Download Identification of bacteria and potential sources in neonates at risk of

Journal of Medical Microbiology (2012), 61, 31–41
DOI 10.1099/jmm.0.034926-0
Identification of bacteria and potential sources in
neonates at risk of infection delivered by Caesarean
and vaginal birth
Cecilia Gonzales-Marin,1 David A. Spratt,2 Michael R. Millar,3
Mark Simmonds,4 Stephen T. Kempley5 and Robert P. Allaker1
1
Correspondence
Institute of Dentistry, Barts and The London School of Medicine and Dentistry,
Queen Mary University of London, 4 Newark Street, London E1 2AT, UK
Cecilia Gonzales-Marin
[email protected]
2
Division of Microbial Diseases, Eastman Dental Institute, University College London,
256 Gray’s Inn Road, London WC1X 8LD, UK
3
Department of Medical Microbiology, Royal London Hospital, Barts and The London NHS Trust,
37 Ashfield Street, London E1 1BB, UK
4
Wolfson Institute of Preventive Medicine, Barts and The London School of Medicine and Dentistry,
Charterhouse Square, London EC1M 6BQ, UK
5
Centre for Paediatrics, Barts and The London School of Medicine and Dentistry,
Queen Mary University of London, 4 Newark Street, London E1 2AT, UK
Received 3 June 2011
Accepted 19 August 2011
Neonatal gastric aspirates (NGA) are routinely screened in UK hospitals to investigate fetal/
neonatal infections associated with cases of adverse pregnancy outcome (APO). The aim of this
study was to describe and compare the microbiology of NGA from Caesarean and vaginal
deliveries using molecular methods, and to evaluate other possible clinical and non-clinical
variables that may have determined the presence of the bacteria in the samples. The value of using
NGA and molecular methods to investigate potential pathogens associated with the risk of early
infection was also evaluated. Bacteria were identified by a combined molecular approach on the
basis of the 16S rRNA gene using both clone analysis and denaturing gradient gel
electrophoresis. A total of 43 and 34 different species were identified in the vaginal (n5121) and
Caesarean (n5119) deliveries, respectively; 26 of the species observed (51 %) were common to
both modalities, although usually less prevalent in the Caesarean cases. Multivariate analysis
confirmed an association between infection and prolonged rupture of membranes in vaginal
deliveries (odds ratio55.7, 95 % confidence interval51.1–29.0). Various associations between
infection and given variables were also shown, including labour, intrapartum antibiotic prophylaxis,
and time and place of sample collection. The molecular methods allowed identification of a
range of bacteria and potential sources not previously observed in NGA, including possible
genito-urinary, gastrointestinal and oral pathogens. NGA represents a valuable sample for
investigating potential pathogens associated with APO and the risk of early infection in neonates
using molecular methods.
INTRODUCTION
Adverse pregnancy outcomes (APOs) denote those complications presented during pregnancy or immediately
after birth that may have caused the termination of the
pregnancy and compromise the well-being of the fetus,
neonate and/or the mother. Complications occur in 11 %
Abbreviations: APO, adverse pregnancy outcome; DGGE, denaturing
gradient gel electrophoresis; GBS, group B streptococci; IAP, intrapartum antibiotic prophylaxis; NGA, neonatal gastric aspirates; ROM, rupture of membranes.
034926 G 2012 SGM
of pregnancies in the UK (ONS, 2005), many of these have
unknown causes, which represent a huge concern to both
society and the health-care system (Petrou, 2003). Intraamniotic infection is an important and frequent cause of
APO. It has been associated with 25–40 % cases of preterm
birth, which accounts for 75 % of perinatal mortality and
morbidity (McCormick, 1985). Also, women with microbial invasion of the amniotic cavity are more likely to
present spontaneous rupture of membranes (ROM),
develop chorioamnionitis and present with an adverse
perinatal outcome (Romero et al., 2006). Furthermore,
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sat, 29 Apr 2017 17:46:51
Printed in Great Britain
31
C. Gonzales-Marin and others
fetal bacteraemia has been detected in 30 % of cases with
preterm ROM and positive amniotic fluid cultures (Carroll
et al., 1996), with a high number of neonates requiring
special care within neonatal intensive care units.
Bacteria identified in neonatal gastric aspirates (NGA) may
be predictive of those pathogens possibly associated with
the occurrence of an APO and early onset neonatal
infection, as positive detection of bacteria in NGA has
shown to correlate 100 % with the presence of chorioamnionitis and/or presentation of preterm ROM in preterm
infants (Miralles et al., 2005). These fluids are routinely
cultured in UK hospitals as the minimum standard
operating procedure to investigate fetal/neonatal infections
in cases of APO, such as preterm birth or low birth weight;
and in newborns that are considered at risk of early
infection. This is mainly based upon non-specific clinical
features, such as respiratory distress, unstable temperature
and maternal fever, or due to other factors such as
congenital anomalies, the use of invasive procedures and
devices during birth, women with previous group B
streptococcal infection and prolonged ROM (HPA, 2004).
There is still a limited knowledge as regards associated
pathogen(s) in APO and their origin, which significantly
limits prevention and provision of treatment. For example,
evaluation of NGA as a tool to predict neonatal sepsis has
been of mixed success (Evans et al., 1988; Leibovich et al.,
1987; Puri et al., 1995). One limitation is related to the
culture techniques routinely used because of the large
number of uncultivable bacteria, as well as the very specific
environmental or nutritional requirements to grow fastidious species (Aas et al., 2005). Therefore, a more detailed
and sensitive analysis is required to better evaluate the
potential use of NGA for clinical investigation and further
research.
In theory, NGA should contain bacteria to which the fetus
has been exposed to in utero (swallowed amniotic fluid).
However, a number of micro-organisms present in the
samples may have been acquired during delivery from the
maternal vagina and skin (Dominguez-Bello et al., 2010);
also, possible bacterial contaminants from the hospital
environment and personnel may have reached the samples during delivery, collection and transportation of the
sample. Likewise, the routine administration of intrapartum antibiotic prophylaxis (IAP) could have altered the
microbiota in mothers and neonates, and may have
influenced the bacterial content.
Hence, the aims of this study were: (i) to describe and
compare bacterial prevalence in samples of NGA between
neonates at risk of early infection that were born by
Caesarean and vaginal birth, (ii) to determine if other
clinical and/or non-clinical variables may have contributed to the presence of bacteria or specific bacteria in the
samples, (iii) to evaluate the value of using NGA and
molecular methods to investigate the source of potential
pathogens associated with the occurrence of APO and the
risk of early infection in neonates.
