylaxis: analysis of 333
cases recorded by the Allergy Vigilance Network from 2002 to
2010. Allergy 2013;68:929e37.
7. Pumphrey RS. Lessons for management of anaphylaxis from a
study of fatal reactions. Clin Exp Allergy 2000;30:1144e50.
Pediatrics and Neonatology (2014) 55, 323e325
Available online at www.sciencedirect.com
ScienceDirect
journal homepage: http://www.pediatr-neonatol.com
BRIEF COMMUNICATION
Comparison of Risk for Early-onset Sepsis
in Small-for-gestational-age Neonates and
Appropriate-for-gestational-age Neonates
Based on Lower Levels of White Blood Cell,
Neutrophil, and Platelet Counts
Nora Hofer a,*, Silvia Edlinger a, Bernhard Resch a,b
a
Research Unit for Neonatal Infectious Diseases and Epidemiology, Medical University of Graz,
Graz, Austria
b
Division of Neonatology, Department of Pediatrics and Adolescent Medicine,
Medical University of Graz, Graz, Austria
Received Jun 5, 2013; received in revised form Nov 5, 2013; accepted Dec 6, 2013
Available online 21 April 2014
Intrauterine growth restriction and its postnatal equivalent, that is, small for gestational age (SGA) at birth, are
known to be risk factors for adverse neonatal outcomes.1
Like other organs, bone marrow is dependent on oxygen
and nutritional supply and some reports suggest that SGA
neonates are at an increased risk for lower levels of white
blood cell (WBC) and neutrophil counts during the first few
days after birth.2e4 However, evidence is scarce and the
clinical importance of these differences is unclear. The aim
of this study was to analyze WBC, neutrophil, and platelet
counts during the first 3 days of life and compare the
risk for early onset sepsis in SGA and appropriate-forgestational-age (AGA) neonates.
We used a preexisting database, which contained laboratory and clinical information from all neonates who were
* Corresponding author. Department of Pediatrics and Adolescent
Medicine, Medical University of Graz, Auenbruggerplatz 34/2, A-8036
Graz, Austria.
E-mail address: [email protected] (N. Hofer).
hospitalized in our neonatal intensive care unit within the
first 24 hours of life during the 5-year-period from 2004 to
2008. We excluded neonates for whom no complete blood
count was determined during the first 3 days of life as well
as neonates with incomplete clinical information. The study
was approved by the local ethics committee. We analyzed
1110 WBCs, 865 neutrophils, and 1104 platelets counts
obtained from 501 preterm and 236 term neonates in the
first 3 days of life. Sixty of the 501 preterm neonates (12%)
and 39 of the 236 term neonates (17%) were SGA (birth
weight <10 percentile, as defined by Voigt et al5).
The SGA neonates had lower median WBC (11.1 103/mL
vs. 13.8 103/mL, p < 0.001), lower median neutrophil
counts (7.3 103/mL vs. 9.5 103/mL, p Z 0.001), and
lower median platelet counts (194 103/mL vs. 225 103/
mL, p < 0.001) in the first 3 days of life compared with AGA
neonates. The same was true when analyzing values separately for preterm (WBC: 9.0 103/mL vs. 12.5 103/mL,
p < 0.001; neutrophil count: 5.3 103/mL vs. 8.3 103/mL,
p < 0.001; platelet count: 167 103/mL vs. 218 103/mL,
http://dx.doi.org/10.1016/j.pedneo.2013.12.006
1875-9572/Copyright ª 2014, Taiwan Pediatric Association. Published by Elsevier Taiwan LLC. All rights reserved.
324
N. Hofer et al
Figure 1 White blood cell (WBC), neutrophil, and platelet counts in (A) preterm and (B) term small-for-gestational-age neonates
(gray bars) and appropriate-for-gestational-age neonates (white bars). Bars represent the interquartile range (25the75th
percentile) with median, and lines indicate 5th percentile and 95th percentile. Statistical analysis was performed using ManneWhitney U test.
p < 0.001) and term neonates (WBC: 16.7 103/mL vs.
18.3 103/mL, p Z 0.007; neutrophil count: 10.8 103/mL
vs. 13.7 103/mL, p Z 0.024; platelet count: 214 103/mL
vs. 239 103/mL, p Z 0.047). The count distribution in the
first 3 days is shown in Fig. 1.
