Content uploaded by Sung Eun Rha
Author content
All content in this area was uploaded by Sung Eun Rha on Jan 07, 2019
Content may be subject to copyright.
120 6 AJR :196, May 20 11
have precisely correlated the MRI data with
those obtained on detailed anatomicohisto-
logic examination (Fig. 2). The junctional
zone has been shown to contain myocytes
with morphologic characteristics differing
from those of the typical myocytes of the
outer myometrium. Junctional zone myo-
cytes present a greater relative nuclear area,
a looser extracellular matrix, and lower wa-
ter content [8]. These elements serve to de-
crease the signal of this zone on T2-weighted
imaging, although the different morphology
of myocytes is not the sole cause. The archi-
tectural organization of the inner myometri-
um is unique with a concentric arrangement
of smooth-muscle fibers in contrast to the
longitudinal orientation of the smooth-mus-
cle fibers of the outer myometrium [10].
To better understand the histologic factors
that explain the signal characteristics of the
junctional zone, it is pertinent to objective-
ly evaluate healthy control subjects. The sim-
plest and most-studied measure is the thickness
of the junctional zone, which is defined as the
subendometrial low signal band [1–4, 11]. The
thickness of this layer is crucial because above
a certain threshold it is a strong criterion for the
diagnosis of adenomyosis. To be reproducible,
the junctional zone measurement should be
made on a midsagittal image through the long
axis of the uterus [12] (Fig. 3).
Physiologic Variations of the Junctional Zone
According to patient age—During premen-
arche, pregnancy, or postmenopause, the zon-
al anatomy of the uterine muscle is often less
MRI Characteristics of the Uterine
Junctional Zone: From Normal to
the Diagnosis of Adenomyosis
Sébastien Novellas1
Madleen Chassang1
Jerome Delotte2
Olivier Toullalan3
Anne Chevallier4
Jerome Bouaziz2
Patrick Chevallier1
Novellas S, Chassang M, Delotte J, et al.
1Service d’Imagerie Diagnos tique et Interventionnelle,
Cent re Hospi talier Régi onal et Univ ersitai re de Nice,
Hôpi tal Arche t 2, 151 Rte d e Saint An toine de Gine stière,
B.P 3 079, 0 620 2 Nice Cedex 3, F rance. Add ress
corre spondence t o S. Novellas ( novellas. s@chu- nice.fr) .
2Service de Gynécologie-Ob stétrique, Centre Hospit alier
Régio nal et Unive rsitair e de Nice, Hôpi tal Arche t 2,
Nice, France.
3Service d’Anatomo -pathologie, Centre Hospitalier
Régio nal et Unive rsitair e de Nice, Hôpi tal Arche t 2,
Nice, France.
4Service de Gynécologie-Ob stétrique, Centre Hospit alier
Régio nal de Cannes , Nice, France .
Women’s Imaging • Review
CME
This a rticle is a vailable for C ME credit .
See www.arrs.org for more information.
AJR 20 11; 196:12 06–12 13
0361–803X/11/1965–1206
© Amer ican Roent gen Ray Socie ty
T
he corpus uterus is a complex or-
gan dedicated to reproduction,
traditionally divided on an ana-
tomicohistologic basis into two
clearly distinct parts: the endometrium and
myometrium. In 1983, Hricak et al. [1] delin-
eated the normal zonal anatomy on MRI that
was confirmed by other authors [2–5]. The
term “junctional zone” was introduced in
this manner to describe this interface ob-
served on MRI: a distinct low signal on T2-
weighted sequences separating the endome-
trium in high signal intensity from the outer
myometrium in intermediate signal. Pelvic
MRI was ahead of anatomicohistologic
methods, which on light microscopy could
not identify this uterine zonal anatomy [6].
Subsequently, it was clearly shown that this
zone corresponds to the innermost layer of
the myometrium and not to the basal layer of
the endometrium [7]. The goal of this article
is, first, to present the typical characteristics
of the junctional zone on MRI as well as its
physiologic variations according to patient
age and according to stage of the reproduc-
tive cycle. Second, we discuss uterine adeno-
myosis and the value of junctional zone as-
sessment in its diagnosis.
Junctional Zone
General Information
Since the publication of Hricak et al. [1],
numerous groups have studied zonal anato-
my of the uterus on T2-weighted imaging,
most notably the low signal intensity of the
junctional zone [8–10] (Fig. 1). These works
Keywords: adenomyosis, junctional zone, pelvic
anato my, pelvic M RI, uter us, women’s i maging
DOI:10.2214/A JR.10.4 877
Recei ved April 2 7, 2010; acce pted af ter revisi on
Augus t 30, 2 010.
WOMEN’S
IMAGING
OBJECTIVE. MRI was the first imaging technique to permit the visualization of the uter-
ine junctional zone and remains the imaging method of choice to evaluate it and its associ-
ated pathology.