32
METHODS
Samples. A total of 240 samples of NGA were analysed in this study:
121 were obtained from vaginal deliveries and 119 from Caesarean
sections. This study was approved by the Outer North East London
Research Ethics Committee (REC reference no. 08/H0701/61). Gastric
aspirates are routinely taken at the maternity and neonatal wards,
Royal London Hospital, Barts and The London NHS Trust, under the
‘Standard operating procedure for the investigation of NGA and
infection screen swabs from neonates’ (reference no. BSOP 23i4.1)
from newborns that present with a clinical manifestation of possible
neonatal sepsis or are at risk of developing an infection (HPA, 2004).
These fluids are obtained through aspiration: inserting a sterile
nasogastric tube into the stomach of the newborn during the first
hours post-delivery and before feeding.
DNA isolation. Due to the high viscosity of the samples, NGA
received pre-treatment with a sputum liquefying agent as suggested
previously (Jones et al., 2010). Genomic DNA was extracted by using the ArchivePure DNA yeast and Gram 2/+ purification kit
(Flowgen Bioscience) following the manufacturer’s recommendations. DNA was then stored at 4 uC until the amplification step was
performed.
Cloning analysis. Bacterial 16S rRNA genes were amplified from the
genomic DNA using modified universal primers as suggested by
Frank et al. (2008). This set of primers amplified the majority of the
16S rRNA gene, approximately 1500 bp in length. The amplification
protocol was performed as described previously (Gafan et al., 2004).
The amplicons generated by PCR were cloned into One Shot
chemically competent TOP10 Escherichia coli using a TOPO TA
cloning kit (Invitrogen) as recommended by the manufacturer. All
PCR products were then purified using the QIAquick PCR
purification kit (Qiagen) for subsequent sequencing. The number of
clones analysed in this study was based upon the richness as seen in 10
initial clones. It was observed that those samples identifying up to
three species in 10 clones did not show any significant increase when
raising the number of clones analysed to 50. However, richer samples
required a higher number of clones until no significant change was
observed (data not shown).
Denaturing gradient gel electrophoresis (DGGE). A shorter
fragment of the 16S rRNA gene (~235 bp) was amplified. Touchdown
PCR and methods to perform DGGE have been explained elsewhere
(Gafan & Spratt, 2005). The observed bands were excised using a
sterile scalpel and eluted in water at 4 uC for 24 h as described by
Ampe et al. (1999). Eluted DNA (1 ml) from each DGGE band was
reamplified. PCR products were then purified and sequenced.
Sequencing. All sequencing was carried out at the Genome Centre,
Barts and The London School of Medicine and Dentistry, Queen
Mary University of London, with an ABI 3700 automated DNA
sequencer, producing sequences of 800 bp (from cloning) and
150 bp (from DGGE) in length. The sequences of the 16S rRNA
gene fragments were analysed by comparison using the Ribosomal
Database Project (Cole et al., 2007) and the Human Oral
Microbiome Database (www.homd.org) (Chen et al., 2010) based
upon ¢97 % similarity with the closest known relative (Patel,
2001).
Statistical analysis. Two-sided Fisher’s exact tests were used to
assess the associations between detection of bacteria and the binomial
variables. The Mann–Whitney U test and the Independent t-test were
also used for quantitative data. Multivariate logistic regression models
were constructed to determine the effect of the covariates on the
detection of bacteria after adjusting for important variables. A P value
of less than 0.05 was taken to indicate statistical significance.
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sat, 29 Apr 2017 17:46:51
Journal of Medical Microbiology 61
Bacteria in neonatal gastric aspirates
RESULTS
Samples analysed in this study were obtained as surplus to
the NGA routinely sent to the Department of Medical
Microbiology, Royal London Hospital, Barts and The London
NHS Trust, during the period August 2007 to February 2009.
Samples obtained from 121 (50 %) vaginal deliveries and 119
(50 %) Caesareans were collected the following day from the
Department of Medical Microbiology. Demographical and
clinical data were obtained from the Standardized Electronic
Neonatal Database (www.neonatal.org.uk/send) and from the
maternity notes (Table 1). Also characteristics of the sample
collection as well as Gram stain and culture results, were
obtained from the Department of Medical Microbiology
database (Tables 1, 2 and 3).
Demographical and clinical data in the vaginal and
Caesarean groups were generally homogeneous with some
exceptions inherent to Caesarean delivery, such as higher
maternal age and multiple deliveries. Also, a higher
number of women from black ethnic origin who delivered
by Caesarean was observed. The time for collection of the
sample was also significantly higher in vaginal deliveries.
When comparing bacteria positive using Gram stain and
culture, the only significant difference was observed for
Gram-positive rods, which were of higher prevalence in
vaginal deliveries (this possibly refers to Lactobacillus spp.
and Gardnerella vaginalis as observed with the molecular
methods).
The samples of NGA were analysed for bacterial content
using two different molecular techniques based upon the
small-subunit (16S) rRNA bacterial gene. With the clone
analysis-PCR, 77 of 240 (32 %) samples were positive. All
the samples were subjected to cloning; however, due to
a technique-sensitive protocol and limited amount of
the samples, bacteria were identified only from 52 of the
Table 1. Participant’s demographic and clinical data arranged by mode of delivery
Total [n5240]
Demographic data
Mode of delivery
Vaginal [n5121]
Neonate’s gender[1]*
Female (n) (%)
Male (n) (%)
Maternal age
Mean age±SD (range)
Maternal ethnic origin
Asian–Bangladeshi/Indian/Pakistani (n) (%)
Asian–other (n) (%)
Black African/Caribbean/other (n) (%)
White British/Irish/other (n) (%)
Other–mixed (n) (%)
Clinical characteristics
Multiple deliveries (n) (%)
Primiparous (n) (%)
Previous complications (n) (%)
In labour (n) (%)
Ruptured membranes (n) (%)
Prolonged ROM (.24 h) (n) (%)
Preterm birth (n) (%)
Mean gestational age (weeks) ±SD
Low birth weight (n) (%)
Mean birth weight (g) ±SD
IAP (n) (%)
Smoking during pregnancy (n) (%)
Sample collection
At the maternal ward (no neonatal ward)
Median time for collection (h) (range)
Samples collected within 12 h of birth (n) (%)
104 (43)
135 (56)
29±6 (17–46)[3]
117
16
30
68
8
(49)
(7)
(13)
(28)
(3)
19 (8)
118 (54)[23]
63 (56)[128]
–
–
39 (22)[61]
140 (58)
35±5
131 (55)
2322±962
35 (17)[39]
16 (8)[50]
58 (24)[2]
4.2 (0.2–23.9)[2]
190 (79)[1]
52 (43)
69 (57)
27±5 (17–42)[1]
61
9
8
39
4
(50)
(7)
(7)
(32)
(3)
4 (3)
53 (49)[13]
28 (51)[66]
–
–
25 (28)[32]
64 (53)
35±5
60 (50)
2397±948
15 (15)[23]
10 (11)[32]
32 (27)[1]
4.7 (0.2–23.9)[1]
92 (76)
P value
Caesarean [n5119]
52 (44)
66 (55)
30±6 (17–46)[2]
56
7
22
29
4
(47)
(6)
(18)
(24)
(3)
15 (13)
65 (60)[10]
35 (61)[62]
57 (53)[11]
46 (51)[29]
14 (16)[29]
76 (64)
35±4
71 (60)
2245±974
20 (19)[16]
6 (6)[18]
26 (22)[1]
3.7 (0.3–23.1)[1]
98 (83)[1]
P ,0.05D
P ,0.05d
P ,0.05d
P ,0.05§
[n]
, Data not known for n number of cases; –, data only analysed for Caesarean cases.
