WOMEN’S IMAGING 781 Physiologic, Histologic, and ......During the second trimester, the placenta...

Note: This copy is for your personal non-commercial use only. To order presentation-ready copies for distribution to your colleagues or clients, contact us at www.rsna.org/rsnarights.

781WOMEN’S IMAGING

Mark A. Sellmyer, MD, PhD • Terry S. Desser, MD • Katherine E. Maturen, MD • R. Brooke Jeffrey, Jr, MD • Aya Kamaya, MD

Retained products of conception (RPOC) are a common and treatable complication after delivery or termination of pregnancy. The pathologic diagnosis of RPOC is made based on the presence of chorionic villi, which indicates persistent placental or trophoblastic tissue. In the setting of postpartum hemorrhage, however, distinguishing RPOC from bleed-ing related to normal postpartum lochia or uterine atony can be clinically challenging. Ultrasonographic (US) evaluation can be particularly helpful in these patients, and a thickened endometrial echo complex (EEC) or a discrete mass in the uterine cavity is a helpful gray-scale US finding that suggests RPOC. However, gray-scale US findings alone are inadequate for accurate diagnosis. Detection of vascularity in a thickened EEC or an endometrial mass at color or power Doppler US increases the posi-tive predictive value for the diagnosis of RPOC. Computed tomography or magnetic resonance imaging may be helpful when US findings are equivocal and typically demonstrates an enhancing intracavitary mass in patients with RPOC. Diagnostic pitfalls are rare but may include highly vascular RPOC, which can be mistaken for a uterine arteriovenous malformation; true arteriovenous malformations of the uterus; invasive moles; blood clot; and subinvolution of the placental implantation site. ©RSNA, 2013 • radiographics.rsna.org

Physiologic, Histologic, and Imaging Features of Retained Products of Conception1

ONLINE-ONLY SA-CME

See www.rsna .org/education

/search/RG

LEARNING OBJECTIVES

After completing this journal-based SA-

CME activity, partic-ipants will be able to:

■ Discuss normal placental architec-ture and develop-ment.

■ List the causes of primary and sec-ondary postpartum hemorrhage.

■ Describe the US appearance of re-tained products of conception.

Abbreviations: AVM = arteriovenous malformation, EEC = endometrial echo complex, hCG = human chorionic gonadotropin, H-E = hematoxy-lin-eosin, PPH = postpartum hemorrhage, PPV = positive predictive value, RPOC = retained products of conception

RadioGraphics 2013; 33:781–796 • Published online 10.1148/rg.333125177 • Content Codes: 1From the Department of Radiology (M.A.S., T.S.D., R.B.J., A.K.), Stanford University School of Medicine, 300 Pasteur Dr, H1307, Stanford, CA 94305; and Department of Radiology, University of Michigan School of Medicine, Ann Arbor, Mich (K.E.M.). Presented as an education exhibit at the 2011 RSNA Annual Meeting. Received August 16, 2012; revision requested September 14 and received November 4; accepted December 19. For this journal-based SA-CME activity, the authors T.S.D. and R.B.J. have disclosed financial relationships (see p 795); all other authors, the editor, and reviewers have no relevant relationships to disclose. Address correspondence to A.K. (e-mail: [email protected]).

©RSNA, 2013

782 May-June 2013 radiographics.rsna.org

IntroductionPostpartum bleeding is a component of the nor-mal physiologic process that occurs after delivery or miscarriage. Normally, bleeding decreases over the first postpartum week. When bleeding is ex-cessive and uterine atony has been excluded as a cause, the next most common cause of postpar-tum bleeding is retained products of conception (RPOC). The placenta usually constitutes the ma-jority of RPOC, and, because of its highly vascular connection to the uterus, it provides a conduit for continued bleeding. Although accurate diagnosis of RPOC is a challenge both clinically and radio-logically, it is important for guiding proper man-agement. In this article, we discuss the physiology of placental development; the clinical, pathologic, and imaging diagnosis of RPOC; and some poten-tial pitfalls that may mimic RPOC.

Normal Physiologic Development of the Placenta

Maternal-fetal interaction is a dramatically complex physiologic process. This interaction is mediated primarily by the interdigitation of the fetal placenta with the decidua and endome-trium in the maternal uterus (Fig 1). The fetus is dependent on the placenta for pulmonary, hepatic, and renal functions. These functions are maintained by the indirect vascular interaction of fetal blood with maternal blood via the utero-placental vessels, allowing the exchange of gas, metabolites, and nutrients. Anatomically, this indirect interaction is characterized by the prox-imity of the fetal circulation contained within the villous tree to the maternal circulation con-tained within the intervillous space. As gestation progresses, the villous branching architecture becomes more complex and the villi become smaller in diameter. This architectural change maximizes the surface area of villi exposed to maternal blood within placental lacunae. At term, very little tissue separates the engorged fetal capillaries from the maternal blood (1).