CONCLUSION. Adenomyosis can be diagnosed using MRI with a diagnostic accuracy
of 85%. The most important MR finding in making the diagnosis is thickness of the junction-
al zone exceeding 12 mm. The principal limitation of MRI is the absence of a definable junc-
tional zone on imaging, which occurs in 20% of premenopausal women.
Novel las et al.
MRI Ch aracter istics o f the Uter ine Junct ional Zon e
Women’s Imaging
Review
Downloaded from www.ajronline.org by 158.46.162.69 on 01/07/19 from IP address 158.46.162.69. Copyright ARRS. For personal use only; all rights reserved
AJR :196, May 20 11 1207
MRI Characteristics of the Uterine Junctional Zone
distinct on MRI [11, 13]. In the postmeno-
pausal female lacking hormone replacement
therapy, the signal of the outer myometrium
approaches the hypointensity of the junction-
al zone because of a progressive dehydra-
tion of the smooth-muscle tissue associated
with a fibrous involution of the extracellular
components of the outer myometrium. More-
over, there is age-related atrophy that reduces
the thickness of the uterine muscle to only a
few millimeters. Thus, in the postmenopaus-
al woman, it is estimated that the junctional
zone cannot be delineated in about 30% of
patients [14, 15]. Only one recent article re-
ports discordant results with only two faulty
measures of the junctional zone in 28 patients
51–80 years old, but the methodology of that
study is questionable [16]. In the premenar-
chal female, the junctional zone is sometimes
detectable as a faint line but is neither measur-
able nor discernible from the outer myometri-
um in a clear manner [11]. During pregnancy,
the junctional zone is poorly visualized as its
signal is augmented and approaches the sig-
nal of the outer myometrium. It progressively
reappears 15 days after delivery and is clearly
defined 6 months postpartum [17, 18]. As a re-
sult, these stages in the reproductive cycle of
the female represent periods when the junc-
tional zone cannot be adequately evaluated.
According to cycle—Hormonal variation in
the female reproductive cycle is one of the ma-
jor factors contributing to the changes in the
thickness of the junctional zone as measured
on MRI [14]. These changes parallel changes
in endometrial thickness but to a lesser degree
[19, 20]. As a result of this hormonal influ-
ence, the maximal thickness of the junction-
al zone is reached during the menstrual phase
[6]. The effect of the hormonal cycle on this
zone of muscular tissue is explained in part
by the uterine ontogeny; that is, it appears that
cells of the endometrium and cells of the junc-
tional zone have a common Müllerian origin,
whereas the outer myometrium is of a non-
Müllerian, mesenchymal origin [21, 22]. As a
result, the outer myometrium presents little or
no dependence on hormonal stimulation and
there is no significant variation in its thickness
during the reproductive cycle.
Role of the Junctional Zone
Ultrasound studies have been paramount in
understanding the role of the junctional zone,
which is intricately linked with its structural
organization and biochemical properties. Pel-
vic ultrasound, performed transabdominally in
the 1980s and later endovaginal ly in the 1990s,
has permitted observation of contractions of
the junctional zone occurring in the nongravid
uterus [23–25]. Video recordings have identi-
fied these contractions with a speed of 1.2–1.7
mm/s with a frequency of 3–5 contractions/
Fig. 1— Sagit tal T2-weight ed image of 35 -year-old woman in first par t of her
mens trual cycle sh ows zonal anat omy with e ndometrium (arrowhead), w hich
is hyperintense; junctional zone (short arrow) as hypointense band; and outer
myome trium (lo ng arrow) with intermediate signal.
Fig. 3 —Sagit tal T2 -
weigh ted image wi th fat
suppression o f 27-year-
old wom an in second
part of her cycle.
Uterus is evaluated
bet ween isth mus and
end of u terine cavi ty
(white lines). Junctional
zone (short arrows)
should be measured
from s everal sites on
anterior and posterior
walls. Junctional
zone measure can
be compared wi th
entire thickness of
myome trium (long
arrows ) evaluated at
same site.
Fig. 2 —Photo micrograph (H and E, × 25) of ut erine section of 3 7-year-old wo man
shows stru ctural pola rity in myomet rium. There is cont rast bet ween dense
arrangement o f smooth- muscle fiber s in inner myometrium (1) and less densit y of
oute r myometr ium (2).
Downloaded from www.ajronline.org by 158.46.162.69 on 01/07/19 from IP address 158.46.162.69. Copyright ARRS. For personal use only; all rights reserved
1208 AJR :196, May 20 11
Novellas et al.
min. Contractions are classified according to
their direction: from the uterine cervix toward
the body, from the body to the cervix, or both
simultaneously [26]. These junctional con-
tractions have been secondarily described on
MRI with particular dynamic sequences [27,
28]. They appear as thin bands perpendicular
to the junctional zone and in low signal inten-
sity on T2-weighted images. This signal has
been partly explained by the venous vasocon-
striction that reduces the quantity of circulat-
ing blood in the myometrium. In 2007, Kido
et al. [29] used MRI to illustrate the influence
of oral contraceptives on uterine contractions.
