Pathology Research and Practice 208 (2012) 458461

Contents lists available at SciVerse ScienceDirect

Pathology Research and Practice

journal homepage: www.elsevier.com/locate/prp

Original article

Spontaneous abortion in multiple pregnancy: Focus on fetal pathology

Jzsef Gbor Jo , kos Csaba, Zsanett Szigeti, Jnos Rig Jr.
Semmelweis University, General Medical Faculty, 1st Department of Gynecology and Obstetrics, Hungary

a r t i c l e

i n f o

Article history:
Received 29 October 2011
Received in revised form 2 January 2012
Accepted 8 May 2012
Keywords:
Spontaneous abortion
Multiple pregnancy
Monochorionic placentation
Perinatal autopsy
Developmental disorders
Intrauterine infection

a b s t r a c t
Multiple pregnancy with its wide array of medical consequences poses an important condition during
pregnancy.
We performed perinatal autopsy in 49 cases of spontaneous abortion resulting from multiple pregnancies during the study period.
Twenty-seven of the 44 twin pregnancies ending in miscarriage were conceived naturally, whereas
17 were conceived through assisted reproductive techniques. Each of the 5 triplet pregnancies ending
in miscarriage was conceived through assisted reproductive techniques. There was a positive history
of miscarriage in 22.4% of the cases. Monochorial placentation occurred more commonly in multiple
pregnancies terminating with miscarriage than in multiple pregnancies without miscarriage. A fetal
congenital malformation was found in 8 cases. Three of these cases were conceived through assisted
reproductive techniques, and 5 were conceived naturally. Miscarriage was due to intrauterine infection
in 36% of the cases.
Our study conrms that spontaneous abortion is more common in multiple than in singleton pregnancies. Monochorial placentation predicted a higher fetal morbidity and mortality. In pregnancies where all
fetuses were of male gender, miscarriage was more common than in pregnancies where all fetuses were
female. Assisted reproductive techniques do not predispose to the development of fetal malformations.
2012 Elsevier GmbH. All rights reserved.

Introduction
Multiple pregnancy with its wide array of medical consequences
poses an important condition during pregnancy. The signicance
of this is further increased by the recent spread of assisted reproductive techniques. In classic obstetric literature, the expected
frequency of multiple pregnancy is calculated by the Hellin formula
[6,8,18,23]. This formula predicts a ratio of 1 twin pregnancy in
every 85 pregnancies, 1 triplet pregnancy in every 852 pregnancies,
and 1 quadruplet pregnancy in every 853 pregnancies. Nowadays,
with the appearance of assisted reproductive techniques (ART),
the incidence of twin pregnancies has increased two-fold, triplet
pregnancies approximately twelve-fold and quadruplet pregnancies approximately eighty-fold compared to expected values by the
Hellin formula [16]. Furthermore, it has been suggested by some
researchers that gestational folic acid supplementation may also
moderately increase the risk for multiple pregnancy at least in case
of natural conception. However, subsequent papers failed to conrm this [19,27].

Corresponding author at: Semmelweis University, 1st Department of Gynecology and Obstetrics, 1088 Budapest, Baross utca 27, Hungary. Tel.: +36 1 266 04 73;
fax: +36 1 317 61 74.
E-mail address: joogabor@hotmail.com (J.G. Jo).
0344-0338/$ see front matter 2012 Elsevier GmbH. All rights reserved.
http://dx.doi.org/10.1016/j.prp.2012.05.011

Based on what factors result in the development of multiple pregnancy, different kinds of placentation may occur
[1,12,20,23,24]. In dizygotic fetuses, as well as in monozygotic
fetuses when separation of amniotic sacks occurs within a few
hours after fertilization of the egg, dichorionic diamniotic placentation develops. If the spatial separation occurs after separation
of trophoblasts from embryoblasts, monoplacental diamniotic
pregnancy will develop. Monochorionic monoamniotic pregnancy
develops when the embryoblasts split immediately before or
immediately after implantation.
The following maternal complications may be associated with
multiple pregnancy: preeclampsia, anemia, cervical incompetence,
abruptio placentae, and placenta previa. In terms of fetal complications with multiple pregnancy, increased perinatal morbidity and
mortality and increased risk for miscarriage and premature delivery
have been observed [5,18,20,23,24].
In this study, we attempted to evaluate and summarize the
results of perinatal autopsies performed after spontaneous abortions that occurred in multiple pregnancies between gestational
weeks 1224. In cases of mid-term miscarriages in multiple
pregnancy, perinatal autopsy may provide useful information
relevant to the risk of future miscarriages. This may be especially valuable if fetal congenital malformation is diagnosed at
the autopsy, providing a valuable tool in risk assessment for
recurrence.