At ultrasonography (US), the first sign of pregnancy before placental development is the presence of a gestational sac. The decidua cap-sularis covers the developing gestational sac, and the decidua parietalis covers the remainder of the uterine cavity. These two echogenic layers form the “double sac sign” in early pregnancy (Fig 2). The gestational sac grows approximately 1 mm per day for the first 8 weeks, during which time the embryo and the trophoblast continue to grow. The placenta is formed from the de-cidua basalis, which arises from the maternal uterine tissue, and the chorion frondosum, which arises from the embryonic villous tropho-blast. Toward the end of the first trimester, the developing placenta grows more rapidly than the fetus and contains convex lobes separated by placental septa (2).

Figure 1. Drawing illustrates the maternal-fetal unit, with the fetus attached to the uterus by the placenta. (Courtesy of Amy Morris, Stanford University.)

RG • Volume 33 Number 3 Sellmyer et al 783

By the beginning of the second trimester at week 13, the placenta comes in direct contact with maternal blood. The placenta is visible at US as a homogeneously echogenic, crescen-tic solid structure located along the wall of the uterine cavity (Fig 3) (3). During the second trimester, the placenta continues to grow to meet the increased fetal metabolic requirements. Its vascularity changes as well, with increased flow velocity in the uterine artery, development of continuous intervillous flow, and the appearance of end-diastolic flow in the umbilical artery by 14 weeks gestation (4). These changes are thought to be related to invasion of the uterine spiral arteries by extravillous trophoblast cells, which cause the arteries to lose smooth muscle cells and elastic lamina. This results in dilatation of the uterine spiral arteries, allowing increased uterine blood flow at reduced pressures (5).

In the third trimester, the placental villi are dominated by thin-walled capillaries, and the pla-centa undergoes trophotropism, the process of pla-cental migration to areas of robust maternal blood supply, to best maintain the fetus. Consequently, the vascular supply and structural integrity of the

placenta vary considerably during gestation. This variability can potentiate numerous disease entities, including placental abruption, placenta accreta, succenturiate lobes, and bilobed placenta, which in turn can increase the risk of incomplete expul-sion of the placenta after parturition. By the third trimester, the blood supply to the fetoplacental unit is 10 times that of the nonpregnant uterus (5).

Clinical and Patho - logic Diagnosis of RPOC

Patients with RPOC often present with postpar-tum hemorrhage (PPH), which is divided into primary and secondary types. Primary PPH is defined as blood loss greater than 500 mL in the first 24 hours, whereas secondary PPH is exces-sive blood loss between 24 hours and 6 weeks postpartum. RPOC is one of the most common causes of both primary and secondary PPH, with risk factors including failure to progress during delivery, placenta accreta, and instru-ment delivery (6). In addition to bleeding, pa-tients may also present with pain or fever. When RPOC manifests as secondary PPH, it may need

Figure 2. Gestational sac. Sagittal US image of the uterus shows an early intrauterine pregnancy with the double sac sign, formed by the decidua capsularis (ar-row) and the decidua parietalis (arrowhead).

Figure 3. Early placental development. Transverse US image of the uterus obtained at 13 weeks gestation shows a developing first-trimester placenta (arrow).


to be distinguished from endometritis, uterine dehiscence or perforation, and, rarely, subinvo-lution of the placental implantation site.

The term retained products of conception refers to intrauterine tissue that develops after concep-tion and persists after delivery or termination of pregnancy. This intrauterine tissue is often of placental trophoblastic origin, since the placenta invades and attaches to the uterine endometrium. The trophoblast forms numerous processes known as chorionic villi that invade the decidua basalis of the endometrium. The key to the mi-croscopic diagnosis of RPOC is the presence of chorionic villi, which indicates the persistence of placental tissue (Fig 4). Each villus is a fingerlike projection of placental tissue containing a core of mesenchymal tissue perfused by fetal capillaries (1). After delivery or termination of pregnancy, the remaining villi exhibit variable vascularity and increasing fibrosis (Fig 5), which may account for the observed variability in the vascularity of re-tained placental tissue at Doppler US (7).

The reported incidence of RPOC seems to de-pend on the gestational age of the pregnancy, with RPOC occurring most frequently after second-trimester delivery or termination of pregnancy.

In one prospective study, RPOC was diagnosed after a third-trimester delivery in 2.7% of women, whereas it was diagnosed after pregnancies end-ing during the second and first trimesters in 40% and 17%, respectively (8). The incidence of RPOC after routine vaginal delivery is thought to be ap-proximately 3%–5% (9). Given that over 4.3 mil-lion births occurred in the United States in 2007, RPOC is a common complication of pregnancy, and suspicion for RPOC accounts for a large num-ber of postpartum US examinations (10).