It was clearly shown that these contractions
have an amplitude, a frequency, and a direc-
tion that correlate with the phase of the men-
strual cycle. Thus, in the first part of the cycle,
contractions occur from the cervix toward the
body and increase in intensity until ovulation.
These contractions participate directly in the
transport of spermatozoids toward the ovum
as reported by Kunz and colleagues [30] by
utilizing inert particles. Ijland et al. [31] con-
firmed this hypothesis by studying the strong
link between contractions issued from the
junctional zone and fertility.
Normal Values of the Junctional Zone
In the 1980s, the first studies proposed be-
tween 2 and 5 mm as a maximum thresh-
old for a normal junctional zone thickness
[7–11]. Over the past 20 years, this criterion
has been regularly revised using new stud-
ies with greater statistical power [12, 15, 16,
32]. Most of these series had the advantage
of correlating junctional zone readings with
histologic analysis, the only method to con-
firm the absence of myometrial disease. In
this manner, the upper limit of normal for the
junctional zone thickness was adjusted up to
5–8 mm. The possible selection bias in these
studies based on patients requiring hyster-
ectomy calls into question the definition of
what constitutes a “normal” junctional zone.
However, it appears that a thickness of up to
8 mm is considered normal.
Adenomyosis
General Information
Adenomyosis is pathology at the inter-
face of the endometrium and myometrium.
Its definition is based on histology findings.
Siegler and Camilien [33] define adenomyo-
sis as the presence of endometrial glandular
cells and cells of the chorion more than 2.5
mm from the endometrium-myometrium in-
terface (Fig. 4). Histologic analysis also al-
lows the evaluation of the degree of invasion
and the extent of disease by examining the
number of foci per field of analysis. A hy-
pertrophic reaction of the smooth-muscle
cells surrounding the ectopic glands is an-
other important element of diagnosis [34].
Involvement of the myometrium can include
the entire interface, termed “diffuse adeno-
myosis,” or it can present in a limited area,
termed “focal adenomyosis.” These two
forms have similar frequency on hysterecto-
my samples but are rarely associated [6, 10,
32, 35]. Adenomyoma, a localized conflu-
ence of adenomyotic glands forming a mass,
constitutes an unusual form of adenomyosis.
The actual prevalence of adenomyosis re-
mains unknown [34]. Past surgical series re-
port a detection rate extremely variable from
5% to 70% [34]. This variation is probably
due in part to the methodology of histolog-
ic analysis: With three uterine sections an-
alyzed the incidence is 31% versus an in-
cidence of 61% given six sections [36]. If
recent studies involving more than 500 pa-
tients are subject to meta-analysis, this range
of detection rates is reduced to 21–47% with
a mean of approximately 30% [34]. Risk fac-
tors for adenomyosis—most notably, the ex-
istence of prior endouterine surgery—were
isolated [37]. Multiparity is another risk fac-
tor, which appears contradictory to the not
so well-established link between adenomy-
osis and infertility [34]. Adenomyosis pref-
erentially affects women between 40 and 50
years old; however, most forms are asymp-
tomatic, explaining perhaps the lower rate
of detection in young women. A positive di-
agnosis of adenomyosis cannot be made re-
liably on the basis of clinical examination
alone. Supporting evidence, including MRI
findings, is necessary to entertain and con-
firm the diagnosis [38, 39].
Diagnostic Strategy
Hysterosalpingography no longer serves a
major role in the diagnosis of adenomyosis
[39, 40]. This more invasive imaging proce-
dure has been largely supplanted by abdomi-
nal ultrasound followed by endovaginal ul-
trasound. The advantages of ultrasound are
ease of use on a daily basis and diagnostic
performance when exploration is of good
quality [15, 32, 35, 41]. However, MRI has
evolved as the current imaging technique of
choice in confirming suspected cases of ad-
enomyosis [15, 39, 42, 43]. To this end, sev-
eral signs of adenomyosis—both direct and
indirect—have been described.
Direct Signs on MRI
Under the term “direct signs” are grouped
all signs that have a specific correlation with
the presence of endometrial glands within
the myometrium.
Microcysts—The presence of submuco-
sal microcysts constitutes the principal di-
rect sign of adenomyosis. Histologic analysis
reveals these cysts to consist of islets of ec-
topic endometrium accompanied by a cystic
glandular dilatation. MRI reveals round cys-
tic foci varying from 2 to 7 mm in diameter,
with a mean around 3 mm, that are embed-
ded within the myometrium [15, 39, 43, 44].
They are usually located within the junctional
zone, although microcysts of the outer myo-
metrium also have been described. These
structures have water signal on MRI with hy-
pointense signal on T1-weighted imaging and
hyperintense signal on T2-weighted imaging
(Fig. 5). Occasionally, at the end of the men-
strual period, hemorrhagic content may be
detected, which appears in hyperintense sig-
nal on T1-weighted imaging (Fig. 6), within
these cystic cavities. This hemorrhagic con-
tent is not found routinely because it appears
that the endometrial cells of adenomyosis
are less dependent on circulating hormone
levels. Consequently, the presence of these
cysts, although almost pathognomonic of ad-
enomyosis, is detected on MRI in only about
half of the cases [15, 43].
Adenomyoma—An adenomyoma [38, 40,
43, 45] is composed of a focal consolidation
of adenomyotic glands located within the
myo metrium. This lesion typically manifests
Fig. 4 —Phot omicrograp h (H and E, ×2 5) of uter ine
sect ion of 33-year-old woman shows ectopic
endometria l tissue (arrow s) (2) extending into
myo metri um (1).
Downloaded from www.ajronline.org by 158.46.162.69 on 01/07/19 from IP address 158.46.162.69. Copyright ARRS. For personal use only; all rights reserved
AJR :196, May 20 11 1209
MRI Characteristics of the Uterine Junctional Zone
as a well-delineated myometrial mass dis-
tinct from the junctional zone. The principal
diagnostic differential is uterine leiomyoma,
and several criteria can be used to distinguish
between the two entities. Although both ad-
enomyoma and leiomyoma are of low signal
intensity on T2-weighted imaging, adenomy-
oma often will have high-signal-intensity foci
on T2. Moreover, in contrast to other myomas,
adenomyoma does not have large vessels at its
periphery. Adenomyoma is rarer than both fo-
cal and diffuse adenomyosis, although its true
prevalence is difficult to establish. Reinhold
et al. [32] isolated only one case among 28
patients with adenomyosis, whereas Bazot et
al. [15] reported seven cases among 40 sur-
gically proven cases. Cases of cystic adeno-
myomas corresponding to an excessive hem-
orrhage of ectopic endometrium within the
myometrium have also been described. Such
lesions present as a cavity with a long axis of
greater than 1 cm that contains hemorrhagic
or liquid contents surrounded by a fibrous tis-
sue with hypointense signal on T2-weighted
imaging (Fig. 7).
Finally, relying solely on the presence of mi-
crocysts and adenomyomas alone proves insuf-
ficient to diagnose adenomyosis with sufficient
sensitivity. Other signs are available that are
more frequent and should be considered in the
overall picture: the indirect signs.
Indirec t Signs on MRI
The signs detailed are secondary to the re-
action of the myometrium provoked by endo-
metrial invasion.
Thickness of the junctional zone—The thick-
ness of the junctional zone is the sign the most
studied in making the diagnosis of uterine
adenomyosis [15, 32, 38, 40, 45–47]. When
this thickness involves the entire junctional
zone, the diagnosis of diffuse adenomyosis is
A
Fig. 5 —38 -year-old woman who p resented w ith chronic dysme norrhea.
A, A xial T2-weight ed image reveals microcysts (arrows ) in hypersig nal within j unctional z one.
B, Microc ysts (arrow s) ar e particularly we ll visualize d in uterine f undus on coronal image. N ote presence of ult rasound gel co ntrast (star ) in vagi na.
B
A
Fig. 6 —45 -year-old w oman who pres ented wit h menomet rorrhagia.
A, Sagi ttal T 2-weight ed image reveals ill- defined thi ckening of junc tional zone associated wi th submucosa l microcyst (a rrow ) in hypersignal.
B, Cys t (arrow) sh ows up in hype rsignal on sagi ttal T1-weigh ted image, re vealing its hemorr hagic nature.
C, Photo micrograph ( H and E, ×25) of myo metrial section s hows blood pool (star) w ithin island of ectopic endomet rial tissue w ith glandular cells (arrowheads ) and st roma
cells (arrows).
CB
Downloaded from www.ajronline.org by 158.46.162.69 on 01/07/19 from IP address 158.46.162.69. Copyright ARRS. For personal use only; all rights reserved
1210 AJR :196, May 20 11
Novellas et al.
entertained (Fig. 8), whereas focal adeno-
myosis is considered when only part of the
junctional zone is involved. Sometimes this
focal thickening is considerable enough to
form a poorly delineated myometrial mass
of low signal intensity and containing mi-
crocysts (Fig. 9). Generally, a junctional zone
thickness of greater than 12 mm is the most
widely accepted criterion in establishing the
presence of adenomyosis. This threshold was
derived after several studies that correlated
the measurement of the junctional zone with
the histologic diagnosis of adenomyosis [15,
32, 48]. When a junctional zone thickness is
greater than 12 mm, adenomyosis may be di-
agnosed with a diagnostic accuracy of 85% and
a specificity of 96%. This threshold, however,
carries a sensitivity, according to Bazot et al.
[15], of only 63%. This low sensitivity is under-
standable when one considers that the thicken-
ing of the junctional zone is associated with a
proliferation of myocytes of the inner layer of
the myo metrium and that adenomyosis is de-
fined by the presence of endometrial glands at
a certain distance from the endometrium sur-
rounded by smooth-muscle cells oriented in a
less-than-coherent fashion. The thickening of
the junctional zone can perhaps be caused by
a process of myometrial infiltration. This con-
cept is supported by several facts.