J.G. Jo et al. / Pathology Research and Practice 208 (2012) 458461

Materials and methods

In this study, we report data from perinatal biopsies performed at The First Department of Gynecology and Obstetrics,
Faculty of General Medicine, Semmelweis University, Budapest,
Hungary after mid-term spontaneous abortions in multiple pregnancies during the study period of January 1, 1995December
31st, 2010. We performed 49 perinatal autopsies from 44 twin
pregnancies and 5 triplet pregnancies. Among the 17 multiple
pregnancies conceived by ART, no case with the application of
donor oocyte occurred. In each case, miscarriage occurred between
gestational weeks 1224. All ultrasound examinations were performed in our departmental ultrasound laboratory. Autopsies were
performed in our departmental pathology laboratory based on
protocols from our own department, as well as on those from
other well-recognized pathology laboratories around the world
[2,9,21,22,25]. During the autopsy, we were able to judge the type
of placentation, the eventual presence of chorioamnionitis, as well
as relevant characteristics of fetal pathology. The diagnostic criteria
of chorioamnionitis were positive vaginal discharge sample taken
during the pregnancy and histologically veried chorioamnionitis
(the evaluation of vaginal discharge samples was performed in the
departmental Chemistry Lab).
We have collected detailed information about the history concerning previous obstetrical complications (spontaneous abortion,
missed abortion, premature birth, intrauterine death, ectopic pregnancy positive obstetric history), and previous pregnancies with
genetic malformations (positive genetic history).
If among the clinical diagnoses a likely obstetric etiology was
found for the miscarriage, this was identied based on review
of the medical records. If a malformation was present in previous pregnancies which was different from the condition giving
rise to miscarriage in the current pregnancy, we regarded it as a
nonspecic genetic risk. If, however, the malformation was identical in the present and previous pregnancies, risk was regarded
specic for that condition. All relevant pieces of information
were included in a computerized database. This database was
then used for all subsequent analyses. For statistical signicance,
we used a p value of <0.05. Abbreviations used in this paper
included: HC hydrocephalus; SB spina bida; CAM cysticus adenomatoid malformation; TTTS twin to twin transfusion
syndrome.

Results
A total of 2141 autopsies were performed in our pathology laboratory during the study period. Out of these, 378 cases (17.6%) were
perinatal autopsies due to miscarriage. During the study period of
16 years, the occurrence of spontaneous abortion was 0.91% per live
birth, whereas the occurrence was higher in multiple pregnancies
(7.3% per live birth). Out of the 378 miscarriage cases, miscarriage
from multiple pregnancy was 49 (12.9%). Forty-four miscarriages
out of the 49 (89.8%) were from twin pregnancies and 5 (10.2%)
were from triplet pregnancies. Out of the 44 twin pregnancies, 27
(61.4%) were conceived naturally, while 17 (38.6%) were conceived
through assisted reproductive techniques. All of the 5 triplet cases
were conceived through assisted reproductive techniques. The total
number of fetal samples amounted to 103.
A positive obstetric history was identied in 11 out of the 49
(22.4%) multiple pregnancy miscarriage cases. In 7 out of these 11
cases (63.6%), assisted reproduction techniques were used, while
4 (36.4%) were conceived naturally. The most common obstetric
history was miscarriage in a previous pregnancy (5 cases).
The mean maternal age of patients undergoing ART was 33 2.8
years vs. the mean median age value of pregnant women having

459

Table 1
Fetal gender distribution in twin pregnancies ending in spontaneous abortion.
Gender distribution