Laboratory values such as white blood cell count and human chorionic gonadotropin (hCG) level may not be helpful in the diagnosis of RPOC because they can be elevated in the nor-mal postpartum setting. However, color Doppler US is an important tool in the accurate diagnosis of RPOC (11). Early diagnosis is critical for di-recting clinical management of bleeding and for preventing associated immediate complications such as perforation or infection, as well as future obstetric complications.

Once RPOC is diagnosed on the basis of clini-cal, laboratory, and US findings, several treatment options are available, including expectant man-

Figure 4. Photomicrograph (original magnification, ×4; hematoxylin-eosin [H-E] stain) of a suction curet-tage specimen obtained 1 month after missed sponta-neous abortion at 8 weeks gestation shows innumer-able fingerlike chorionic villi (arrows) intermixed with maternal blood (B) and fragments of uterine decidua (D). Trophoblasts are seen infiltrating the decidua (ar-rowheads) at the implantation site.

Figure 5. Photomicrograph (original magnification, ×20; H-E stain) of a vacuum aspiration specimen obtained 1 month after missed spontaneous abortion at 8 weeks gestation shows villi with extensive central fibrotic changes (pink) instead of the loose mesenchy-mal stroma and prominent capillary vascularity seen in normal villi.


agement, use of uterotonic medications such as prostaglandin E1 analogs, and surgical interven-tions such as dilation and curettage or hystero-scopic removal (Fig 6) (12,13). Given the risks associated with surgical interventions, including perforation, infection, Asherman syndrome, and development of scar tissue, any of which can have a long-term negative impact on future pregnan-cies, accurate diagnosis is vital (14).

US Diagnosis of RPOCUS is a useful diagnostic tool for the triage of pa-tients with PPH and is more accurate than clini-cal presentation alone for the diagnosis of RPOC (11,15). Combined gray-scale–color Doppler US

is the first-line imaging modality for the diagnosis of suspected RPOC and allows real-time assess-ment of the uterine structures and blood flow, whereas magnetic resonance (MR) imaging can be a useful adjunct imaging modality in compli-cated cases (Fig 7). The sensitivity and specificity of US for the diagnosis of RPOC varies widely based on the diagnostic criteria and the clinical setting. However, more recent studies have de-fined more robust gray-scale and color Doppler US criteria with a high sensitivity and specificity for RPOC (7,8,16–21) (Table).

Figure 6. Hysteroscopic image shows RPOC. (Courtesy of Mary Jacobson, MD, Stanford University.)

Figure 7. Pathologically proved RPOC in a 39-year-old woman. (a) Longitudinal endovaginal color Doppler US image of the uterus shows a hypervascular area in the myometrium (arrow) that extends to the endometrium. (b) Sagittal contrast material–enhanced fat-saturated T1-weighted MR image shows vascularized material in the uterine cavity (arrow).


Fin

din

gs

fro

m S

elec

ted

Stu

die

s fo

r th

e D

iag

no

sis

of

RP

OC

Stu

dy/Y

ear*

Cas

esC

rite

ria

Sen

siti

vity

†S

peci

ficit

y†P

PV

†N

PV

†A

ccur

acy†

Com

men

ts

Atr

i et

al/2

011

(16)

91M

ass

8171

6485

.10

75R

PO

C p

atho

logi

cally

con

firm

ed in

36/

91 p

atie

nts

Vas

cula

rity

9467

6595

.10

78…

P v

alue

0.07

>0.

05>

0.05

0.1

>0.

05…

Kam

aya

et a

l/ 20

09 (

7)26

9M

ass

29…

80…

…A

ll R

PO

C a

nd n

on-R

PO

C p

atho

logi

cally

con

firm

edV

ascu

lari

ty (

type

s 1–

3)96

……

……

No

sign

ifica

nt d

iffe

renc

e in

end

omet

rial

thi

ckne

ss b

e-tw

een

pati

ents

wit

h an

d pa

tien

ts w

itho

ut R

PO

CD

urfe

e et

al/

2005

(19

)16

3M

ass

7989

5995

.10

…F

ocal

ech

ogen

ic/h

eter

ogen

eous

lesi

on in

end

omet

rium

Com

plex

flui

d29

7922

84.1

0…

…T

hick

ened

end

omet

rium

(>

10 m

m)

778

680

.10

……

Neg

ativ

e fin

ding

s0

4728

100.