Among the preterm neonates, 25 and 103 patients (5.0%
and 20.6%, respectively) had culture-proven and clinical
early onset neonatal sepsis, and among the term neonates
the figures were 14 and 50 (5.9% and 21.2%), respectively.
The SGA neonates were not at an increased risk for cultureproven and clinical sepsis (odds ratio was 1.0 for both in
preterm neonates and 1.1 for both in term neonates). In
neonates with culture-proven sepsis, WBC, neutrophil, and
platelet count did not differ between SGA and AGA neonates. In clinical sepsis, preterm SGA neonates had lower
WBC, neutrophil, and platelet counts compared with preterm AGA neonates (WBC: 5.8 103/mL vs. 11.0 103/mL,
p < 0.001; neutrophils: 3.6 103/mL vs. 7.7 103/mL,
p Z 0.023; platelet count: 155 103/mL vs. 205 103/mL,
p Z 0.006), but no differences were found in term neonates.
We divided neonates in groups of WBC, neutrophil, and
platelet counts within the lower quartile and within the
upper three quartiles (cutoff for WBC 10.0 103/mL, for
neutrophils 5.9 103/mL, and for platelet count 177 103/
mL). When comparing these groups, there was no difference
in risk for culture-proven sepsis between preterm and term
SGA and AGA infants.
The present analysis adds to the growing body of evidence to demonstrate that intrauterine growth restriction
influences WBC, neutrophil, and platelet counts with lower
values in SGA neonates. Few references are found on this
topic and in most analyses the study population is
restricted, by either limits of gestational age and birth
weight, or presence or absence of other conditions (i.e.,
chorioamnionitis, funisitis). In this brief report, we include
a high number of neonates of all gestational ages. Despite
the differences in blood counts between SGA and AGA babies, SGA neonates were not at an increased risk of early
onset neonatal sepsis.
Previous studies revealed conflicting results on the impact
of lower WBC and absolute neutrophil count on sepsis or
mortality.3,6,7 Treatment of neutropenia using hematopoietic
growth factors has been implemented in certain cases.
Further research on the clinical relevance is needed
prior to when such studies can be put into clinical practice.
In this study, we only analyzed laboratory values determined during the first 3 days of life. The SGA neonates are
known to be at an increased risk for late-onset infection, and
more research is necessary on the immunological aspect in
SGA neonates, which may contribute to an increased risk.
In conclusion, SGA neonates had lower WBC, neutrophil,
and platelet counts compared with their AGA counterparts
during the first 3 days of life. The SGA neonates were not at
an increased risk for early onset neonatal sepsis.
Conflicts of interest
All contributing authors declare no conflicts of interest.
References
1. Zeitlin J, El Ayoubi M, Jarreau PH, Draper ES, Blondel B,
Ku
¨nzel W, et al. Impact of fetal growth restriction on mortality
and morbidity in a very preterm birth cohort. J Pediatr 2010;
157:733e9.
Blood counts in SGA neonates
2. Wirbelauer J, Thomas W, Rieger L, Speer CP. Intrauterine
growth retardation in preterm infants 32 weeks of gestation
is associated with low white blood cell counts. Am J Perinatol
2010;27:819e24.
3. Procianoy RS, Silveira RC, Mussi-Pinhata MM, Souza Rugolo LM,
Leone CR, de Andrade Lopes JM, et al. Sepsis and neutropenia
in very low birth weight infants delivered of mothers with
preeclampsia. J Pediatr 2010;157:434e8.
4. Christensen RD, Henry E, Wiedmeier SE, Stoddard RA,
Lambert DK. Low blood neutrophil concentrations among
extremely low birth weight neonates: data from a multihospital health-care system. J Perinatol 2006;26:682e7.
325
5. Voigt M, Schneider KT, Ja
¨hrig K. Analysis of a 1992 birth sample
in Germany. 1: New percentile values of the body weight of
newborn infants. Geburtshilfe Frauenheilkd 1996;56:550e8
[Article in German].
6. Doron MW, Makhlouf RA, Katz VL, Lawson EE, Stiles AD.
Increased incidence of sepsis at birth in neutropenic infants of
mothers with preeclampsia. J Pediatr 1994;125:452e8.
7. Teng RJ, Wu TJ, Garrison RD, Sharma R, Hudak ML. Early
neutropenia is not associated with an increased rate of nosocomial infection in very low-birth-weight infants. J Perinatol
2009;29:219e24.