Male/male
Female/female
Male/female

23
18
3

52.3
40.9
6.8

a multiple pregnancy conceived spontaneously at 29 3.2 years

(p < 0.05).
The indication for applying ART technique was pelvic factor in
8/17 cases (47.1%), male factor in 5/17 cases (29.4%), ovulatory factor in 3/17 cases (17.6%), and unexplained infertility in 1/17 cases
(5.9%).
Fetal gender distribution in the multiple pregnancies resulting in spontaneous abortion is displayed in Table 1. In our series,
the number of cases where fetal gender was identical was signicantly higher (p < 0.05) than the number of cases with heterogenous
fetal gender. In triplet pregnancies, male/male/male fetal gender
distribution and female/female/female distribution were equally
common (40%; 2 cases each). There was a single case with heterogenous fetal gender (20%).
The type of placentation could be identied in 34 cases based
on available information. These are described in Table 2. The ratio
of monochorionic vs. dichorionic multiple pregnancies was 11:23.
The difference in occurrence between mono- and dichorionic pregnancies was not signicant (p > 0.05).
Given the small number of triplet pregnancies, maternal age was
analyzed for the whole multiple pregnancy group, i.e., twin and
triplet pregnancies taken together. In cases where no specic fetal
pathology was identied on perinatal biopsy, the median maternal age was 30 3.64 years (range: 1645 years). In cases where
a specic fetal pathology was seen, the maternal median age was
29 5.46 years (range: 1744 years). The difference between these
two median values was not statistically signicant (p > 0.05).
In terms of gestational age at the time of spontaneous abortion, there was no statistical difference between cases where a
specic fetal pathology was found vs. cases where fetal pathology
could not be identied. The respective values were 19.5 2.77 gestational weeks in the positive fetal pathology group vs. 20 2.35
gestational weeks in the group with no fetal pathology identied
(p > 0.05). Fetal congenital malformation was identied in 8 out of
the 44 twin pregnancy cases (18.2%); major malformations (which
may explain the abortion) were veried in 6/44 (13.6%) and minor
ones (not incompatible with postnatal life) in 2/44 (4.6%) cases. In
the remaining 36 cases (81.8%), no specic fetal pathology could
be identied. As is demonstrated in Table 3, a congenital malformation was present in both fetuses in 1 case, whereas in 7 cases,
only one of the fetuses was affected (fetus A: the fetus closer to
birth canal). In the cases where a congenital malformation could
be identied, the pregnancy was naturally conceived in 63.5% of
cases, while assisted reproduction techniques were used in 36.5%
of cases. In the triplet pregnancy cases, a congenital malformation
was identied in a single case. This congenital malformation was
identied as spina bida lumbosacralis. In this case, as in all other
cases of triplet pregnancies, the pregnancy was conceived through
assisted reproductive techniques.

Table 2
Types of placentation in multiple pregnancies ending in spontaneous abortion.
Placentation

Monochorionicmonoamniotic placentation
Monochorionicdiamniotic placentation
Dichorionicdiamniotic placentation

1
10
23

2.9
29.4
67.7

460

J.G. Jo et al. / Pathology Research and Practice 208 (2012) 458461

Table 3
Fetal pathology found on perinatal autopsy after spontaneous abortion in multiple
pregnancy.

1
2
3
4
5
6
7
8

Affected fetus

Observed pathology

B
B
A
B
B
A
B
A and B

Auricular aplasia, oligodactyly, pes equinovarus

Hydrocephalus internus
Accessory renal pelvis (double pelvis)
CAM type II
Renal agenesis, right-sided
Single umbilical artery
Single umbilical artery
Lumbosacral spina bida cystica

Table 4 describes the most likely etiologies leading to miscarriage in our study. This data tends to conrm previous reports
revealing intrauterine infections as the most common etiology.
Among the 16 cases of diagnosed intrauterine infection, 4 (25%)
occurred due to premature rupture of membrane. Another relatively common diagnosis was twintwin transfusion syndrome
(TTTS) in 13.6% of the cases.
Discussion
Our ndings support the widely recognized phenomenon that
obstetric complications, including spontaneous abortion, are more
common in multiple vs. single pregnancy. In our study population,
there was an eight-fold higher incidence of miscarriage; this is
somewhat higher than the previously reported two- to four-fold
difference in occurrence [10,13,20,23,24].
It is remarkable that the ratio of twin pregnancies conceived naturally vs. conceived through assisted reproductive techniques (64%
vs. 36%) was identical irrespective of whether miscarriage occurred
vs. pregnancy was carried to term. This suggests that assisted reproductive technique does not affect the likelihood of spontaneous
abortion in this context. Fitzsimmons et al. came to a similar conclusion [7], proposing that naturally conceived multiple pregnancies
may even carry a higher risk for miscarriage compared to those
conceived through assisted reproductive techniques.
There was a positive history for spontaneous abortion in nearly
64% of multiple pregnancies conceived through assisted reproductive techniques. In some of these cases, assisted reproductive
technique was employed precisely because of previous failure to
produce live birth due to underlying pathology.
The median age of patients undergoing ART was signicantly
higher than that of women having spontaneously conceived twin
pregnancy. As ART is usually applied after years of several attempts
to become pregnant spontaneously, this result was as was to be
expected.
The risk for miscarriage in multiple pregnancies is signicantly
higher in cases where fetal gender is identical in all fetuses within
the given pregnancy [26]; this is supported by our ndings. This
phenomenon can be explained by the fact that when fetal gender
is the same, monochorial, and monoamniotic multiple pregnancy
is more likely. It is well recognized that the risk for miscarriage is
higher in monochorionic forms of multiple pregnancy [1,15,20].
Table 4
Most likely clinical diagnosis resulting in spontaneous abortion in multiple
pregnancy.
Clinical diagnosis