100

…N

o m

ass,

com

plex

flui

d, e

ndom

etri

al t

hick

ness

<10

mm

Blo

od fl

ow t

o en

dom

etri

um63

6075

46.1

0…

Col

or D

oppl

er U

S p

erfo

rmed

in o

nly

29/1

63 p

atie

nts

Van

den

Bos

ch e

t al

/200

2 (1

8)38

5F

ocal

vas

cula

rity

at

Dop

pler

U

S96

……

……

US

per

form

ed d

urin

g fir

st 2

pos

tpar

tum

wee

ks

Van

den

Bos

ch e

t al

/200

8 (8

)10

70U

teri

ne c

avit

y m

ass

wit

h or

w

itho

ut a

dis

tinc

t va

scul

ar

pedi

cle

or e

nhan

ced

myo

-m

etri

al v

ascu

lari

ty a

t co

lor

Dop

pler

US

98…

……

…E

nhan

ced

myo

met

rial

vas

cula

rity

defi

ned

as a

n ar

ea o

f m

arke

d flo

w o

ver

the

full

thic

knes

s of

the

myo

me-

triu

m, r

each

ing

the

uter

ine

cavi

ty o

n co

lor

Dop

pler

U

S im

ages

Alc

ázar

et

al/

1995

(17

)33

Het

erog

eneo

us E

EC

wit

h

an ir

regu

lar

endo

met

rial

-m

yom

etri

al in

terf

ace

8194

8393

……

Mat

ijevi

c et

al/

2009

(20

)93

PP

H77

3025

74.1

0…

…L

ower

abd

omin

al p

ain

7.4

7987

12.1

0…

…F

ever

2764

6831

.10

……

Cer

vica

l dila

tati

on70

3030

69.1

0…

…G

ray-

scal

e U

S f

eatu

res

9833

1584

.10

…E

ndom

etri

al m

ass

>10

mm

wit

h a

hype

r/hy

poec

hoic

an

d m

ixed

ech

ogen

ic p

atte

rn in

the

ute

rine

cav

ity

Gra

y-sc

ale/

Dop

pler

US

fe

atur

es66

6948

51.1

0…

Gra

y-sc

ale

crit

eria

and

low

-res

isti

ve-i

ndex

(<

0.4

5)

bloo

d flo

wde

Vri

es e

t al

/ 20

00 (

21)

64U

S fi

ndin

gs o

f R

PO

C85

8868

92…

25%

fal

se-p

osit

ive

rate

Not

e.—

EE

C =

end

omet

rial

ech

o co

mpl

ex, N

PV

= n

egat

ive

pred

icti

ve v

alue

, PP

V =

pos

itiv

e pr

edic

tive

val

ue. *

Num

bers

in p

aren

thes

es in

dica

te r

efer

ence

num

bers

. †N

umbe

rs

repr

esen

t pe

rcen

tage

s (w

ith

the

exce

ptio

n of

P v

alue

s).


Gray-Scale US Findings of RPOCWe have found that the most sensitive finding of RPOC at gray-scale US is a thickened endome-trial echo complex (EEC) (Figs 8, 9). The exact definition of “thickened” varies in the literature, ranging from 8 to 13 mm (22,23). If a patient

at our institution is clinically suspected of hav-ing RPOC, we use 10 mm as the cut-off value, which has a reported sensitivity of over 80% (7,15,23,24). However, the specificity of this measurement is relatively low (20%) because a thickened EEC can be observed in postpartum patients with no RPOC (Fig 10) (17,23). On the other hand, if the EEC is less than 10 mm in thickness, the negative predictive value for RPOC is 63%–80% (17,23). In practice, how-ever, a diagnosis of RPOC is not made solely on the basis of gray-scale US findings; color Dop-pler US findings are critical in further refining the diagnosis.

Another gray-scale US finding that increases the likelihood of RPOC is the presence of an endometrial or intrauterine mass (16,17). A sen-sitivity of up to 79% has been reported for this finding (17). In a previously published study (7), we found this finding (when present) to have a sensitivity of only 29% but a high posi-tive predictive value (PPV) of 80% for RPOC. A potential explanation for the differences in sensitivity between various studies may be the varying definitions of an intrauterine mass, which we defined as a structure separate from the endometrium that is distinguishable in two orthogonal planes (Fig 11).

Figure 8. Pathologically proved RPOC with a thick-ened EEC in a 29-year-old postpartum woman. Lon-gitudinal transabdominal gray-scale US image shows a markedly thickened (26-mm) EEC (arrows).

Figure 9. Pathologically proved RPOC with a thick-ened EEC in a 40-year-old woman. Longitudinal endovaginal US image shows a mildly thickened (12-mm) EEC (cursors).

Figure 10. Pathologically proved organizing blood clot with no RPOC in a 30-year-old woman with vagi-nal bleeding. Longitudinal transabdominal color Dop-pler US image obtained 8 days postpartum shows an avascular, thickened (33-mm) EEC (arrows).