Pediatrics and Neonatology (2014) 55, 326e327
Available online at www.sciencedirect.com
ScienceDirect
journal homepage: http://www.pediatr-neonatol.com
LETTER TO THE EDITOR
Pulmonary Hemorrhage in Very-low-birthweight Infants
To the Editor
1
The study by Yen et al is very interesting and possibly
clinically useful. According to their study, between 2006
and 2011, 18 (3.2%) of 469 very-low-birth-weight (VLBW)
infants developed severe pulmonary hemorrhage (PH).
The mean gestational age (GA) was 27 weeks, the mean
birth weight (BW) was 822 g, and the onset age was 2.5
days after birth. Among these 18 infants, 16 (88.9%) had
respiratory distress syndrome (RDS) and 13 received
surfactant therapy; eight patients (44.4%) had patent
ductus arteriosus (PDA), which was confirmed by
echocardiography.
All 13 patients who received surfactant therapy developed severe PH within 72 hours after surfactant treatment.
This is about the timing that overlapped with the emergence of hemodynamically significant PDA. The paper,
however, did not report how many infants among these 13
patients developed hemodynamically significant PDA and
whether or not they received any treatment for it. When
did the other five patients (including 2 without RDS and 3
with RDS but who did not receive surfactant therapy)
develop severe PH? Did they also have hemodynamically
significant PDA?
At present, there is no consensus on the treatment for
severe PH in VLBW infants. Treatment for severe PH in their
neonatal intensive care unit included not only intratracheal
epinephrine spraying/irrigation but also blood component
therapy for coagulopathy or thrombocytopenia. The paper
also did not report the number of infants who developed
severe PH and presented with bleeding problems.
The precise etiology of PH in VLBW infants remains unclear. Lin et al2 reported that the hemodynamically significant PDA had an odds ratio of 8.5 (p < 0.006) and RDS with
surfactant therapy had an odds ratio of 7.4 (p < 0.006) for
severe PH in a cohort of 1997e1998, with an incidence of
massive PH of 5.9% (20/340), mean GA of 26.9 weeks, and
mean BW of 909 g. Chen et al3 reported that there was no
difference in the incidence of PDA between PH infants and
non-PH infants, but did show a significant difference in the
use of surfactant therapy (p < 0.001) in 399 VLBW infants
admitted to Kaohsiung Veterans General Hospital between
2000 and 2010, with a 4% incidence of massive PH (16/399),
mean GA of 26.1 weeks, and mean BW of 865 g. These data,
as compared with that of Yen et al,1 imply that the incidence of severe PH seems to have been reducing in the
recent 5 years in Taiwan.
As the authors described, the risk factors for PH were
prematurity, intrauterine growth restriction, respiratory
problems, PDA, bleeding problems, ventilator usage, and
surfactant treatment.1 Without an answer to these questions, it is difficult to realize the causes of severe PH and
the exact role of their current treatment including intratracheal epinephrine therapy, high-frequency oscillation
ventilator, or surfactant supplement therapy in VLBW infants with severe PH.
Conflicts of interest
All contributing authors declare no conflicts of interest.
References
1. Yen TA, Wang CC, Hsieh WS, Chou HC, Chen CY, Tsao PN. Shortterm outcome of pulmonary hemorrhage in very-low-birthweight preterm infants. Pediatr Neonatol 2013;54:330e4.
2. Lin TW, Su BH, Lin HC, Hu PS, Peng CT, Tsai CH, et al. Risk
factors of pulmonary hemorrhage in very-low-birth-weight infants: a two-year retrospective study. Acta Paediatr Taiwan
2000;41:255e8.
http://dx.doi.org/10.1016/j.pedneo.2014.01.005
1875-9572/Copyright ª 2014, Taiwan Pediatric Association. Published by Elsevier Taiwan LLC. All rights reserved
.
Letter to the Editor
3. Chen YY, Wang HP, Lin SM, Chang JT, Hsieh KS, Huang FK, et al.
Pulmonary hemorrhage in very low-birthweight infants: risk
factors and management. Pediatr Int 2012;54:743e7.