Chorioamnionitis
Twintwin transfusion syndrome (TTTS)
Incompetent cervix
Preeclampsia
Fetal malformation
Etiology unclear

16
6
3
1
8
10

36.40
13.60
6.80
2.30
18.20
22.70

However, our ndings suggest that the inuence of fetal gender

goes beyond the question of chorinicity or amnionicity. We found
that in the setting where fetal gender was malemale, there was
a 13% higher risk for miscarriage compared to cases where fetal
gender was femalefemale.
In the 34 cases in which chorionicity could be accurately judged
based on available information, dichorionic placentation was signicantly more common than monochorionic (23 vs. 11 cases or
68% vs. 32%). Considering that the ratio of monochorionic multiple pregnancies to all multiple pregnancies is expected to be
between 15 and 25% [11,12], this tends to support previous beliefs
that monochorionicity in multiple pregnancy is a risk factor for the
development of miscarriage in this setting [4,10,17].
There is no difference in maternal age between cases where
spontaneous abortion occurred through singular or multiple pregnancy. Our study shows that in multiple pregnancy there was no
difference in maternal age between cases where, in the background
of miscarriage, a fetal congenital malformation could be identied
vs. cases in which fetal pathology was not seen.
Similarly, there was no signicant difference with regard to gestational age at the time of miscarriage in the cases where a fetal
pathology was identied vs. cases where there was no such pathology. It is noteworthy that in singular pregnancies, miscarriage
associated with congenital malformation generally takes place 3
weeks earlier, compared to cases where a malformation is not identied [14]. In multiple pregnancies, this time difference is usually
not seen.
In twin pregnancies ending in spontaneous abortion, we found
a developmental disorder in at least 1 fetus in 18% of the cases
although the rate of major fetal malformations was 13.2%. These
data represent a higher incidence of fetal congenital malformation compared to miscarriage in singular pregnancies (10%) [14].
Livingston and Poland found a similar incidence of congenital malformation in at least one fetus in twin pregnancies ending in
miscarriage (21%) [20]. Arteaga et al. observed substantially higher
incidences; in more than 46% cases were congenital malformations
identied [3].
In regard to the relationship of congenital malformations and
the nature of conception, we found no difference in the incidence
of malformations on the basis of whether conception was natural or
induced by assisted reproductive techniques. This contradicts prior
observations made by Fitzsimmons et al., demonstrating a higher
fetal morbidity in spontaneously conceived multiple pregnancies
[7].
In our study, disorders of the central nervous system were
prominent among the type of fetal pathology identied. This represented 3 out of the 9 cases of congenital malformation. This
conrms previous nding by Arteaga et al., who observed a similar
prominence of central nervous system (or craniospinal) disorders
in this case scenario [3].
As can be seen in Table 4, the most common clinical diagnosis
found in the background of spontaneous abortion was intrauterine infection (chorioamnionitis). This underlines the importance of
treatment and especially the prevention of intrauterine infection in
an attempt to decrease the incidence of miscarriage. It is important
to note that TTTS occurred in more than 30% of the cases, suggesting that this phenomenon deserves further attention, and future
studies to evaluate the nature of this relationship are warranted.
In conclusion, our study demonstrated that multiple pregnancies are associated with a higher rate of obstetric complications
compared to singular pregnancies. Conception through assisted
reproductive techniques does not necessarily predict a higher risk
for spontaneous abortion in multiple pregnancies. In cases where
fetuses have identical gender within a given pregnancy, there
is a signicantly higher incidence of miscarriage. Among multiple pregnancies ending in spontaneous abortion, there was no