Color Doppler US Findings of RPOCAs mentioned earlier, color Doppler US further enhances diagnostic confidence in identifying RPOC. For example, blood clots will appear avas-cular at color Doppler US, whereas the detection of vascularity in a thickened EEC or endometrial mass is likely to represent RPOC. Compared with clinical presentation alone, color Doppler

US increases the likelihood of predicting RPOC by approximately a factor of 2 (20). The degree of vascularity in a thickened EEC or endometrial mass can help increase diagnostic confidence for the presence of RPOC (7,16). To better describe the typical color Doppler US findings of RPOC, the degree of vascularity of the endometrial com-ponent can be compared with the myometrial vas-cularity in the same image section and graded as type 0, 1, 2, or 3 (Fig 12) (7).

Figure 11. Pathologically proved RPOC in a 22-year-old woman with vaginal bleeding. Transverse (a) and longitudinal (b) endovaginal gray-scale US images of the uterus obtained 2 weeks postpartum clearly depict a 3.8 × 1.8 × 2.0-cm echogenic mass (arrows) in the lower uterine segment and cervix.

Figure 12. Drawings illustrate RPOC with various types of vascularity (types 0–3). The degree of vascularity is measured by comparing endometrial with myometrial blood flow at color Doppler US. (Courtesy of Amy Morris, Stanford University.)


Figure 13. Pathologically proved RPOC with type 0 vascularity in a 42-year-old woman with persistent bleeding 2 weeks after a miscarriage (22 weeks gestation). (a) Longitudinal endovaginal US image shows persistent thickening (28 mm) of the EEC (arrows) and distention of the endometrial canal by heterogeneous echogenic material. (b) Longitu-dinal color Doppler US image shows avascularity of the endometrium.

Figure 14. Pathologically proved RPOC with type 1 vascularity in a 40-year-old woman with passing blood clots and an open cervical os. (a) Longitudinal transabdominal US image shows a thickened (39-mm) EEC (arrows). (b) Longi-tudinal color Doppler US image shows a small focus of color flow within the EEC (arrow).

Type 0 vascularity, defined as no detectable vascularity in a thickened EEC or mass, may represent either a blood clot or avascular RPOC (Fig 13) (7). Although an avascular thickened endometrium could still represent RPOC, avas-cular RPOC will probably pass spontaneously without intervention and is unlikely to cause severe bleeding.

If any vascularity is detected in a thickened EEC or mass, the likelihood of RPOC increases substantially to a PPV of 96%. Three vascularity patterns have been shown to increase diagnostic confidence and help direct clinical management

of RPOC: type 1 (Fig 14), or minimal vascular-ity (less than that of the myometrium), which has a PPV greater than 90%; type 2 (Fig 15), or moderate vascularity (nearly equal flow in the endometrium and myometrium), which has been shown to have a PPV of 100% (7); and type 3 (Fig 16), or marked endometrial vascu-larity (greater than that of normal myometrium in the same image section), which also has a PPV of 100%. Of note, type 3 vascularity can be so robust that it mimics an AVM at color


Doppler US. The arterial flow velocity of type 3 RPOC can be 100 cm/sec or higher, with a very low-resistance spectral waveform and large ves-sels occasionally being evident (7). Vascularity should always be seen extending from the myo-metrium into the endometrium. If vascularity is isolated to the myometrium, other diagnoses besides RPOC should be considered (see “Pit-falls”). Given the marked vascularity of type 3

Figure 16. RPOC with type 3 vascularity in a 22-year-old woman with a 1-week history of vaginal bleeding who had had a miscarriage 2 months earlier. (a) Longitudinal endovaginal color Doppler US image shows a thickened EEC with marked vascularity involving both the endometrium and the myometrium. (b) Color Doppler US image with spectral waveform shows marked vascularity, with velocities of up to 79 cm/sec. This marked vascularity should not be mistakenly attributed to an arteriovenous malformation (AVM) given the clear extension of vascularity to the endometrium; the presence of marked vascularity should, however, be conveyed to the obstetrician.

RPOC, the obstetrician should be notified of this finding due to the theoretic risk of bleeding if a large vessel inadvertently becomes unroofed during dilation and curettage.