Bai-Horng Su*
Hsiang-Yu Lin
Fu-Kuei Huang
Ming-Luen Tsai
Department of Neonatology, China Medical University
Children’s Hospital, Taichung, Taiwan
327
Department of Pediatrics, School of Medicine,
China Medical University, Taichung, Taiwan
*Corresponding author. Department of Neonatology,
China Medical University Children’s Hospital, Number 2,
Yuh-Der Road, Taichung 404, Taiwan.
E-mail address: [email protected] (B.-H. Su)
Nov 27, 2013
Pediatrics and Neonatology (2014) 55, 328
Available online at www.sciencedirect.com
ScienceDirect
journal homepage: http://www.pediatr-neonatol.com
LETTER TO THE EDITOR
Reply: Pulmonary Hemorrhage in Very-lowbirth-weight Infants
We thank the authors for their comments on pulmonary
hemorrhage (PH) in very-low-birth-weight infants.1 Indeed,
the timing of surfactant use and the occurrence of hemodynamically significant patent ductus arteriosus (PDA) may
occur consequently. Prior to PH, no heart murmur or hemodynamically significant PDA was noted in our cases. In this
study,2 13 patients who received surfactant therapy developed PH within 72 hours of surfactant therapy. Among these
patients, half (7 /13: 53%) were diagnosed with coexistent
PDA by echocardiography during PH. PDA was bidirectional
shunt in five patients. For the other five patients who
developed severe PH despite no surfactant use, the median
age for severe PH was 3.2 days (range, 2e7 days). Only one
patient was diagnosed with hemodynamically significant PDA
during the occurrence of severe PH. This patient was not
diagnosed with respiratory distress syndrome but he presented with thrombocytopenia soon after birth. PH developed at age 3 days and cardiac echography showed a
bidirectional flow with left atrium:aortic root ratio of 1.8.
The patient received PDA ligation on the next day.
Why was no heart murmur found prior to PH in these infants? One possible explanation is that the pulmonary pressure
was still high in these very-low-birth-weight infants with respiratory distress syndrome. After surfactant therapy, the
pulmonary compliance improved and it caused dramatically
decreased pulmonary pressure. If PDA coexists, the dramatically decreased pulmonary pressure will result in significant
left-to-right shunt and it may lead to PH due to volume overload. This hemodynamic change may happen quickly even
prior to when hemodynamically significant PDA is found.
As mentioned by Cole et al,3 defective coagulation
function may only serve to exacerbate or prolong the
hemorrhage rather than initiate it. In their study, only one
(1/10: 10%) patient was found to have definite coagulopathy preceding the occurrence of severe PH. In our study,
samples preceding PH were taken from 18 patients within
5e48 hours and six (33%) patients had thrombocytopenia
(<150 109 platelets/L). Therefore, thrombocytopenia
and coagulopathy may worsen the condition of severe PH
but are insufficient to cause PH without other precipitating
factors.
Therefore, the incidence of severe PH is higher in patients receiving surfactant therapy with coexisting PDA.
Early detection of hemodynamically significant PDA and
appropriate treatment may also play an important role in
the management of severe PH. The outcome of severe PH
may worsen if accompanied by bleeding tendency.
Conflicts of interest
All contributing authors declare no conflicts of interest.
References
1. Su BH, Lin HY, Huang FK, Tsai ML. Pulmonary hemorrhage in very
low birth weight infants. Pediatr Neonatal 2014;55:326e7.
2. Yen TA, Wang CC, Hsieh WH, Chou HC, Chen CY, Tsao PN. Shortterm outcome of pulmonary hemorrhage in very-low-birthweight preterm infants. Pediatr Neonatol 2013;54:330e4.
3. Cole VA, Normand IC, Reynolds EO, Rivers RP. Pathogenesis of
hemorrhagic pulmonary edema and massive pulmonary
hemorrhage in the newborn. Pediatrics 1973;51:175e87.
Ting-An Yen
Po-Nien Tsao*
Department of Pediatrics, National Taiwan University
Hospital, National Taiwan University College of Medicine,
Taipei, Taiwan
*Corresponding author. Department of Pediatrics, National
Taiwan University Hospital and National Taiwan University
College of Medicine, Number 7, Chung-Shan South Road,
Taipei 100, Taiwan.
E-mail address: [email protected] (P.-N. Tsao)
http://dx.doi.org/10.1016/j.pedneo.2014.01.006
1875-9572/Copyright ª 2014, Taiwan Pediatric Association. Published by Elsevier Taiwan LLC. All rights reserved.
Dec 10, 2013