*There was one Caesarean case where the neonate’s gender was unknown.
DP value determined using an independent t-test.
dP value determined using Fisher’s exact test.
§P value determined using the Mann–Whitney U test.
http://jmm.sgmjournals.org
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sat, 29 Apr 2017 17:46:51
33
C. Gonzales-Marin and others
Table 2. Gram staining results related to the samples of NGA analysed in this study
Gram stain result*
Total (n) (%) [n5240]
Mode of delivery
Vaginal [n5121]
Total positive (n) (%)
G(+) cocci (total)
G(+) cocci pairs
G(+) cocci chains
G(+) cocci clusters
G(+) rods
G(2) cocci
G(2) rods
G(2) bacillus
62
43
5
10
4
14
3
20
2
(26)
(18)
(2)
(4)
(2)
(6)
(1)
(8)
(1)
38
27
3
6
4
12
1
13
1
P value
Caesarean [n5119]
(31)
(22)
(2)
(5)
(3)
(10)
(1)
(11)
(1)
24
16
2
4
0
2
2
7
1
(20)
(13)
(2)
(3)
(0)
(2)
(2)
(6)
(1)
P ,0.05D
*Data obtained from the Department of Medical Microbiology database, Royal London Hospital, Barts and The London NHS Trust.
DP value determined using Fisher’s exact test.
PCR-positive samples (68 %) successfully. It is possible that
some inhibitors might have been present in the samples or
were acquired during one of the steps involved in cloning. A
total of 932 clones (10 to 47 clones per sample) were selected
and sequenced for identification purposes. Forty-two species
were detected using this approach. With the DGGE analysis,
227 samples were investigated. Bands (n5372) were cut
from the polyacrylamide gels for sequencing. A total of 92
(41 %) of these were positive, detecting a total of 37 species
in the 88 positive samples (96 %) that were successfully
characterized with this technique.
The results obtained from both techniques were analysed as
a combined molecular approach in order to determine the
microbiology of NGA. A total of 59 % per cent (142/240)
of the samples were negative for bacterial detection with
the molecular techniques used. From the 98 (41 %) positive
samples, 80 (82 %) samples were characterized and the
total number of species identified was 51. The mean (±SD)
number of species per sample characterized was 3.7±3.0
for the positive samples, containing up to 15 species in one
of the samples. The most prevalent species were significantly higher in vaginal deliveries compared to Caesarean
Table 3. Culture results related to the samples of NGA analysed in this study
Total (n) (%) [n5240]
Culture result*
Mode of delivery
Vaginal [n5121]
Total positive (n) (%)
‘Skin flora’
GBS
a-Haemolytic streptococcus (including viridans
streptococcus)
Coagulase-negative staphylococcus
‘Anaerobes’
Coliform group bacteria
E. coli
Candida albicans
‘Mouth flora’
Streptococcus spp. (Streptococcus bovis)
Corynebacterium spp.
Enterococcus spp.
Klebsiella pneumoniae
Staphylococcus aureus
Bacillus spp.
Gardnerella vaginalis
Yeast
Caesarean [n5119]
60
11
9
7
(25)
(5)
(4)
(3)
38
5
4
2
(31)
(4)
(3)
(2)
22
6
5
5
(18)
(5)
(4)
(4)
7
6
5
4
4
3
2
2
2
2
1
1
1
1
(3)
(3)
(2)
(2)
(2)
(1)
(1)
(1)
(1)
(1)
(0)
(0)
(0)
(0)
6
4
5
3
4
3
1
2
2
1
1
1
1
1
(5)
(3)
(4)
(2)
(3)
(2)
(1)
(2)
(2)
(1)
(1)
(1)
(1)
(1)
1
2
0
1
0
0
1
0
0
1
0
0
0
0
(1)
(2)
(0)
(1)
(0)
(0)
(1)
(0)
(0)
(1)
(0)
(0)
(0)
(0)
*Data obtained from the Department of Medical Microbiology database, Royal London Hospital, Barts and The London NHS Trust.
34
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sat, 29 Apr 2017 17:46:51
Journal of Medical Microbiology 61
Bacteria in neonatal gastric aspirates
sections (Table 4). The percentages of the specific bacteria
per group were calculated out of the percentage of samples
able to be characterized; however, it was not possible to
carry this out with the statistical analysis where all the
samples were included.
Mode of delivery has previously been shown to determine
different patterns of initial colonizers in neonates; therefore,
the prevalence of the bacteria in Caesarean and vaginal
deliveries was analysed separately in order to evaluate other
possible clinical and non-clinical variables. Detection of
general bacteria and Caesarean delivery did not show any
significant association as determined with the molecular
approach; although associations were close to significance
for those with ruptured membranes at delivery (P50.073),
prolonged ROM (P50.058), those who received IAP
(P50.059) and women who smoked during pregnancy
(P50.062). However, significant associations were observed
for detection with Gram staining and culture techniques as
shown by the Fisher’s exact test (Table 5). On analysing
specific bacteria, few associations were observed. Presence of
the Streptococcus mitis group, was significantly associated
with IAP (P50.025), Streptococcus agalactiae (group B
streptococci, GBS) was significantly higher in the cases of
Caesarean patients who were in labour (P50.028), and
Pseudomonas spp. were associated with samples collected at
more than 12 h post-delivery (P50.015). Nevertheless, none
of these associations remained in the multivariate analyses for Gram staining, culture or molecular detection in
newborns delivered by Caesarean section.