CT and MR Imaging Findings of RPOC

The role of computed tomography (CT) and MR imaging in the evaluation of PPH is limited, but it increases with the amount of time that has elapsed since delivery or pregnancy termination

Figure 15. Pathologically proved RPOC with type 2 vascularity in a 29-year-old woman (G2, P2) who had under-gone a cesarean section 6 weeks earlier. (a) Longitudinal endovaginal gray-scale US image shows a large amount of echogenic material and a thickened EEC (arrows). (b) Longitudinal color Doppler US image obtained in the same area shows endometrial material with moderate vascularity similar to that of the adjacent myometrium.


and with the clinical complexity of the case (25). US is usually sufficient in routine or typical cases of RPOC, but radiologists should also be aware of the characteristic appearance of RPOC at CT and MR imaging. A confident diagnosis of RPOC can be made with either of these modali-ties if an enhancing soft-tissue mass is present within the endometrial canal (Fig 17) (26,27). Without precontrast images, enhancement can be difficult to distinguish from high-attenuation clot at CT. Furthermore, enhancement may be partial or delayed (27). Postcontrast MR images are very helpful for the assessment of enhance-ment (Fig 18a), but the signal characteristics of RPOC on T2-weighted and precontrast T1-weighted images may vary depending on the degree of hemorrhage and tissue necrosis (Fig 18b). Because the findings of RPOC at MR im-aging in particular overlap entirely with those of gestational trophoblastic disease (28), clini-cal context is essential. Specifically, the serum

b-hCG level is usually normal or low in patients with RPOC and considerably elevated in those with gestational trophoblastic disease.

PitfallsAlthough distinguishing RPOC from blood clots in the postpartum uterus is the most common challenge in this setting, the radiologist should be aware of several other rare but potential pitfalls that can mimic RPOC.

The biggest pitfall in diagnosis is mistaking the marked vascularity of RPOC for an AVM (29,30). The true incidence of uterine AVM is not known, but in our experience, this entity is rare and is likely overdiagnosed because the un-derlying disease is often either RPOC or related to subinvolution of the placental implantation site (31). It is thought that most AVMs develop secondary to uterine tissue injury (eg, from prior

Figure 17. RPOC in a patient with secondary PPH. Axial CT image demonstrates a briskly en-hancing, lentiform mass (arrow) in the anterior uterus. This finding was confirmed at surgery to represent a large component of a retained placenta.

Figure 18. RPOC. (a) Axial postcontrast T1-weighted MR image shows RPOC with brisk enhancement (arrow). (b) Sagittal T2-weighted MR image demonstrates a nodular component of high-signal-intensity soft tissue (arrow) within the endometrial canal.


of these entities may manifest as an endometrial mass with vascularity (Fig 20).

Subinvolution of the placental implantation site is an exceptionally rare postpartum condition in which the uterine vessels fail to involute following delivery (34). The postpartum uterus may appear unusually large, with dilated myometrial vessels (Fig 21) (18). Although its exact cause is unknown, subinvolution of the placental implantation site is thought to be related to an abnormal immunologic recognition process similar to preeclampsia. Risk factors for subinvolution include atony, multiparity, cesarean section, uterine prolapse, uterine fibroids, endometritis, coagulopathies, and RPOC (35).

Figure 19. Presumed AVM in a 30-year-old woman who had undergone dilation and curettage 1 month ear-lier for RPOC after an uncomplicated normal spontane-ous vaginal delivery with a sudden large volume of vagi-nal bleeding. (a) Longitudinal color Doppler US image shows marked vascularity isolated to the myometrium (white arrow) and clot in the endometrial cavity (black arrow). (b) Color Doppler US image with spectral wave-form shows high-velocity, low-resistance flow. (c) Digital subtraction angiogram shows an early draining vein (ar-row). The patient underwent embolization, with resolu-tion of bleeding.

dilation and curettage) rather than as congenital malformations. The standard for AVM diagnosis is angiographic identification of an early drain-ing vein (Fig 19), although color-power Doppler US is increasingly being relied on as a surrogate (31,32). Both AVMs and RPOC are persistent postpartum findings; however, RPOC can be distinguished from AVMs on the basis of the vascular endometrial component seen in RPOC, whereas uterine AVMs primarily involve only the myometrium (31). Proper identification of AVMs is important because of the potential risk of life-threatening hemorrhage, whether spontaneous or as a result of dilation and curettage. Treatment options for uterine AVMs depend on the clinical situation and include uterine artery embolization and expectant management with close monitor-ing (32,33).

Another potential mimic of RPOC is an un-derlying endometrial abnormality, such as an endometrial polyp or submucosal fibroid. Either


Invasive moles may occur as a complication of a complete or partial hydatidiform mole in-vading the uterus, with occasional spread to the lungs or vagina (36). Invasive moles and RPOC may have some overlap of imaging findings, but the clinical picture usually allows differentia-

tion; the presence of a prior molar pregnancy with persistently elevated and increasing hCG levels is vital in making the diagnosis. Pathologic specimens obtained in these patients exhibit

Figure 20. Pathologically proved vascular endometrial polyp in a 23-year-old woman with vaginal bleeding who had undergone medical termination of pregnancy 2 months earlier. (a) Longitudinal gray-scale US image shows an isoechoic mass with small cystic spaces (calipers) within the endometrium. (b) Longitudinal color Doppler US im-age shows marked vascularity within the mass. Pathologic analysis showed only fragments of weakly proliferative and decidualized endometrium, with no RPOC.