A
A
Fig. 7—42-year- old woman who p resented w ith chron ic pelvic pain .
A and B, Coronal (A) an d axial (B) T2-weighted i mages reveal cy stic area surround ed by low-si gnal-int ensity r ing (arrow). Over all appearance on MRI is that of cystic
adenomyoma.
Fig. 8 —46 -year-old w oman who pres ented with pelv ic heavines s.
A, Sagi ttal T 2-weighted image s hows extensive and dif fuse enlargement of junctional zon e (arrowheads) with ill- defined contours signaling diffuse adenomyosis.
B, On axial image, j unctional zone thick ening (arrowheads) is associated w ith well -circumscribed leio myoma (arrow ) in frank hypointense signal on T2-weighted imaging.
B
B
Downloaded from www.ajronline.org by 158.46.162.69 on 01/07/19 from IP address 158.46.162.69. Copyright ARRS. For personal use only; all rights reserved
AJR :196, May 20 11 1211
MRI Characteristics of the Uterine Junctional Zone
To begin, a linear relationship does not ex-
ist between the thickening of the junctional
zone and the histologic diagnosis of adeno-
myosis or between the presence of dysmen-
orrhea and the histologic diagnosis of adeno-
myosis [49]. Second, histologic correlations
reveal that junctional zone thickness is often
more considerable than the degree of pene-
tration of the endometrial glands within the
myometrium [6]. This observation is the ba-
sis for the hypothesis that there is an anteced-
ent event within the endometrium-myometri-
um interface that triggers the appearance of
adenomyosis. Despite these uncertainties,
the threshold of 12 mm should still be stan-
dard given its excellent specificity in estab-
lishing the diagnosis of adenomyosis.
Junctional zone differential—In 2001, Due-
holm et al. [46] introduced the concept of the
junctional zone differential. The junctional
zone differential is calculated by measur-
ing the difference in maximal and minimal
thicknesses in both the anterior and posterior
portions of the uterus. In their study, a junc-
tional zone differential of more than 5 mm
was a more reliable measure in diagnosing
adenomyosis than using a junctional zone
thickness of greater than 12 mm. This crite-
rion however has not been evaluated in fur-
ther studies; thus, the value of this measure
needs to be confirmed.
Ratio of junctional zone thickness and
myo metrium thickness—During analysis of
pelvic MRI, there is sometimes a dispropor-
tion in the overall thickness of the different
uterine layers. Thus, there can be a junction-
al zone at less than 12 mm that is neverthe-
less relatively enlarged when compared with
the other thinner uterine layers. Reinhold et
al. [35] in 1996 as well as Bazot et al. [15]
in 2001 studied the relationship between the
thickness of the junctional zone relative to the
thickness of the entire myometrium measured
at the same place; that is, the ratio of junction-
al zone thickness to myometrium thickness.
Reinhold et al. [35] found a significant differ-
ence in this ratio between patients with ade-
nomyosis (0.69) and a control group (0.44).
However, in that study, no relative threshold
expressed as a ratio performed better in di-
agnosing adenomyosis than a simple absolute
threshold of 12 mm [35]. Bazot et al. [15] had
similar findings with a ratio of 40% allowing
a diagnosis of adenomyosis with a sensitivity
of 65% and a specificity of 92%.
Other signs—A focal thickening of the
junctional zone of greater than 12 mm is a
strong indicator of adenomyosis [15, 32, 38,
43, 45, 46]. It is necessary, however, in these
cases to eliminate the possibility of a uterine
contraction as the cause of this thickening by
repeating T2-weighted imaging several min-
utes after the initial scan. Another solution
to eliminate contractions is to inject an anti-
peristaltic drug [27]. A blurred interface of
the junctional zone with the outer myometri-
um is another possible sign to consider, al-
though it is rather subjective. When the de-
gree of adenomyosis is significant, the uterus
may be globally enlarged, but this sign is not
very sensitive or specific [15]. Finally, hyper-
intense linear striations radiating from the
endometrium toward the myometrium may
give a false impression of enlargement of the
endometrium (Fig. 10). This sign is called
“pseudowidening of the endometrium” [43].
Performance of MRI and Its Limits
in Diagnosing Adenomyosis
Few studies exist that have used a rigorous
surgical reference when evaluating MRI per-
formance for the diagnosis of adenomyosis.
The studies that have been performed show a
sensitivity of from 70% to 86% and a speci-
ficity of 86–93%, with a mean accuracy of
87.5% [15, 35, 46]. These numbers are simi-
lar to ultrasound studies performed on adeno-
myosis [15, 32, 38, 46], and current opinion
remains divided about which is the diagnostic
test of choice. MRI does have the advantage
of less interoperator variability and also per-
mits a more specific diagnosis. Uterine leio-
myomas are present in almost 50% of cases
involving adenomyosis of the uterus, render-
ing ultrasound analysis more difficult and
less efficient [15, 39, 42]. In these situations,
the recourse to MRI is useful especially when
conservative treatment has been chosen. MRI
however has certain limitations.
As we detailed earlier, the less-than-ide-
al sensitivity of MRI in detecting the direct
signs of adenomyosis needs emphasis. Micro-
A
Fig. 9 —Multipar ous 45-year-old woman.
A, Image s hows consid erable focal t hickening of ju nctional zo ne forming mass of ill -defined shapes (arrow). N ote opacification of vagina (star).
B, In corona l plane, this co nsolidate d mass deforms cont ours of ret roflexed uterus an d results i n hypertrophy of anter ior uterin e wall (arrow ). St ar indicates v agina.
B
Downloaded from www.ajronline.org by 158.46.162.69 on 01/07/19 from IP address 158.46.162.69. Copyright ARRS. For personal use only; all rights reserved
1212 AJR :196, May 20 11
Novellas et al.
cysts are detected in only 50% of cases. This
is most likely due to insufficient spatial reso-
lution with current MRI technology.
The second limitation of MRI concerns
the evaluation of a fundamental indirect
sign: the thickness of the junctional zone.
From 20% to 30% of patients will not have
a measurable junctional zone during their re-
productive cycle and this percentage climbs
to 50% in postmenopausal women [6, 15,
39]. This failure rate has not been complete-
ly elucidated: In some patients, the junction-
al zone is not visible and in others it may not
be distinguishable from the outer myome-
trium; finally, the presence of myomas can
render the junctional zone measurement dif-
ficult if not impossible. Reinhold et al. [35]
did not identify any patient with adenomy-
osis and a nonmeasurable junctional zone.
However, Bazot and colleagues [15] report-
ed 22% of patients having proven adenomy-
osis and a nonmeasurable junctional zone.
In one study, investigators evaluated sever-
al different types of T2-weighted sequences:
true fast imaging with steady-state free pre-
cession, turbo inversion recovery, and turbo
spin-echo T2 [50]. They found that sequenc-
es performed with breath-holding improved
the ability to determine the thickness of the
junctional zone and reduced interobserver
variability [50].
Conclusion
The junctional zone constitutes a func-
tionally and morphologically unique entity
within the uterus. MRI was the first imag-
ing technique to reveal this structure and re-
mains a standard tool in its evaluation. The
junctional zone may be invaded by endome-
trial cells that extend into the uterine mus-
cle. Adenomyosis thus defined can be confi-
dently diagnosed when the altered junctional
zone thickness is greater than 12 mm. This
criterion is however not measurable in all pa-
tients. Fast breath-hold T2-weighted MRI
might be useful to improve performance in
measuring the junctional zone to diagnose
uterine adenomyosis.
Acknowledgment
We thank Pierre Vahala for assistance
with the iconography.
References
1. Hrica k H, Alpers C, Crooks LE, Sheldon PE .
Magnetic resonance imaging of the female pelvis:
initial experience. AJR 19 83 ; 141:1119 –112 8
2. Lee JK, Gersell DJ, Balfe DM, Worthington JL,
Picus D, Gapp G. The ut erus: in vitro MR-ana-
tomic correlation of norma l and abnorm al speci-
mens. Radiology 198 5; 157:175 –179
3. Holliday J, Saxon R, Luf kin R B, et al. Anatomic
correlations of magnetic resonance images with
cadaver cryosections. RadioGraphics 1985; 5:
887–921
4. Ar rivé L, Guinet C, Buy JN, Malbec L, Vadrot D,
Laval-Jeantet M. Magnetic resona nce imagi ng of
the pelvic area: prelim inary results [in French]. J
Radiol 1985; 66:771–777
5. Bryan PJ, Butler HE, Li Puma JP. Magnetic reso -
nance imaging of the pelvis. Radiol Clin Nor th
Am 1984 ; 22 :897– 915
6. Fusi L, Cloke B, Brosens JJ. The uter ine junction-
al zone. Best Pract Res Clin Obs tet Gynaecol
2006; 20:479–491
7. Scoutt LM, Flynn SD, Luthringer DJ, McCauley
TR, McCa rthy SM. Junctional zone of the uter us:
correlation of MR imaging and histologic exam i-
nation of hysterectomy specimens. Radiology
1991; 179:403–407
8. McCarthy S, Scott G, Majumdar S, et al. Uterine
junctional zone: MR study of water content and re-
laxation properties. Radiology 1989; 171:241–243
9. Mitchell DG, Schonholz L, Hilper t PL, Pennell
RG, Blum L, Rif kin MD. Zones of the uterus: dis-
crepancy between US and MR images. Radiology
199 0; 174 :827– 831
10. Brown HK, Stoll BS, Nicosia SV, et al. Uteri ne
junctional zone: correlation between histologic
findings and MR imaging. Radiology 1991; 179 :
409–413
11. Demas BE, Hricak H, Jaffe RB. Uterine MR im-
aging: effects of hormonal st imulation. Radiology
1986; 159 :123–126
A
Fig. 10 —36 -year-old w oman who under went MRI evaluat ion for polymyom atous uterus disco vered on ult rasound.
A, On sagit tal T2- weighted image, pos terior my oma (black arrow) is v isible in fr ank hyposig nal. Focal aden omyosis (white arrow) is visible on an terior ut erine wall,
prese nting as ill -defined ar ea in hyposig nal-cont aining cysts. Hyperintens e linear or ret iculated st reaks ext end from end ometrium i nto markedly thick ened anteri or
myome trium (arrowheads).
B, In axial plane, t hese striation s (arrowhead) are l ocated within are a of adenomyo sis (arrow).
B
Downloaded from www.ajronline.org by 158.46.162.69 on 01/07/19 from IP address 158.46.162.69. Copyright ARRS. For personal use only; all rights reserved
AJR :196, May 20 11 1213
MRI Characteristics of the Uterine Junctional Zone
12. Masui T, Katayama M, Kobayashi S, et al. Chang-
es in myometria l and junctional zone thick ness
and signal intensity: demonstration with kinemat-
ic T2-weighted MR i maging. Radiology 20 01;
221:75 –8 5
13. McCarthy S, Tauber C, Gore J. Female pelvic
anatomy: MR assessment of variations during the
menstrual cycle and with use of oral contracep-
tives. Radiology 1986; 160:119–123
14. Brosens JJ, de Souza NM, Barker FG. Uterine
junctional zone: function and disease. Lancet
1995; 346:558–560
15. Bazot M, Cortez A, Darai E, et al. U ltrasonogra-
phy compared with magnetic resonance imaging
for the diagnosis of adenomyosis: correlation with
histopat hology. Hum Reprod 2001; 16:2427–2433
16. Hauth EA , Jaeger HJ, Libera H, Lange S, Forsting
M. MR imag ing of the uterus an d cervix in hea lthy
women: determination of nor mal values. Eur Ra-
diol 2007; 17:734–742
17. Willms AB, Brown ED, Kettritz U I, Kuller JA,
Semelka RC. Anatomic changes in t he pelvis after
uncomplicated vaginal delivery: evaluation with
serial M R imaging. Radiology 1995; 195:91–94
18. Pellerin M, Colau JC, Rebibo G, Savey L. An
MRI study of the normal pelvis in the im mediate
postpartum per iod [in French] . J Radiol 1999;
80: 579–584
19. Janus CL, Wiczyk HP, Laufer N. Magnetic reso-
nance imaging of the menst rual cycle. Magn
Reson Imaging 1988 ; 6 :6 69 –674
20. Wiczyk HP, Janus CL, Richards CJ, et al. Com-
parison of magnetic resonance imaging and ultra-
sound in evaluat ing follicula r and endometr ial
development throughout the norma l cycle. Ferti l
Steril 198 8; 49:9 69–972
21. Brosens JJ, Barker FG, de Souza N M. Myometria l
zonal differentiation a nd uterine ju nctional zone
hyperplasia in the non-preg nant uterus. Hum Re-
prod Upda te 1998; 4:496–502
22. Noe M, Kunz G, Herbertz M, Mall G, Leyendeck-
er G. The cyclic patt ern of the im munocytochem-
ical expression of oestrogen and progesterone re-
ceptors in hu man myometrial and endometrial
layers: chara cterization of the endometrial-suben-
dometria l unit. Hum Reprod 199 9; 14:190 –197
23. Kunz G, Noe M, Herbertz M, Leyendecker G.
Uterine peristalsis during the follicular pha se of
the menstr ual cycle: effects of oestrogen, ant ioes-
trogen and oxytocin. Hum Reprod Update 1998;
4:647– 654
24. Birnholz JC. Ultrasonic visualizat ion of endome-
trial movements. Fertil Steril 1984; 41:157–158
25. Chalubinski K, Deutinger J, Bernaschek G. Vagi-
nosonography for re cording of cycle-related myo-
metria l contractions. Fertil Steril 1993; 59:225 –
228
26. Ijland M M, Evers JL, Dunselman GA, van Katwijk
C, Lo CR, Hoogland HJ. Endometrial wavelike
movements during the menstr ual cycle. Fert il
Steril 19 96 ; 65 :746 –749
27. Naka i A, Koyama T, Fujimoto K, Togashi K.
Functional MR imaging of the uterus. Magn
Reson Imagi ng Clin N Am 2008; 16:673–684, ix
28. Nakai A, Togashi K, Yamaoka T, et al. Uterine
peristalsis shown on cine M R imaging usi ng ul-
trafast sequence. J Magn Reson Imaging 2003;
18:726 –7 33
29. Kido A, Togashi K, Kataoka M, et al. T he effect
of oral contraceptives on uter ine contractility and
menstrual pain: an assessment wit h cine MR im-
aging. Hum Reprod 2007; 22:2066 –2071
30. Kunz G, Beil D, Deini nger H, Wildt L, Leyen-
decker G. The dynamics of rapid sp erm transport
through the female genital tract: evidence from
vaginal son ography of uter ine peristalsis a nd hys-
terosalpingoscintigraphy. Hum Reprod 1996 ; 11:
627–632
31. Ijland MM, Evers JL , Dunselma n GA, Volovics
L, Hoogland HJ. Relation between endometrial
wavelike activity a nd fecundabil ity in sponta ne-
ous cycles. Fertil Steril 1997; 67:492 –496
32. Reinhold C, Tafazoli F, Mehio A, et al. Uterine
adenomyosis: endovaginal US and MR imaging
features with histopathologic correlation. Radio-
Graphics 19 99; 19[ sp ec no]: S147– S160
33. Siegler AM, Camilien L. Adenomyosis. J Reprod
Med 1994; 39:841–853
34. Fernandez H, Donnadieu AC. Adenomyosis [in
French]. J Gynecol Obstet B iol Reprod (Paris)
2007; 36:179–185
35. Rein hold C, McCar thy S, Bret PM, et al. Diffuse
adenomyosis: compa rison of endovaginal US and
MR imag ing with histopathologic co rrelation. Ra-
diology 19 96 ; 199 :151–158
36. Bird CC, McElin TW, Manalo-Estrella P. The elu-
sive adenomyosis of the uterus: revisited. A m J
Obstet Gynecol 1972; 112: 583–593
37. Panganama mula UR, Ha rmanli OH, Isik-Akbay
EF, Grotegut CA, Da ndolu V, Ga ughan JP. Is prior
uterine surgery a risk factor for adenomyosis? Ob-
stet Gynecol 20 04 ; 104 :1034–1038
38. Kinkel K, Vincent B, Balleyguier C, Helenon O,
Moreau J. Value of MR imaging in the diagnosis
of benign uterine conditions [in French]. J Radiol
2000 ; 81:773–779
39. Taourel P, Laffargue G, Dechaud H. Adenomyo-
sis: what imaging modal ity in the diagnosis and
staging [in French]? Gynecol Obstet Fertil 2004 ;
32:976– 980
40. Reinhold C, Tafazoli F, Wang L. Imaging feat ures
of adenomyosis. Hum Repro d Update 1998; 4 :
337–349
41. Ascher SM, Ar nold LL, Patt R H, et al. Adenomy-
osis: prospe ctive comparison of MR imagi ng and
transvaginal sonography. Radiology 1994; 19 0:
803–806
42. Togashi K, Ozasa H, Konishi I, et a l. Enlarged uter-
us: differentiation between adenomyosis and leio-
myoma with MR imaging. Radiology 1989; 171:
531–534
43. Tamai K, Koyama T, Umeoka S, Saga T, Fujii S,
Togashi K. Spectr um of MR features in adenomy-
osis. Bes t Pract Res Clin Obstet Gynaecol 2006;
20:583–602
44. Togashi K, Nishimura K, Itoh K, et al. Adenomyo-
sis: diagnosis with MR imaging. Radiology 1988 ;
166 :111–114
45. Byun JY, Kim SE, Choi BG, Ko GY, Jung SE,
Choi KH. Di ffuse and focal adenomyosis: MR
imaging findings. RadioGraphics 19 99; 19[ spe c
no ]: S161– S170
46. Dueholm M, Lundorf E , Hansen ES, Sorensen
JS, Ledertoug S, Olesen F. Magnetic resonance
imaging a nd transvaginal u ltrasonography for the
diagnosis of adenomyosis. Fertil Steril 20 01; 76:
588–594
47. Kang S, Turner DA, Foster GS, Rapoport MI,
Spencer SA, Wang JZ. Adenomyosis: specificity
of 5 mm as the maximum nor mal uter ine junc-
tional zone t hickness in MR images. AJR 19 96;
166 :114 5–1150
48. Tamai K, Togashi K, Ito T, Morisawa N, Fujiwara
T, Koyama T. MR imaging findings of adenomy-
osis: correlation with histopathologic featu res
and diagnostic pitfalls. Ra dioGraphics 2005;
25:21– 40
49. Kissler S, Zangos S, Kohl J, et al. D uration of dys-
menorrhoea and extent of adenomyosis visualised
by magnetic resonance imaging. Eur J Obstet Gy-
necol Reprod B iol 2008; 137:204–209
50. Bazot M, Daraï E, Clément de Givr y S, Boud-
ghène F, Uzan S, Le Blanche A F. Fast breath-hold
T2-weighted MR i maging reduces interobserver
variability in the diagnosis of adenomyosis. AJR
2003; 180:1291–1296
FOR YOUR INFORMATION
This article is available for CME credit. See www.arrs.org for more information.
Downloaded from www.ajronline.org by 158.46.162.69 on 01/07/19 from IP address 158.46.162.69. Copyright ARRS. For personal use only; all rights reserved