In contrast, a different pattern of associations was observed
in the vaginal deliveries. Presence of general and specific
bacteria showed several associations with clinical and nonclinical variables using the bivariate analysis (Table 6). No
associations were observed between positive Gram staining
and culture analyses and the variables investigated in this
group. Table 7 summarizes the results of the multivariate
logistic regression; infection was significantly greater for
prolonged ROM in cases delivered vaginally.
DISCUSSION
In accordance with the UK’s National Standard Methods,
samples of NGA are investigated using minimal microbiology standard procedures to aid treatment and prognosis.
Gram staining allows a rapid detection of bacteria; while
microbial culture permits the growth and isolation of
micro-organisms, particularly focusing on a range of
bacteria historically associated with neonatal sepsis (HPA,
2004). Traditionally, cultured isolates have been limited to
a few species including E. coli, Haemophilus influenzae,
Listeria monocytogenes, GBS, and species of the genera
Klebsiella, Pseudomonas and Corynebacteria; and usually
only one single micro-organism per sample was detected
(Borderon et al., 1994). However, Jones et al. (2010)
recently demonstrated that culture analysis of NGA is
only positive when bacterial numbers exceeded 4.506105
http://jmm.sgmjournals.org
c.f.u. ml21. Routine analysis may thus underestimate the
number and richness (diversity) of positive samples. In the
current study, we have detected bacteria in 41 % of the
samples using PCR-based methods, each containing up
to 11 different species, compared to only 25–26 % with
routine detection and up to 2 species reported per sample.
Using molecular methods, Oue et al. (2009) have identified
20 species in samples of NGA (n542). Those given were also
observed in this study; however, we also identified another
31 species using the combined approach. Interestingly, many
of these bacteria have not been previously detected in NGA.
A combination of methods including a recombinant DNA
technique (clone analysis) and a profiling DGGE demonstrated improved bacterial detection. The procedures
required vary considerably for each method, which may
well explain the detection variations between both techniques (Table 4). In any case, the predominant bacteria
were identified by both methods while the differences were
observed with the less prevalent species.
Due to the invasiveness of the procedure, no healthy
neonates are normally sampled for this specimen. In theory,
gastric fluids in the healthy fetus should be sterile since it is
thought to develop in a bacteria-free environment and
bacterial colonization of the amniotic cavity is considered a
pathological finding (Romero et al., 2007). Nevertheless,
bacteria acquired at the time of delivery and through other
sources of contaminations may be present in the samples. It
was recently demonstrated that the initial microbiota in
healthy neonates reflects the mother’s vaginal bacteria if
delivered vaginally and the skin microbiota in Caesarean
sections (Dominguez-Bello et al., 2010). However, in the
aforementioned study (Dominguez-Bello et al., 2010) the
samples were obtained from relatively superficial sites (skin,
oral mucosa and nasopharyngeal aspirate) and before
removing the vernix caseosa, which would explain the
observations. Conversely, our data demonstrated that a high
percentage of samples obtained from the vaginal delivery
group were negative (49 %, 59/121) even when using highly
sensitive molecular techniques. Furthermore, the microbiota
our study has identified in the NGA resembles that described
elsewhere in studies of the amniotic fluid (Han et al., 2009;
Kotecha et al., 2004; Miralles et al., 2005) rather than solely
of a vaginal/skin origin. Therefore, this supports the assumption that NGA contains swallowed amniotic fluid.
Interestingly, in the amniotic fluid studies, an association
between intra-amniotic infection and clinical events, such as
early onset neonatal sepsis, chorioamnionitis and funisitis,
was demonstrated, which supports the potential value of
NGA as an alternative to investigate those infections
associated with APO, as suggested by Miralles et al.
(2005). However, the bacterial complexity observed in the
samples suggests the possibility that bacteria in general and
specific bacteria may have originated from other sources of
clinical and non-clinical origin or a combination of these.
Although the results indicate that the predominant bacteria have probably originated from the birth canal during
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sat, 29 Apr 2017 17:46:51
35
C. Gonzales-Marin and others
Table 4. Bacteria identified in NGA samples using clone analysis and DGGE in a combined molecular approach
Total (n) (%) [n5240]
Species
Total PCR detection positive (n) (%)
No. of samples with bacteria characterized (n) (%)*
Total samples included in study (n)D
S. pneumoniae/S. mitis/S. oralis/S. infantis§||
U. urealyticum/Ureaplasma parvum§||
E. coli||
Lactobacillus iners||
S. agalactiae||
Gardnerella vaginalis||
Sneathia sanguinegens/ ‘Sneathia amnionii’§||
Pseudomonas aeruginosa/Pseudomonas otitidis§||
Lactobacillus crispatus||
Streptococcus sanguinis/Streptococcus gordonii§||
Staphylococcus epidermidis||
Peptoniphilus spp.||
Veillonella montpellierensis (Veillonella atypica/
Veillonella parvula)§||
Pseudomonas fluorescens||
A. christensenii||
Streptococcus salivarius/Streptococcus vestibularis§||
Brevundimonas diminuta||
H. influenzae||
Prevotella bivia#
S. sinensis/S. parasanguinis§#
Peptostreptococcus anaerobius||
Sphingomonas spp.||
Streptococcus constellatus/Streptococcus intermedius/
Streptococcus anginosus§||
Granulicatella elegans||
Veillonella parvula/Veillonella dispar§||
Sphingopyxis alaskensis||
Massilia sp.#
Moraxella osloensis||
Propionibacterium acnes||
Prevotella sp.||
F. nucleatum||
Hyphomicrobium zavarzinii
Acinetobacter baumannii#
Corynebacterium riegelii
Bacteroides sp.
Staphylococcus sp.
Enterococcus faecalis
Lactobacillus gasseri#
Anaerococcus vaginalis
Faecalibacterium prausnitzii
Megasphaera sp.#
Fusobacterium necrophorum
Caldimonas taiwanensis
Silanimonas lenta
Bifidobacterium sp.