Figure 21. Pathologically proved subinvolution of the placental implantation site in a 36-year-old woman who had undergone a cesarean section for twins 1½ weeks earlier. (a) Longitudinal transabdominal US image shows an en-larged uterus with large, tubular hypoechoic areas (arrow) in the myometrium. (b) Longitudinal color Doppler US image demonstrates tubular hypoechoic areas representing large vessels within the myometrium. Pathologic analysis demonstrated ectasia of the uterine veins with subinvolution of the uteroplacental arteries and no evidence of RPOC.


characteristic changes: hydropic villi, inclu-sion of trophoblasts within villi (“trophoblastic islands”), scalloping of villous borders, and cellular atypia (Fig 22) (37). The swollen and relatively hypocellular villi correlate well with

the imaging and gross appearances of molar tis-sue, namely, multiple discrete fluid-filled sacs (Fig 23). At US, an invasive mole can appear as a poorly defined mass with anechoic areas that is located in the uterus. Color Doppler findings

Figure 22. Partial mole. Photomicrograph (orig-inal magnification, ×10; H-E stain) of a suction curettage specimen obtained after spontaneous abortion at 9 weeks gestation shows a markedly hydropic villus (arrows). Mesenchymal cells are highly edematous, with sparse pink cytoplasm (ar-rowheads). An island of isolated trophoblasts (T) lies in the center of the villus. These findings sug-gest a partial molar pregnancy.

Figure 23. Complete mole in an 18-year-old woman at 6 weeks gestation with a b-hCG serum level over 400,000 mIU/mL. (a) Longitudinal transvaginal color Doppler US image demonstrates innumerable anechoic sacs (arrow) representing enlarged hydropic villi filling and distend-ing the endometrial canal. (b) Photomicrograph (original magnification, ×4; H-E stain) shows a large, edematous villus with an acellular central cistern (arrow), findings that are diagnostic for a complete molar pregnancy when seen throughout sampled villi. (c) Photomicrograph (p57 im-munostain) shows the absence of brown (positive) nuclear staining in villous mesenchymal cells (arrow), a finding that suggests paternal disomy, or a complete mole. As ma-ternally expressed protein, p57 should be seen in normal placental tissue.


include increased vascularity, predominantly in the myometrium, resulting from myometrial in-vasion (Fig 24).

ConclusionGiven an appropriate clinical history, the pres-ence of a thickened EEC or an endometrial mass with detectable vascularity at color and power Doppler US allows confident diagnosis of RPOC. Potential pitfalls in this diagnosis are rare but can include uterine AVMs, preexisting endometrial polyps, invasive moles, and subinvolution of the placental implantation site.

Acknowledgments.—We thank Jeslyn A. Rumbold for her editorial assistance; Amy N. Morris for her illustra-tions; Mary Jacobson, MD, for the image in Figure 6; and Abhishek Shukla, MD, Pathology Department, University of Michigan, for pathologic images and consultation.

Disclosures of Conflicts of Interest.—R.B.J.: Related financial activities: none. Other financial activities: con-sultant for Innervision Medical Technologies. T.S.D.: Related financial activities: none. Other financial activi-ties: course developer for Amirsys.

References 1. Burkitt H, Young B, Heath J. Wheater’s functional

histology. London, England: Churchill Living-stone, 1993.

2. Cunningham FG, Leveno KJ, Bloom SL, et al. Wil-liams obstetrics. New York, NY: McGraw-Hill, 2010.

3. Gibbs R, Danforth D, Karlan B, Haney A. Dan-forth’s obstetrics and gynecology. Philadelphia, Pa: Lippincott Williams & Wilkins, 2008.

4. Foidart JM, Hustin J, Dubois M, Schaaps JP. The human placenta becomes haemochorial at the 13th

week of pregnancy. Int J Dev Biol 1992;36(3): 451–453.

5. Whitley GS, Cartwright JE. Cellular and molecu-lar regulation of spiral artery remodelling: lessons from the cardiovascular field. Placenta 2010;31(6): 465–474.

6. Sheiner E, Sarid L, Levy A, Seidman DS, Hallak M. Obstetric risk factors and outcome of pregnan-cies complicated with early postpartum hemor-rhage: a population-based study. J Matern Fetal Neonatal Med 2005;18(3):149–154.

7. Kamaya A, Petrovitch I, Chen B, Frederick CE, Jeffrey RB. Retained products of conception: spec-trum of color Doppler findings. J Ultrasound Med 2009;28(8):1031–1041.

