zyxwvutsrqponmlkj
zyxwvu
zyxwvutsrqpon
zyxw
zyxw
zyxw
Acia Obsiei Gynecol Scand 1996: 7s: 657-661 Copyrrghr B ACro Obsrer Gynecol Scand 1996
Printed in Denmork - a" righls reserved
Acta Obstetricia et
Gynecologica Scandinavica
ISSN 0001-6349
ORIGINAL ARTICLE
Stillbirths and maternal antibodies to
Chlamydia trachomatis. A new EIA test for
serology
MARJALEENA KOSKINIEMI', PlRKKO AMMALA', ALE NARVANEN5, PEKKA sAIKKU4, MARIA
SODERLUND', PENTTI KOSKELA4,
PEKKA LAURILA3 AND ANTTIVAHERI'
From the Departments of 'Virology, 20bstetrics and Gynecology and 3Pathology, University of Helsinki, 4National Public Health
Institute, Oulu and 'Labsystems Research Laboratories, Helsinki, Finland
Acta Obstet Gynecol Scand 1996; 75: 657-661.0 Acta Obstet Gynecol Scand 1996
Objective. The relationship between stillbirths and infections during pregnancy.
Methods. Antibodies to 13 different viruses and to Mycoplasma pneumoniae, Treponema pal-
lidum, Toxoplasma gondii, and Chlamydia trachomatis, psittaci, and pneumoniae were stud-
ied from sera of 42 mothers with stillbirth and of matched case-control pregnancies.
Results. Elevated antibody levels to Chlamydia trachomatis and human parvovirus B19 were
observed in seven mothers with stillbirth as compared with one in matched case controls @-
0.03 and p<0.06, Fisher's exact one- and two-tailed~values). By a novel EIA test for C. tra-
rhomatis, based on synthetic polypeptide, 15 positive cases were detected in patients com-
pared to seven in controls. Placental calcifications and fetal growth retardation appeared more
often @<0.001 and pe0.05, respectively) in the C. trachomatis- associated cases than in the
others. There was no significant relation between stillbirth and the other antibodies.
Conclusions. We suggest C. trachomatis as a hitherto unrecognized possible cause of still-
births besides human parvovirus B19. A new EIA test for C. trachomatis, one covering all
strains, is promising and makes the future assays more convenient, and may thus make it pos-
sible to reduce the rate of stillbirths.
Key words: Chlamydia trachomatis; parvovirus; stillbirths; viral antibodies
Submitted 4 Septernbec 1995
Accepted II Januaq 1996
The causes of stillbirths often remain unknown,
even after autopsy (1,2). Descriptive diagnoses,
such as immature birth, growth retardation, and
chorioamnionitis, are suggestive of an infectious
etiology, but the microbes responsible have been
largely elusive.
Maternal infection with human parvovirus B 19
may lead to fetal hydrops and death (3), but this virus
has been assigned a minor role as a cause of stillbirths
zyxw
to >20% of cervical specimens obtained from preg-
nant women and is accepted as acause of neonatal con-
junctivitis and pneumonia ( 5 ) . The association with
premature birth has been reported though evidence for
that is still mainly indirect (6,7). C. trachomutis may
be transmitted in utero, but only one case leading to
fetal death has been reported (8,9).C. psittuci has been
cultured from fetal animal tissues and ovine recurrent
abortion is a widely known disease that may be trans-
zyxwv
zyx
(4). Chlamydia trachomatishas been cultured from 2 mitted to humans and cause abortion (10). We report
a relationship between stillbirths and elevated mater-
Abbreviations:
nal antibodies to human parvovirus B 19 or to C. tra-
IF: immunofluorescence; HI: hemagglutination inhibition; EIA: chomutis, comparing the results with those of matched
enzyme immunoassay; EB: elementary bodies; SFD: small for case controls, and present data obtained by a new EIA
' gestational age. test for C.trachornatis antibodies.
0Acta Obstet Gynecol Scand 75 (1996)
, 658 zyxwvuts
zyxwvutsrqpo
M. Koskiniemi et al.
Materials and methods
(13). The presence of antibodies was regarded as
zy
positive with a cut off 20.300 and elevated if 21.300
Patients. Between April 1, 1988, and January 31, at 405 nm at a serum dilution of 1:lOO. Alkaline
1990, all mothers who suffered stillbirths were en- phosphatase-conjugate was used to detect the bound
rolled in this study from the greater Helsinki area, parvovirus IgG or IgM antibodies. Commercial kits
with 1.2 million inhabitants. The mothers were re- were used for rubella virus (Rubaset@EIA-G and
ferred to Departments I and I1 of Obstetrics and EIA-M, Orion Diagnostica, Helsinki, Finland) and
Gynecology, University of Helsinki, Finland. Dur- I: gondii (Toxoplasma gondii IgG and IgM EIA kit;
ing the 22-month period, 44 women had stillbirths Labsystems).
(fetal deaths after the 21st gestational week). Two For chlamydia1 antibodies, a micro-IF method
mothers were excluded because no antenatal sera measuring IgG and IgM antibodies specific to C.
zyxwvu
were obtained; one delivered an anencephalic fetus trachomatis, C. pneumoniae, and C. psittaci was
and the other had a grave cord complication. Thus used (14). The elementary bodies (EB) used as anti-
the study group included 42 mothers (95% of all) gen were for the c. trachomatis pooled immuno-
and all had serum specimens taken at weeks 8 to 36 types CJHI, GFK, and BED, and for the C. pneumo-
of pregnancy. Nine of the 42 cases (21.4%) were niae strain TWAR (all from Washington Research
twins. Forty-six fetuses were born dead and five of Foundation, Seattle, WA), and for C. psittaci pooled
the twins were alive. Each mother had a case-con- OA and 6BC strains (Slovak Academy of Sciences,
trol, matched by age (k4 yrs, mean age 29.5 and Bratislava). All IgM-positive sera were retested af-
29.4 years, respectively), number of pregnancies ter treatment with Gullsorb (Gull Laboratories, Salt
(2.2 and 2.3, respectively), number of deliveries Lake City, UT). The presence of antibodies at titers
(0.9 and 1.O, respectively), gestational age (the same 232 (reciprocal dilution) was regarded as elevated,
trimester), date of specimen (f10 days), and home and the presence of IgM (216) antibodies was con-
city (same city). Forty-three of the 46 dead fetuses sidered diagnostic of an acute infection.
were autopsied, placentas and fetal tissues were Antibodies to C. trachomatis were also measured
studied histopathologically (P.L.), and the mothers using a new EIA test specific for C. trachomatis
were examined thoroughly (P.A.) in the hope of de- species (Labsystems, Helsinki, Finland, unpub-
termining the cause of stillbirth. lished). A synthetic peptide of 12 amino acids de-
Protocol. Maternal sera were studied for antibod- rived from the VP4 region of C. trachomatis
ies against 13 different viruses, Mycoplasma pneu- MOMP strain L2 was used as an antigen to coat
moniae, Treponema pallidum, Toxoplasma gondii, polystyrene microtiter plates (Labsystems, Hel-
and Chlamydia trachomatis, psittaci, and pneumo- sinki). In solid-phase EIA 50 pL serum was diluted
niae. Syphilis serology is done routinely (TPHA- 1:4 and peroxidase-conjugated anti-human IgG was
test, Fujirebio, Inc., Tokyo, Japan) and all the moth- used to detect the bound C. trachomatis IgG anti-
ers were negative. An immunofluorescence (IF) test bodies (Niirvanen et al. manuscript in preparation).
was carried out for Puumala virus (a hantavirus All microbiological studies were carried out at
causing Nephropathia epidemica), a hemagglutina- the Department of Virology, University of Helsinki,
tion inhibition (HI) assay for Sindbis virus (an al- except the serology for Treponemapallidum and ru-
phavirus causing Pogosta disease with a rash mim- bella virus, which was done at the National Public
zy
icking rubella), and a solid-phase enzyme immu- Health Institute, Oulu, and the EIA test for C. tra-
noassay (EIA) of IgG specific for adeno, influenza chomatis at Labsystems, Helsinki, Finland.
A and B (1 l), HSV-1 and HSV-2, CMV, parainflu- Analysis. The test results for the mothers who had
enza 1, respiratory syncytial and varicella zoster vi- stillbirths were compared with those for the
ruses (commercial antigens, purified for EIA by Vir- matched case controls. The data were analyzed us-
ion@).The M. pneumoniae antigen was a gift from ing Fisher's exact test, and the chi-square test. The p
M. Kleemola (12). Antigens at a protein concentra- values reported are one- and two-tailed. In this anal-
tion of 1-5 pg/mL in buffers were used to coat poly- ysis, we used the Epistat 5 program (CDCWHO).
styrene microtiter plates (NUNC, Copenhagen, The results obtained with the new EIA test for C.
Denmark and Labsystems, Helsinki, Finland). Four- trachomatis were compared with those of the micro-
fold dilutions of sera, starting from 1:200, were IF method, and specificity and sensitivity were cal-
zy
studied and end-point titers were estimated as recip- culated.
rocals of the highest serum dilution giving an ab-
zyxwvutsrq
sorbance value at 405 nm of at least the mean +2 Results
s.d. units of negative sera and an absorbance value
of at least 0.200. For human parvovirus B19, a re- The antibody status of the mothers who had still-
combinant protein was used in solid-phase EIA births and their matched controls was similar (in re-
0Acta Obstet Gynecol Scand 75 (1996)