Atopobium minutum
Clostridium perfringens
Fusobacterium periodonticum
Klebsiella pneumoniae#
36
98 (41)
80 (82)
196d
27 (14)
18 (9)
14 (7)
13 (7)
11 (6)
10 (5)
9 (5)
8 (4)
8 (4)
7 (4)
6 (3)
6 (3)
5 (3)
Mode of delivery
P value
Vaginal [n5121]
Caesarean [n5119]
62 (51)
51 (82)
100d
19 (19)
15 (15)
13 (13)
10 (10)
4 (4)
8 (8)
8 (8)
6 (6)
5 (5)
5 (5)
6 (6)
3 (3)
3 (3)
36 (30)
29 (81)
96d
8 (8)
3 (3)
1 (1)
3 (3)
7 (7)
2 (2)
1 (1)
2 (2)
3 (3)
2 (2)
0 (0)
3 (3)
2 (2)
5
4
4
4
4
3
3
3
3
3
(3)
(2)
(2)
(2)
(2)
(2)
(2)
(2)
(2)
(2)
3
4
3
2
2
2
2
2
2
1
(3)
(4)
(3)
(2)
(2)
(2)
(2)
(2)
(2)
(1)
2
0
1
2
2
1
1
1
1
2
(2)
(0)
(1)
(2)
(2)
(1)
(1)
(1)
(1)
(2)
2
2
2
2
2
2
2
2
2
2
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
1
1
1
1
1
2
2
2
2
0
1
1
1
1
1
1
1
1
1
1
1
0
0
0
0
0
(1)
(1)
(1)
(1)
(1)
(2)
(2)
(2)
(2)
(0)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(0)
(0)
(0)
(0)
(0)
1
1
1
1
1
0
0
0
0
2
0
0
0
0
0
0
0
0
0
0
0
1
1
1
1
1
(1)
(1)
(1)
(1)
(1)
(0)
(0)
(0)
(0)
(2)
(0)
(0)
(0)
(0)
(0)
(0)
(0)
(0)
(0)
(0)
(0)
(1)
(1)
(1)
(1)
(1)
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sat, 29 Apr 2017 17:46:51
P ,0.05
P ,0.05
P ,0.05
P ,0.05
P ,0.05
Journal of Medical Microbiology 61
Bacteria in neonatal gastric aspirates
Table 4. cont.
Species
Proteus mirabilis#
Mycoplasma hominis#
Total (n) (%) [n5240]
1 (1)
1 (1)
Mode of delivery
Vaginal [n5121]
Caesarean [n5119]
0 (0)
0 (0)
1 (1)
1 (1)
P value
*Percentage of the samples that could be characterized in this study was calculated out of the number of PCR-positive samples.
DTotal number of samples studied was calculated as the percentage of samples characterized out of the total number of samples.
dPercentage of specific bacteria was calculated out of the total number of samples characterized.
§Species considered in a same group due to difficulty of differentiation by the 16S rRNA gene approach.
||Species detected with both molecular techniques.
Species identified with cloning analysis only.
#Species identified with DGGE approach only.
vaginal delivery and as a result of a prolonged ROM (odds
ratio55.7), most of the prevalent species described in vaginal
deliveries are also observed in Caesarean sections, exceptions
were Staphylococcus epidermidis and Aerococcus christensenii,
frequent colonizers of the female genito-urinary tract
(Collins et al., 1999). This observation supports the ascending route as the most common pathway of intrauterine
infection in which micro-organisms from the vaginal/
cervical tract gain access to the intrauterine cavity (Romero
et al., 2006); for example bacterial vaginosis has frequently
been associated with APO (Kimberlin & Andrews, 1998).
Another explanation, however, may be provided by previous
work wherein it was demonstrated that a suction-like effect
during uterine contractions may introduce vaginal fluid into
the amniotic cavity (Zervomanolakis et al., 2007). Gram
staining results also support the hypothesis since there was a
significant association between the presence of bacteria and
Caesarean patients in labour.
Our data confirm that prolonged ROM (in vaginal and
Caesarean deliveries) and ruptured membranes (in
Caesarean deliveries) determine the presence of bacteria
in the neonates as observed with both the Gram stain and
molecular data. It is well know that ROM in labour or not
in labour can result in increased rates of maternal and fetal
infection (Bryant-Greenwood & Millar, 2000; Romero
et al., 1991). Administration of IAP has been instituted as a
routine hospital procedure since the mid-1990s, particularly to decrease the incidence of early onset group B
streptococcal septicaemia (Ohlsson & Shah, 2009; Schrag
et al., 2000). In this study, no cases of elective Caesarean
received IAP; all the cases were emergency Caesareans. The
presence of S. agalactiae was close to significance with IAP
in Caesarean (P50.086) and also in vaginal deliveries. The
routine administration of IAP in women with a history of
infection with GBS may be an explanation as it is not
known if the bacteria are viable; also, IAP may disrupt
the perinatal microbiota, allowing antibiotic-resistant GBS
strains and other pathogenic bacteria to proliferate.
Streptococcus pneumoniae/Streptococcus mitis/Streptococcus
oralis/Streptococcus infantis were the most prevalent
http://jmm.sgmjournals.org
bacteria (27 %) in the samples and were associated with
IAP in Caesarean sections. Although viridans streptococci
(S. mitis and S. oralis) are of low pathogenicity, S. mitis has
been reported to cause early onset neonatal infection and
neonatal death as a result of its resistance to antibiotics
(Adams & Faix, 1994). High prevalence of the S. mitis
group in perinatal infections has not been reported before;
however, S. pneumoniae is one of the major bacterial
pathogens worldwide causing bacteraemia and community-acquired infections in infants (Selva et al., 2010).
Unfortunately, members of the S. mitis group are genetically closely related to each other and we were not able
to differentiate them in this study.
Ureaplasma urealyticum has been implicated in several
forms of APO (Waites et al., 2005). However, carriage in
the women’s genito-urinary tract does not necessarily
trigger an APO (Govender et al., 2009), nor does the
treatment of the infection alter the occurrence of
prematurity (Romero et al., 1989). High prevalence of U.
urealyticum in NGA from vaginal deliveries indicates this
pathogen is likely to be derived from the birth canal during
delivery as is present at low prevalence in the Caesarean
cases; although, it is also possible it was acquired from
contamination during the collection of the sample since U.
urealyticum, as well as Lactobacillus spp., Pseudomonas spp.,
Gardnerella vaginalis and Sneathia spp., were present at
significantly higher levels in vaginal deliveries collected on
the maternal wards. This may be due to the difficulty in
achieving sterile conditions in the maternal wards and
also because they usually take a longer time for collection.
Similarly, an association was shown between Pseudomonas
spp. and Caesarean samples collected at more than 12 h
post-delivery (P50.015). Therefore, greater care during
collection and transport of the samples should be advised.
The most common pathogens recognized as causing
nosocomial bloodstream infection in neonatal intensive
care units (E. coli, Staphylococcus spp., Klebsiella and
Pseudomonas spp.) were also identified in the samples. E.
coli was associated with prolonged ROM in vaginal
deliveries. This is one of the most common bacterial
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sat, 29 Apr 2017 17:46:51
37
38
0.415
0.319
species responsible for neonatal infections and is increasingly implicated in nosocomial outbreaks due to multidrug-resistant strains, and should be investigated further.
55 (66)
52 (63)
Another interesting observation was the association between Lactobacillus spp. and neonates delivered at term and
presenting a birth weight ¢2500 g in vaginal deliveries,
although not significant in the Caesarean group. Lactobacillus
spp. are usually observed as dominating the microbiota in
the vagina and cervix of healthy women, and may prevent
colonization of more pathogenic species (Boris & Barbés,
2000). Significant association was observed between
Lactobacillus crispatus and full-term birth. This is consistent
with the previous observation that Lactobacillus crispatus
promotes the stability of the normal vaginal inhabitants,
which may prevent further complications; however,
Lactobacillus iners has been shown to be less protective,
predisposing to the occurrence of abnormal vaginal microbiota and pregnancy complications (Verstraelen et al., 2009).
However, an association between Lactobacillus spp. and
collection of the samples at the maternal ward was also
observed, and none of these associations were maintained in a
multivariate analysis.
, Data not known for n number of cases; (%), percentage within group (Gram stain, culture or molecular).
*P values determined using Fisher’s exact test.
DSignificant difference as determined by P ,0.05.
dPrevalence associated with no observation of the clinical variable.
[n]
0.006Dd
0.153
70 (74)
63 (66)
6 (25)
8 (33)
,0.001Dd
0.005Dd
8 (37)
10 (46)
68 (70)
61 (63)
21 (58)
19 (53)
0.029D
6 (7)
0.722
9 (10)
5 (21)
0.154
3 (14)
11 (11)
8 (22)
0.303
0.073
37 (49)
27 (44)
20 (61)
19 (66)
0.305
1.000
46 (51)
37 (51)
11 (65)
9 (53)
0.012D
0.038D
41 (47)
32 (45)
Positive (n536) Negative (n583) P values*
P values*
Positive (n522) Negative (n597)
P values*
Negative (n595)
Positive (n524)
Gram stain
16 (80)
14 (74)
In labour (n) (%)
Ruptured membranes
(n) (%) [29]
Prolonged ROM
(.24 h) (n) (%)
Preterm birth (n) (%)
Low birth weight (n) (%)
[11]
Molecular
Culture
Caesarean deliveries (n5119)
Clinical characteristic
Table 5. Associations between clinical characteristics of the pregnancy and Gram stain, culture and molecular results in samples delivered by Caesarean section
C. Gonzales-Marin and others
Finally, secondary pathways for intrauterine infection
have also been proposed. It has been suggested that an
association exists between periodontal disease and APO
(Offenbacher et al., 1996). Although in low prevalence,
bacteria from a possible oral origin were detected in the
samples. Granulicatella elegans and Streptococcus sinensis/
Streptococcus parasanguinis require particular attention
since they are known to become blood borne and have
been associated with infective endocarditis (Faibis et al.,
2008; Ohara-Nemoto et al., 2005). Also, Fusobacterium
nucleatum, a recognized periodontal pathogen, was observed in the samples. In fact, this is the most common
species isolated from amniotic fluid in women with preterm labour and intact membranes (Hill, 1998), and it has
been associated with previous pregnancy complications
(Bearfield et al., 2002). Recently, it has been reported as the
aetiological cause of a stillbirth (Han et al., 2010). Due to
the strong evidence indicating the possible involvement of
the maternal oral bacteria in the complication, our group
have also evaluated the presence of known periodontal
pathogens in NGA, and demonstrated F. nucleatum present
in samples of NGA to have more likely originated from a
woman’s oral cavity rather than vagina (Gonzales-Marin
et al., 2011). Further studies are necessary to investigate
this potential pathogen’s involvement in APO.
Conclusions
Although no associations were observed with preterm
birth, low birth weight or previous complications, identification of the bacteria in the samples may indicate
colonization and therefore identify a potential infectious
agent. It is likely that the bacteria causing an infection in
the fetus and/or neonate may have originated from a
maternal site harbouring a complex community (e.g. the
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sat, 29 Apr 2017 17:46:51
Journal of Medical Microbiology 61
http://jmm.sgmjournals.org
Table 6. Associations between molecular detection of general and specific bacteria in NGA and clinical characteristics of the pregnancy in vaginal deliveries
Clinical
characteristic
Vaginal deliveries (n5121)
Molecular
Species specific (P values)*
Positive Negative P value* Streptococcus Ureaplasma Lactobacillus Lactobacillus Lactobacillus Escherichia
(n562) (n559)
mitis
spp.
spp.
iners
crispatus
coli
Previous
complications
(n) (%)[66]
Prolonged
ROM (.24 h)
(n) (%)
Preterm
birth (n) (%)
Low birth
weight (n) (%)
IAP (n) (%)[23]
Samples
collected at
the maternal
ward (n) (%)[1]
Streptococcus Pseudomonas Gardnerella Sneathia
agalactiae
spp.
vaginalis
spp.
13 (50)
15 (52)
1.000
0.469
1.000
0.611
0.611
–
0.023Dd
1.000
0.111
0.491
1.000
18 (29)
7 (12)
0.025D
0.204
0.213
0.313
0.313
–
0.017D
0.556
1.000
0.281
0.344
32 (52)
32 (54)
0.856
0.328
0.593
0.020Dd
0.188
0.021Dd
0.379
0.342
1.000
0.146
0.721
28 (45)
32 (54)
0.365
0.133
0.789
0.008Dd
0.095
0.057
0.075
0.619
1.000
0.061
0.491
6 (13)
22 (36)
9 (18)
10 (17)
0.582
0.038D
0.688
0.582
0.604
0.004D
0.572
0.002D
0.491
0.022D
1.000
0.018D
1.000
0.105
0.060
0.289
0.060
0.031D
0.153
,0.001D
0.586
0.031D
, Data not known for n number of cases; (%), percentage within group (molecular).
*P values determined using Fisher’s exact test.
DSignificant difference as determined by P,0.005.
dPrevalence associated with no observation of the clinical variable.
39
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sat, 29 Apr 2017 17:46:51
Bacteria in neonatal gastric aspirates
[n]
C. Gonzales-Marin and others
Table 7. Multivariate logistic regression for detection of broad-range bacteria in NGA obtained from vaginal deliveries
Variable
Detection of bacteria
95 % confidence interval
Previous complications
Prolonged ROM
PTB
LBW
IAP
Maternal ward
P value
OR
Lower
Upper
0.263
0.036*
0.878
0.480
0.342
0.955
2.11
5.71
1.24
0.38
0.46
0.94
0.57
1.12
0.08
0.03
,0.01
0.11
7.81
29.02
20.15
5.56
2.30
7.79
LBW, Low birth weight; OR, odds ratio; PTB, preterm birth.