8. Van den Bosch T, Daemen A, Van Schoubroeck D, Pochet N, De Moor B, Timmerman D. Occur-rence and outcome of residual trophoblastic tissue: a prospective study. J Ultrasound Med 2008;27(3): 357–361.

9. Epperly TD, Fogarty JP, Hodges SG. Efficacy of routine postpartum uterine exploration and manual sponge curettage. J Fam Pract 1989;28(2):172–176.

10. Martin JA, Hamilton BE, Sutton PD, et al. Births: final data for 2007. Natl Vital Stat Rep 2010;58 (24):1–85.

11. Malvern J, Campbell S, May P. Ultrasonic scan-ning of the puerperal uterus following secondary postpartum haemorrhage. J Obstet Gynaecol Br Commonw 1973;80(4):320–324.

12. Sotiriadis A, Makrydimas G, Papatheodorou S, Io-annidis JP. Expectant, medical, or surgical manage-ment of first-trimester miscarriage: a meta-analysis. Obstet Gynecol 2005;105(5 pt 1):1104–1113.

13. Golan A, Dishi M, Shalev A, Keidar R, Ginath S, Sagiv R. Operative hysteroscopy to remove retained products of conception: novel treatment of an old problem. J Minim Invasive Gynecol 2011;18(1): 100–103.

Figure 24. Invasive mole in a 40-year-old woman with an elevated hCG level. (a) Longitudinal gray-scale US im-age shows a heterogeneous and indistinct EEC. (b) Longitudinal color Doppler US image shows areas of marked vascularity and invasion of the myometrium.


14. Asherman JG. Amenorrhoea traumatica (atretica). J Obstet Gynaecol Br Emp 1948;55(1):23–30.

15. Achiron R, Goldenberg M, Lipitz S, Mashiach S. Transvaginal duplex Doppler ultrasonography in bleeding patients suspected of having residual trophoblastic tissue. Obstet Gynecol 1993;81(4): 507–511.

16. Atri M, Rao A, Boylan C, Rasty G, Gerber D. Best predictors of grayscale ultrasound combined with color Doppler in the diagnosis of retained prod-ucts of conception. J Clin Ultrasound 2011;39(3): 122–127.

17. Alcázar JL, Baldonado C, Laparte C. The reliabil-ity of transvaginal ultrasonography to detect re-tained tissue after spontaneous first-trimester abor-tion, clinically thought to be complete. Ultrasound Obstet Gynecol 1995;6(2):126–129.

18. Van den Bosch T, Van Schoubroeck D, Lu C, De Brabanter J, Van Huffel S, Timmerman D. Color Doppler and gray-scale ultrasound evaluation of the postpartum uterus. Ultrasound Obstet Gynecol 2002;20(6):586–591.

19. Durfee SM, Frates MC, Luong A, Benson CB. The sonographic and color Doppler features of retained products of conception. J Ultrasound Med 2005;24 (9):1181–1186.

20. Matijevic R, Knezevic M, Grgic O, Zlodi-Hrsak L. Diagnostic accuracy of sonographic and clinical parameters in the prediction of retained products of conception. J Ultrasound Med 2009;28(3):295–299.

21. de Vries JI, van der Linden RM, van der Linden HC. Predictive value of sonographic examination to visualize retained placenta directly after birth at 16 to 28 weeks. J Ultrasound Med 2000;19(1):7–12.

22. Sadan O, Golan A, Girtler O, et al. Role of sonog-raphy in the diagnosis of retained products of con-ception. J Ultrasound Med 2004;23(3):371–374.

23. Ustunyurt E, Kaymak O, Iskender C, Ustunyurt OB, Celik C, Danisman N. Role of transvaginal sonogra-phy in the diagnosis of retained products of concep-tion. Arch Gynecol Obstet 2008;277(2):151–154.

24. McEwing RL, Anderson NG, Meates JB, Allen RB, Phillipson GT, Wells JE. Sonographic appearances of the endometrium after termination of pregnancy in asymptomatic versus symptomatic women. J Ul-trasound Med 2009;28(5):579–586.

25. Lee NK, Kim S, Lee JW, et al. Postpartum hemor-rhage: clinical and radiologic aspects. Eur J Radiol 2010;74(1):50–59.

26. Menias CO, Elsayes KM, Peterson CM, Huete A, Gratz BI, Bhalla S. CT of pregnancy-related com-plications. Emerg Radiol 2007;13(6):299–306.

27. Noonan JB, Coakley FV, Qayyum A, Yeh BM, Wu L, Chen LM. MR imaging of retained products of conception. AJR Am J Roentgenol 2003;181(2): 435–439.

28. Elsayes KM, Trout AT, Friedkin AM, et al. Imag-ing of the placenta: a multimodality pictorial re-view. RadioGraphics 2009;29(5):1371–1391.

29. Fleming H, Ostör AG, Pickel H, