Key words
adenomyosis - fetal growth restriction - magnetic resonance imaging - preeclampsia
- junctional zone
Schlüsselwörter
Adenomyose - fetale Wachstumsretardierung - Magnetresonanztomografie - Präeklampsie
- Übergangszone
Introduction
Adenomyosis is a benign invasion of endometrial glands and stroma into the myometrium
surrounded by a hypertrophic and hyperplastic myometrial tissue and represents a spectrum
of lesions, ranging from increased thickness of the junctional zone (JZ) to overt
adenomyosis and adenomyomas [1], [2]. The use of modern imaging techniques, especially magnetic resonance imaging (MRI),
enables its identification as diffuse or focal thickening of the myometrial JZ and
less common forms of adenomyosis such as adenomyoma, adenomyomatous polyps, and cystic
adenomyoma have also been described [3].
During pregnancy, trophoblast invasion to the endometrium and the myometrial JZ causes
decidualization and unique vascular changes [4]. Defective deep placentation was reported to be associated with a spectrum of obstetrical
complications such as late miscarriage, preterm labour, fetal growth restriction (FGR)
and preeclampsia [5].
Preeclampsia is defined as new hypertension and substantial proteinuria at or after
20 weeks of gestation [6]. Although the cause of preeclampsia remains largely unknown, the leading hypotheses
strongly rely on disturbed placental function in early pregnancy [7]. Failure of placentation, particularly the physiological transformation of spiral
arteries, leads to a stressed, unperfused placenta. Most probably because of the similar
pathogenesis, preeclampsia could be seen together with FGR, stillbirth, placental
abruption and preterm labor [8].
To the best of our knowledge, there is no prior clinical study investigating the relationship
between adenomyosis and preeclampsia. The aim of this study was to compare the presence
of adenomyosis on magnetic resonance imaging in patients with and without history
of preeclampsia in order to investigate the possible role of adenomyosis in the pathogenesis
of preeclampsia.
Material and Methods
This study was a prospective, randomized study conducted at an Obstetrics and Gynecology
Clinic of a University Hospital in Turkey, between January and July 2015.
A total of 69 women were included in the study. All patients (n = 76) with diagnosis
of preeclampsia who were delivered at our obstetrics clinic during study period were
called by phone and asked for enrollment in the study. Among this group, a total of
35 patients agreed with MRI examination and were enrolled. The control group consisted
of 34 patients having presented with non-gynecologic complaint with a history of at
least one pregnancy without preeclampsia and without any history of infertility, endometriosis
and/or leiomyoma, uterine surgery except prior low transverse incision cesarean section
and history of hydatiform mole. The control group was randomized based on age, gravidity
and parity numbers by computer.
All study subjects underwent pelvic MRI examination at least 6 months after their
pregnancy during out of their menstrual phase [9], [10]. The primary outcome was the presence of adenomyosis. The secondary outcomes were
early or late onset of preeclampsia, FGR and preeclampsia-related complications (abruptio
placenta, HELLP syndrome and eclampsia) in the presence of adenomyosis.
Full medical histories were obtained. Preeclampsia was defined as hypertension with
a systolic blood pressure of 140 mmHg or higher or a diastolic blood pressure of 90 mmHg
or higher occurring after 20 weeks of gestation in a woman with previously normal
blood pressure and proteinuria of at least 300 mg/24 hour urine collection [6]. Proteinuria in spot urine sample instead of 24 hour urine collection was used for
diagnostic purposes in patients having undergone emergency delivery. The onsets of
preeclampsia were considered as early (< 34 gestational weeks) or late (≥ 34 gestational
weeks). FGR was defined as at least 2 weeks of retardation in ultrasonographic measures
of the fetus compared to expected gestational week. All clinical assessments of the
participants were performed by the same investigator (PSH), and all MRI evaluations
were assessed by the same experienced radiologist (MF). The radiologist was blind
for the study groups.
MRI assessment
MRI was performed with 1.5-Tesla scanners (Signa, General Electric Medical Systems).
We acquired 7-mm slices with 1-mm spacing in the sagittal, coronal, and axial planes
relative to the orientation of the uterine cavity, using T2-weighted fast (turbo)
spin echo sequences (TR/TEef, 3500–4000 ms/100 ms, echo train length 17) in all three
planes. Surface coils (phase array pelvic coils) were used for data acquisition and
examinations were completed within 30 minutes in each case.
Previously described MRI criteria for the diagnosis of adenomyosis were used [9], [11]. Both direct and indirect signs of adenomyosis on MRI were measured. Submucosal
microcysts, adenomyoma and cystic adenomyoma were considered as direct signs. Maximum
thickness of JZ (≥ 12 mm), JZ differential (the highest value calculated as JZmax minus JZmin measured in both anterior and posterior walls in the sagittal slices) (> 5 mm), the
ratio of JZmax thickness to myometrial thickness from the same plane (> 40 %), focal thickenning
of JZ, globally enlarged uterus and non-uniform JZ contours were considered as indirect
signs of adenomyosis. Presence of adenomyosis was defined as presence of at least
one direct and/or indirect signs on MRI ([Fig. 1]).
Fig. 1 Pelvic MRI at mid-sagittal plane showing (arrows) increased focal junctional zone
thickness.
Ethical consent
The research protocol was approved by the institutional review board (no. 20478486-22)
on January 14th, 2015. Informed consent was obtained from each participant.
Statistical analysis
Statistical analysis was performed with IBM SPSS Statistics 21.0 (SPSS Inc., Chicago,
IL). The Shapiro–Wilk test was used to calculate whether the numeric variables were
normally distributed. For normally distributed variables, numeric data were analyzed
with the Studentʼs t test and cross-tables and Pearson χ2 analysis were employed in the evaluation of the categorical data. The Mann–Whitney
U test was used for abnormally distributed variables. Fisherʼs exact test was used
in data with small sample size with an expected frequency of 5 % or less. P-value
< 0.05 was considered as statistically significant.
Thirty-three patients in each group had a power of 0.80 with the effect size of d = 0.70
and α = 0.05 based on sample size calculations.
Results
Description of the sample
Study population consisted of 69 patients (median age 29 years old, range 17–41).
Baseline demographic and clinical characteristics in patients with and without preeclampsia
were similar except for the older gestational age at delivery in control group most
likely due to larger proportion of uncomplicated pregnancies ([Table 1]).
Table 1 Baseline demographic and clinical characteristic of the preeclampsia and control
groups.
|
Preeclampsia group (n = 35)
|
Control group (n = 34)
|
p-value
|
|
DM: diabetes mellitus, GDM: gestational diabetes mellitus, MRI: magnetic resonance
imaging
* studentʼs t-test, ** χ2 test, *** Fisherʼs exact test
|
|
Age
|
27.97 ± 5.81
|
30.17 ± 4.92
|
0.094*
|
|
Smoking
|
12 %
|
26 %
|
0.138**
|
|
Gravidity
|
2.02 ± 1.54
|
2.20 ± 1.14
|
0.591*
|
|
Parity
|
1.48 ± 0.85
|
1.76 ± 0.95
|
0.205*
|
|
Abortion
|
0.54 ± 1.06
|
0.47 ± 0.89
|
0.762*
|
|
Gestational age at delivery
|
33.75 ± 4.26
|
36.73 ± 2.95
|
0.001*
|
|
Time interval (delivery to MRI)
|
12.25 ± 3.48
|
13.26 ± 3.06
|
0.207*
|
|
DM and/or GDM
|
15 %
|
3 %
|
0.197***
|
|
Presence of myoma
|
17.1 %
|
5.9 %
|
0.259***
|
|
Presence of endometriosis/endometrioma
|
2.9 %
|
8.8 %
|
0.356***
|
Mean age, number of gravidity and parity rates of patients with and without FGR were
similar (28.77 ± 5.52 vs. 28.84 ± 5.42, p = 0.772; 2 ± 1.37 vs. 2.21 ± 1.49, p = 0.646;
1.55 ± 0.85 vs. 1.57 ± 0.94, p = 0.983), respectively. Mean age, number of gravidity
and parity rates of patients with early and late onset preeclampsia were similar (26.27 ± 4.31
vs. 29.76 ± 6.75, p = 0.068; 1.77 ± 1.16 vs. 2.29 ± 1.86, p = 0.562; 1.27 ± 0.75 vs.
1.70 ± 0.92, p = 0.105), respectively.
In patients with preeclampsia, the rate of HELLP syndrome, abruptio placenta and eclampsia
were 45.7, 14.3 and 8.6 %, respectively.
Comparison of adenomyosis and preeclampsia, its related complications and FGR
Adenomyosis was present in 85.5 % of the study population. The prevalence of adenomyosis
was not different between patients with and without preeclampsia (88.6 vs. 82.4 %,
p = 0.513), respectively ([Table 2]). The JZ differential tended to be higher in patients with preeclampsia compared
to control subjects (4.14 ± 2.54 vs. 3.11 ± 1.96; p = 0.066), respectively ([Table 2]).
Table 2 Comparison of MRI findings based on adenomyosis characteristics in preeclampsia and
control groups.
|
Preeclampsia group (n = 35)
|
Control group (n = 34)
|
p-value
|
|
* studentʼs t-test, ** χ2 test, *** Fisherʼs exact test, max: maximal, mm: milimeters
|
|
Submucosal microcyst
|
0 %
|
2.9 %
|
0.490***
|
|
Cystic adenomyoma
|
2.9 %
|
2.9 %
|
1.000***
|
|
Presence of adenomyoma
|
80 %
|
70.6 %
|
0.364**
|
|
Adenomyoma max (mm)
|
5.20 ± 4.22
|
4.94 ± 4.24
|
0.800*
|
|
Adenomyoma location
|
%
|
%
|
0.177**
|
|
|
28.6
|
8.7
|
|
|
|
25
|
13
|
|
|
|
17.9
|
39.1
|
|
|
|
7.1
|
4.3
|
|
|
|
7.1
|
26.1
|
|
|
|
7.1
|
4.3
|
|
|
|
7.1
|
4.3
|
|
|
Total size of adenomyoma (mm)
|
11.82 ± 12.32
|
12.12 ± 7.82
|
0.918*
|
|
Focal JZ (JZmax) (mm)
|
10.40 ± 3.98
|
11.32 ± 3.99
|
0.340*
|
|
Localization of JZmax
|
%
|
%
|
0.208**
|
|
|
21.7
|
29.2
|
|
|
|
52.2
|
16.7
|
|
|
|
8.7
|
8.3
|
|
|
|
0
|
7.2
|
|
|
|
4.3
|
8.3
|
|
|
|
8.7
|
29.2
|
|
|
|
4.3
|
4.2
|
|
|
JZmax/myometrium
|
0.53 ± 1.18
|
0.55 ± 0.13
|
0.726*
|
|
JZ differential (mm)
|
4.14 ± 2.54
|
3.11 ± 1.96
|
0.066*
|
|
JZ not-uniform
|
51.4 %
|
52.9 %
|
0.900**
|
|
Uterine size (mm)
|
47.48 ± 7.83
|
50.33 ± 7.62
|
0.130*
|
|
Adenomyosis diagnostic criteria
|
%
|
%
|
0.239**
|
|
|
11.4
|
17.6
|
|
|
|
34.3
|
20.6
|
|
|
|
11.4
|
2.9
|
|
|
Presence of overall adenomyosis
|
88.6 %
|
82.4 %
|
0.513***
|
There was no relationship between the presence of adenomyosis and preeclampsia-related
complications including abruptio placenta (p = 0.477), HELLP syndrome (p = 1.000)
and eclampsia (p = 0.313). Presence of adenomyosis were compared in patients with
preeclampsia with or without FGR and only the presence of adenomyoma was more common
in patients with FGR compared to patients without FGR (94.4 vs. 64.7 %; p = 0.041),
respectively ([Table 3]). There was no correlation between FGR and early/late onset preeclampsia; 9 out
of of 18 patients with FGR (50 %) had pregnancy complicated with early-onset preeclampsia
while 9 out of 18 patients with FGR were complicated with late-onset preeclampsia
(p = 0.862).
Table 3 Comparison of MRI findings based on adenomyosis characteristics in patients with
preeclampsia with and without fetal growth restriction.
|
Adenomyosis characteristics
|
FGR (+) (%) (n = 18) (median [min–max])
|
FGR (−) (%) (n = 17) (median [min–max])
|
p-value
|
|
FGR: fetal growth restriction, JZ: Junctional Zone, max: maximal, mm: milimeters.
* Mann-Whitney U test, ** Fisherʼs exact test, *** χ2 test.
|
|
Presence of adenomyoma
|
94.4 %
|
64.7 %
|
0.041**
|
|
Adenomyoma max (mm)
|
5.5 (0–20)
|
3 (0–11)
|
0.115*
|
|
Total size of adenomyoma (mm)
|
8 (3–54)
|
6 (2–35)
|
0.273*
|
|
JZmax/myometrium
|
0.47 (0.21–0.83)
|
0.54 (0.22–0.83)
|
0.175*
|
|
JZ differential (mm)
|
3 (1–9)
|
3 (2–13)
|
0.366*
|
|
JZ non-uniform
|
66.7 %
|
50 %
|
0.241***
|
|
Focal JZ (JZmax) (mm)
|
10 (4–18)
|
10 (4–19)
|
0.295*
|
|
Uterine size (mm)
|
47.75 (37–73)
|
44 (35.5–61)
|
0.590*
|
Subgroup analysis based on the early or late onset of preeclampsia
Diagnostic signs (direct and indirect) of adenomyosis were compared in patients with
early and late onset preeclampsia. There was a strong association between three indirect
signs of adenomyosis and late-onset preeclampsia; JZmax/myometrial thickness, JZ differential and JZmax were found to be significantly higher in patients with late-onset preeclampsia than
in patients with early-onset preeclampsia [0.42 (0.21–0.72) vs. 0.66 (0.36–0.83),
p = 0.001; 3 (1–6) vs. 5 (2–13), p = 0.020; 9 (4–15) vs. 13 (7–19), p = 0.005], respectively
([Table 4]).
Table 4 Comparison of MRI findings based on adenomyosis characteristics in patients with
early and late-onset preeclampsia.
|
Adenomyosis characteristics
|
< 34 weeks (n = 18) (median [min–max])
|
≥ 34 weeks (n = 17) (median [min–max])
|
p-value
|
|
JZ: Junctional Zone, max: maximal, mm: milimeters
* Mann-Whitney U test, ** Fisherʼs exact test, *** χ2 test
|
|
Presence of adenomyoma
|
72.2 %
|
88.2 %
|
0.402**
|
|
Adenomyoma max (mm)
|
4 (0–11)
|
6 (0–20)
|
0.175*
|
|
Total size of adenomyoma (mm)
|
7 (2–27)
|
9 (2–54)
|
0.262*
|
|
JZmax/myometrium
|
0.42 (0.21–0.72)
|
0.66 (0.36–0.83)
|
0.001*
|
|
JZ differential (mm)
|
3 (1–6)
|
5 (2–13)
|
0.020*
|
|
JZ non-uniform
|
38.9 %
|
64.7 %
|
0.127***
|
|
Focal JZ (JZmax) (mm)
|
9 (4–15)
|
13 (7–19)
|
0.005*
|
|
Uterine size (mm)
|
45.25 (38.5–61)
|
49 (35.5–73)
|
0.107*
|
Discussion
MRI is currently an important technique with high diagnostic value for the diagnosis
of adenomyosis [12]. Histopathologic evaluation of uterus has been the gold standard diagnostic method
for adenomyosis until Hricak et al. described the normal zonal anatomy as a distinct
low signal on T2-weighted sequences seperating endometrium in high signal intensity
from outer myometrium in intermediate signal on MRI in 1983 [13].
Adenomyosis is not a uniform disease. It rather represents a spectrum of lesions,
ranging from disruption of the JZ architecture with little or no endometrial invasion
to overt diffuse adenomyosis and focal adenomyomas [4]. JZ is the inner ⅓ part of myometrium and found to be different than outer myometrium
by its embryonic origin and functions [4], [14]. Although the distinct myometrial zonal anatomy on MRI disappears in pregnancy,
disruption of the JZ prior to conception may have profound repercussion on deep placentation
and subsequent pregnancy outcome. Thickening and disruption of the JZ appearance is
strongly associated with uterine adenomyosis [4].
The placental bed, the area of the uterus underlying the placenta, plays a key role
in supporting placental function by supplying oxygenated blood to the intervillous
space via the spiral arteries [15], [16]. Decidualization initiates some morphologic changes in JZ spiral arteries outside
the placental bed [17]. Defective deep placentation was reported to be associated with a spectrum of obstetrical
complications such as late miscarriage, preterm labour, fetal growth restriction and
preeclampsia [5]. Interstitial trophoblast invasion of the JZ appears adequate, although impaired
decidualization of the myometrial spiral arteries predisposes for failed intravascular
trophoblast invasion in those conditions [18], [19].
Although the cause of preeclampsia remains largely unknown, failured remodeling of
the spiral artery has especially been considered as an early defect causing preeclampsia.
This abnormal placentation leads to reduced uteroplacental arterial flow and episodes
of irregular placental perfusion resulting in oxidative stress, subsequent apoptotic
and necrotic disruption of syncytial architecture and release of various components
from the intervillous space into the maternal circulation, stimulating production
of inflammatory cytokines. This leads to systemic maternal disease as the second stage
of preeclampsia [20].
Preeclampsia consists of many different clinical subtypes like early-onset preeclampsia
often complicated by FGR, preeclampsia accompanied with HELLP syndrome and late-onset
preeclampsia. Early and late-onset preeclampsia are defined as preeclampsia that develops
before and at or after 34 weeks of gestation, respectively. Although the presenting
features overlap, they are associated with different maternal and fetal outcomes,
biochemical markers, heritability, and clinical features [21]. Early onset preeclampsia has a worse prognosis characterized with early birth,
perinatal death and/or severe neonatal morbidity. Lisonkova et al. reported that congenital
anomalies were more strongly associated with early-onset disease, suggesting the presence
of associated placental abnormalities that affect perfusion. In contrast, a stronger
positive association between diabetes mellitus and late-onset preeclampsia suggests
that relative placental insufficiency is more likely to occur in diabetic pregnancies
with a larger fetus. They also observed a strong association between early-onset disease
and small-for-gestational age fetuses, likely because of the profound effects of poor
placental perfusion early in gestation [22]. We found a relationship between late-onset preeclampsia and indirect signs of adenomyosis
on MRI suggesting that adenomyosis might be related with a relative placental insufficiency.
Although the causes of preeclampsia remain one of the great medical mysteries of our
time, some promising researches are being published recently. Fisher et al. reported
that cytotrophoblasts that are isolated from the placentas of affected pregnancies
normalized their gene expression over 48 hours in vitro; this unexpected finding suggests
that some aspects of the aberrant differentiation of cytotrophoblasts within the uterine
wall that is observed in situ may be reversible. This result points to the importance
of environment factors in driving the phenotype. The next challenge is to identify
what they are. There are new clues to the longstanding mystery of the reason of abnormal
placentation in preeclampsia and the added urgency to find the answers, because these
pathways could be valuable therapeutic targets for reversing abnormal placental function
in patients [23]. In light of our study, we speculate that adenomyosis (especially defective junctional
zone) might be one of these mysterious factors. Thus, careful assessment for defective
junctional zone during routine ultrasonographic examinations in pregnancy as a predictive
factor for late-onset preeclampsia should be made.
A recent retrospective study investigated the relationship between adenomyosis and
uterine enlargement and poor pregnancy outcomes on 36 cases with adenomyosis and 144
control pregnancies. They found that the adenomyosis group had significantly higher
rates of preterm delivery, preterm premature rupture of membranes, small-for-gestational
age, fetal malpresentation and higher cesarean delivery as compared with the control
group. The authors concluded that adenomyosis was associated with a higher preterm
delivery rate and more frequent occurrences of FGR and fetal malpresentation. There
were no statistical differences in terms of preeclampsia between groups [24]. Similarly, we could not find any relationship between preeclampsia and overall
frequency of adenomyosis. This was probably due to the fact that neither preeclampsia
in itself is not a uniform diease nor adenomyosis has only one diagnostic sign [25]. We believe that subgroup analyses in larger group of patients are important to
make a comment in this point of view.
Yorifuji et al. reported two pregnant cases with non-contrast magnetic resonance angiography
technique. Both of the cases had large adenomyomas at the posterior site of the uteruses
and both pregnancies resulted in FGR. One of the cases had preeclampsia. They speculated
that “vascular steal” by adenomyosis might be the reason for the FGR in the cases
[26]. We also found a relationship between the presence of adenomyoma on MRI and FGR,
suggesting that presence of adenomyoma might be related with a poor placental perfusion.
We found no relationship between signs of adenomyosis on MRI and development of preeclampsia-related
complications such as placental abruption and eclampsia.
There are two important results of our study. Firstly, there was no overall relationship
between preeclampsia and adenomyosis. Comparison of patients with late and early-onset
preeclampsia and MRI characteristics showed a significant relationship between late-onset
preeclampsia and indirect signs of adenomyosis. Secondly, we found a significant relationship
between presence of adenomyoma and FGR. Thus, not only the mere presence of adenomyosis,
but also the type and/or size of the adenomyotic lesions might have a role in the
development of preeclampsia and FGR. As a limitation of our study, there was no control
group of SGA without preeclampsia and numbers of the patients in subgroups were relatively
small.
Conclusion
Presence of adenomyoma on MRI is more common in cases with preeclampsia complicated
with FGR and indirect signs of adenomyosis detected on MRI might play a role in the
pathogenesis of late-onset preeclampsia. Further studies with larger groups of patients
are needed to confirm these findings.