*Significant difference as determined by P,0.05.
Borderon, E., Desroches, A., Tescher, M., Bondeux, D., Chillou, C. &
Borderon, J. C. (1994). Value of examination of the gastric aspirate
vagina, the gut or the oral cavity), and that more than
one single pathogen may translocate. As observed, DGGE
profiling may provide a very useful tool for potential
clinical use due to its sensitivity in detecting the most
prevalent bacteria, ease of use and relatively low cost. In
this study, several clinical variables (mode of delivery,
prolonged ROM, labour and IAP) have been shown to
influence the presence of bacteria in neonates; however,
bacteria possibly acquired during collection are also found
in the samples. Under a more strict collection protocol, the
study of NGA should provide a new insight into the
investigation of the species of bacteria possibly associated
with APO. Specifically, the effect of antibiotic prophylaxis and antibiotic resistance in clinical trials should be
evaluated.
membranes: their preterm premature rupture. Biol Reprod 63, 1575–
1579.
ACKNOWLEDGEMENTS
Cole, J. R., Chai, B., Farris, R. J., Wang, Q., Kulam-Syed-Mohideen,
A. S., McGarrell, D. M., Bandela, A. M., Cardenas, E., Garrity, G. M. &
Tiedje, J. M. (2007). The ribosomal database project (RDP-II):
We would like to thank the UNESCO-L’OREAL partnership for
supporting this study through the International Fellowship awarded
to C. G-M. Consumables for this study were supported with
departmental funds. Special thanks to the staff of the hospital’s
laboratory, maternity and neonatal wards at Barts and The London
NHS Trust for their support, especially to Dr Mark Wilks and Miss
Anita Sanghi.
for the diagnosis of neonatal infection. Biol Neonate 65, 353–366.
Boris, S. & Barbés, C. (2000). Role played by lactobacilli in
controlling the population of vaginal pathogens. Microbes Infect 2,
543–546.
Bryant-Greenwood, G. D. & Millar, L. K. (2000). Human fetal
Carroll, S. G., Papaioannou, S., Ntumazah, I. L., Philpott-Howard, J.
& Nicolaides, K. H. (1996). Lower genital tract swabs in the predic-
tion of intrauterine infection in preterm prelabour rupture of the
membranes. Br J Obstet Gynaecol 103, 54–59.
Chen, T., Yu, W.-H., Izard, J., Baranova, O. V., Lakshmanan, A. &
Dewhirst, F. E. (2010). The Human Oral Microbiome Database: a web
accessible resource for investigating oral microbe taxonomic and
genomic information. Database 2010, baq013.
introducing myRDP space and quality controlled public data. Nucleic
Acids Res 35 (Database issue), D169–D172.
Collins, M. D., Jovita, M. R., Hutson, R. A., Ohlén, M. & Falsen, E.
(1999). Aerococcus christensenii sp. nov., from the human vagina. Int J
Syst Bacteriol 49, 1125–1128.
Dominguez-Bello, M. G., Costello, E. K., Contreras, M., Magris, M.,
Hidalgo, G., Fierer, N. & Knight, R. (2010). Delivery mode shapes the
REFERENCES
acquisition and structure of the initial microbiota across multiple
body habitats in newborns. Proc Natl Acad Sci U S A 107, 11971–
11975.
Aas, J. A., Paster, B. J., Stokes, L. N., Olsen, I. & Dewhirst, F. E.
(2005). Defining the normal bacterial flora of the oral cavity. J Clin
Evans, M. E., Schaffner, W., Federspiel, C. F., Cotton, R. B., McKee,
K. T., Jr & Stratton, C. W. (1988). Sensitivity, specificity, and predictive
Microbiol 43, 5721–5732.
value of body surface cultures in a neonatal intensive care unit. JAMA
259, 248–252.
Adams, J. T. & Faix, R. G. (1994). Streptococcus mitis infection in
newborns. J Perinatol 14, 473–478.
Ampe, F., Ben Omar, N., Moizan, C., Wacher, C. & Guyot, J. P. (1999).
Polyphasic study of the spatial distribution of microorganisms in
Mexican pozol, a fermented maize dough, demonstrates the need for
cultivation-independent methods to investigate traditional fermentations. Appl Environ Microbiol 65, 5464–5473.
Faibis, F., Mihaila, L., Perna, S., Lefort, J. F., Demachy, M. C., Le
Flèche-Matéos, A. & Bouvet, A. (2008). Streptococcus sinensis: an
emerging agent of infective endocarditis. J Med Microbiol 57, 528–531.
Frank, J. A., Reich, C. I., Sharma, S., Weisbaum, J. S., Wilson, B. A. &
Olsen, G. J. (2008). Critical evaluation of two primers commonly
used for amplification of bacterial 16S rRNA genes. Appl Environ
Microbiol 74, 2461–2470.
Bearfield, C., Davenport, E. S., Sivapathasundaram, V. & Allaker,
R. P. (2002). Possible association between amniotic fluid micro-
Gafan, G. P. & Spratt, D. A. (2005). Denaturing gradient gel
organism infection and microflora in the mouth. BJOG 109, 527–533.
electrophoresis gel expansion (DGGEGE)–an attempt to resolve the
40
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sat, 29 Apr 2017 17:46:51
Journal of Medical Microbiology 61
Bacteria in neonatal gastric aspirates
limitations of co-migration in the DGGE of complex polymicrobial
communities. FEMS Microbiol Lett 253, 303–307.
Ohara-Nemoto, Y., Kishi, K., Satho, M., Tajika, S., Sasaki, M.,
Namioka, A. & Kimura, S. (2005). Infective endocarditis caused by
Gafan, G. P., Lucas, V. S., Roberts, G. J., Petrie, A., Wilson, M. &
Spratt, D. A. (2004). Prevalence of periodontal pathogens in dental
Granulicatella elegans originating in the oral cavity. J Clin Microbiol
43, 1405–1407.
plaque of children. J Clin Microbiol 42, 4141–4146.
Gonzales-Marin, C., Spratt, D. A., Millar, M. R., Simmonds, M.,
Kempley, S. T. & Allaker, R. P. (2011). Levels of periodontal
Ohlsson, A. & Shah, V. S. (2009). Intrapartum antibiotics for known
maternal group B streptococcal colonization. Cochrane Database Syst
Rev, CD007467.
pathogens in neonatal gastric aspirates and possible maternal sites of
origin. Mol Oral Microbiol 26, 277–290.