J.G. Jo et al. / Pathology Research and Practice 208 (2012) 458461

difference in the incidence of fetal congenital malformations

between pregnancies conceived through assisted reproductive
techniques vs. naturally conceived pregnancies.
References
[1] J. Ando, K. Nonaka, K. Ozaki, et al., The Tokyo twin cohort project: overview and
initial ndings, Twin Res. Hum. Gen. 9 (2006) 817826.
[2] J.G. Archie, J.S. Collins, R.R. Lebel, Quantitative standards for fetal and neonatal
autopsy, Anat. Pathol. 126 (2006) 256265.
[3] M. Arteaga, I. Garca-Pelez, L.R. Cornejo, et al., Fetal dysmorphism in spontaneous abortions in a Mexican population, Int. J. Obstet. Gynecol. 73 (2001)
131140.
[4] R. Bajoria, L.Y. Wee, S. Anwar, et al., Outcome of twin pregnancies complicated
by single intrauterine death in relation to vascular anatomy of the monochorionic placenta, Hum. Reprod. 14 (1999) 21242130.
[5] F.G. Cunningham, N.F. Gant, K.J. Leveno, et al., Multifetal pregnancy, in: F.G.
Cunningham (Ed.), Williams Obstetrics, 21st ed., McGraw and Hill, New York,
2001, pp. 765810.
[6] J.O. Fellman, A.W. Eriksson, A mathematical model for recurrent twinning, Acta
Genet. Med. Gemellol. 39 (1990) 307316.
[7] B.P. Fitzsimmons, M.W. Bebbington, F.R. Fluker, Perinatal and neonatal outcomes in multiple gestations: assisted reproduction versus spontaneous
conception, Am. J. Obstet. Gynecol. 179 (1998) 11621167.
[8] H.A. Freye, Genetic and peristatic aspects of research on twinning, Zentralbl.
Gynakol. 109 (1987) 14011408.
[9] S.J. Gordijn, J.J. Erwich, Value of perinatal autopsy: critique, Pediatr. Dev. Pathol.
5 (2002) 480488.
[10] W.A. Grobman, A.M. Peaceman, What are the rates and mechanisms of rst
and second trimester pregnancy loss in twins? Clin. Obstet. Gynecol. 41 (1998)
3745.
[11] A. Gul, A. Cebeci, H. Aslan, I. Polat, et al., Perinatal outcomes of twin pregnancies
discordant for major fetal anomalies, Fetal Diagn. Ther. 20 (2005) 244248.

461

[12] P.A. Hatkar, A.D. Bhide, Perinatal outcome in twins with relation to chorionicity,
J. Postgrad. Med. 45 (1999) 3337.
[13] E. Hruby, Multifetal pregnancy in evidence based medicine, in: Z. Papp
(Ed.), Dialogue in Perinatology, Golden Book Kiad, Budapest, 2001,
pp. 412419.
[14] J.G. Jo, A. Beke, Z.s. Szigeti, et al., Miscarriages in fetopathological examinations, Fetal Diagn. Ther. 25 (2009) 186191.
[15] H.K. Kaufman, R.F. Hume, B.C. Calhoun, et al., Natural history of twin gestation
complicated by in utero fetal demise: associations of chorionicity, prematurity
and maternal morbidity, Fetal Diagn. Ther. 18 (2003) 442446.
[16] T. Kernyi, . Csaba, Fetal demise in multiple gestation, in: Z. Papp (Ed.), Dialogue in Perinatology, Golden Book Kiad, Budapest, 2001, pp. 428440.
[17] B.W. Kovacs, T.H. Kirschbaum, R.H. Paul, Twin gestations. I: antenatal care and
complications, Obstet. Gynecol. 74 (1989) 313317.
[18] L. Lamp, Z. Papp, Multiple gestation, in: Z. Papp (Ed.), Obstetrics and Gynecology, Semmelweis Kiad, Budapest, 1992, pp. 96102.
[19] T. Levy, I. Blickstein, Does the use of folic acid increase the risk of twinning?
Int. J. Fertil. Womens Med. 51 (2006) 130135.
[20] J.E. Livingston, B.J. Poland, A study of spontaneously aborted twins, Teratology
21 (1980) 139148.
[21] T. Marton, B. Hargitai, P. Patks, et al., Practice of fetopathological examination,
Orv. Hetil. 140 (1999) 14111416.
[22] P.R. Meier, D.K. Manchester, R.H. Shikes, et al., Perinatal autopsy: its clinical
value, Obstet. Gynecol. 67 (1986) 349351.
[23] F. Paulin, Multifetal pregnancy, in: Z. Papp (Ed.), Obstetrics and Gynecology,
Semmelweis Kiad, Budapest, 2006, pp. 326335.
[24] M.L. Pernoll, R.C. Benson, Multiple pregnancy, in: Current Obstetric & Gynecologic Diagnosis and Treatment, Appleton&Lange, Norwalk, USA, 1997, pp.
357367.
[25] D.N. Saller, K.B. Lesser, U. Harrel, et al., The clinical utility of the perinatal
autopsy, JAMA 273 (1995) 663666.
[26] J.G. Santema, I. Koppelaar, H.C.S. Wallenburg, Hypertensive disorders in twin
pregnancy, Eur. J. Obstet. Gynecol. Reprod. Med. 58 (1995) 913.
[27] S.E. Vollset, H.K. Gjessing, A. Tandberg, et al., Folate supplementation and twin
pregnancies, Epidemiology 16 (2005) 201205.