Oue, S., Hiroi, M., Ogawa, S., Hira, S., Hasegawa, M., Yamaoka, S.,
Yasui, M., Tamai, H. & Ogihara, T. (2009). Association of gastric fluid
Govender, S., Theron, G. B., Odendaal, H. J. & Chalkley, L. J. (2009).
microbes at birth with severe bronchopulmonary dysplasia. Arch Dis
Child Fetal Neonatal Ed 94, F17–F22.
Prevalence of genital mycoplasmas, ureaplasmas and chlamydia in
pregnancy. J Obstet Gynaecol 29, 698–701.
Patel, J. B. (2001). 16S rRNA gene sequencing for bacterial pathogen
Han, Y. W., Shen, T., Chung, P., Buhimschi, I. A. & Buhimschi, C. S.
(2009). Uncultivated bacteria as etiologic agents of intra-amniotic
Petrou, S. (2003). Economic consequences of preterm birth and low
identification in the clinical laboratory. Mol Diagn 6, 313–321.
inflammation leading to preterm birth. J Clin Microbiol 47, 38–47.
birthweight. BJOG 110 (Suppl. 20), 17–23.
Han, Y. W., Fardini, Y., Chen, C., Iacampo, K. G., Peraino, V. A.,
Shamonki, J. M. & Redline, R. W. (2010). Term stillbirth caused by
Puri, J., Revathi, G., Faridi, M. M., Talwar, V., Kumar, A. & Parkash, B.
(1995). Role of body surface cultures in prediction of sepsis in a
oral Fusobacterium nucleatum. Obstet Gynecol 115 (Suppl.), 442–445.
neonatal intensive care unit. Ann Trop Paediatr 15, 307–311.
HPA (2004). Investigation of Gastric Aspirates and Infection Screen
Romero, R., Mazor, M., Oyarzun, E., Sirtori, M., Wu, Y. K. & Hobbins,
J. C. (1989). Is genital colonization with Mycoplasma hominis or
Swabs from Neonates, national standard method, BSOP 23, issue 4.
London: Health Protection Agency.
Hill, G. B. (1998). Preterm birth: associations with genital and possibly
oral microflora. Ann Periodontol 3, 222–232.
Jones, V., Wilks, M., Johnson, G., Warwick, S., Hennessey, E.,
Kempley, S. & Millar, M. (2010). The use of molecular techniques for
bacterial detection in the analysis of gastric aspirates collected from
infants on the first day of life. Early Hum Dev 86, 167–170.
Kimberlin, D. F. & Andrews, W. W. (1998). Bacterial vaginosis:
association with adverse pregnancy outcome. Semin Perinatol 22,
242–250.
Kotecha, S., Hodge, R., Schaber, J. A., Miralles, R., Silverman, M. &
Grant, W. D. (2004). Pulmonary Ureaplasma urealyticum is associated
with the development of acute lung inflammation and chronic lung
disease in preterm infants. Pediatr Res 55, 61–68.
Leibovich, M., Gale, R. & Slater, P. E. (1987). Usefulness of the gastric
aspirate examination in the diagnosis of neonatal infection. Trop
Geogr Med 39, 15–17.
McCormick, M. C. (1985). The contribution of low birth weight to
infant mortality and childhood morbidity. N Engl J Med 312, 82–90.
Miralles, R., Hodge, R., McParland, P. C., Field, D. J., Bell, S. C.,
Taylor, D. J., Grant, W. D. & Kotecha, S. (2005). Relationship between
antenatal inflammation and antenatal infection identified by detection of microbial genes by polymerase chain reaction. Pediatr Res
57, 570–577.
Ureaplasma urealyticum associated with prematurity/low birth
weight? Obstet Gynecol 73, 532–536.
Romero, R., Ghidini, A., Mazor, M. & Behnke, E. (1991). Microbial
invasion of the amniotic cavity in premature rupture of membranes.
Clin Obstet Gynecol 34, 769–778.
Romero, R., Espinoza, J., Kusanovic, J. P., Gotsch, F., Hassan, S.,
Erez, O., Chaiworapongsa, T. & Mazor, M. (2006). The preterm
parturition syndrome. BJOG 113 (Suppl. 3), 17–42.
Romero, R., Espinoza, J., Gonçalves, L. F., Kusanovic, J. P., Friel, L. &
Hassan, S. (2007). The role of inflammation and infection in preterm
birth. Semin Reprod Med 25, 21–39.
Schrag, S. J., Zywicki, S., Farley, M. M., Reingold, A. L., Harrison, L. H.,
Lefkowitz, L. B., Hadler, J. L., Danila, R., Cieslak, P. R. & Schuchat, A.
(2000). Group B streptococcal disease in the era of intrapartum
antibiotic prophylaxis. N Engl J Med 342, 15–20.
Selva, L., Esteva, C., Gené, A., De Sevilla, M. F., Hernandez-Bou, S.
& Muñoz-Almagro, C. (2010). Direct detection of Streptococcus
pneumoniae in positive blood cultures by real-time polymerase chain
reaction. Diagn Microbiol Infect Dis 66, 204–206.
Verstraelen, H., Verhelst, R., Claeys, G., De Backer, E., Temmerman, M.
& Vaneechoutte, M. (2009). Longitudinal analysis of the vaginal
microflora in pregnancy suggests that L. crispatus promotes the stability
of the normal vaginal microflora and that L. gasseri and/or L. iners are
more conducive to the occurrence of abnormal vaginal microflora. BMC
Microbiol 9, 116.
Offenbacher, S., Katz, V., Fertik, G., Collins, J., Boyd, D., Maynor, G.,
McKaig, R. & Beck, J. (1996). Periodontal infection as a possible risk
Waites, K. B., Katz, B. & Schelonka, R. L. (2005). Mycoplasmas and
factor for preterm low birth weight. J Periodontol 67 (Suppl.), 1103–
1113.
ONS (2005). Live births, Stillbirths and Infant Deaths by ONS Cause
Zervomanolakis, I., Ott, H. W., Hadziomerovic, D., Mattle, V., Seeber,
B. E., Virgolini, I., Heute, D., Kissler, S., Leyendecker, G. & Wildt, L.
(2007). Physiology of upward transport in the human female genital
Groups. Newport: Office for National Statistics.
tract. Ann N Y Acad Sci 1101, 1–20.
http://jmm.sgmjournals.org
ureaplasmas as neonatal pathogens. Clin Microbiol Rev 18, 757–789.
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sat, 29 Apr 2017 17:46:51
41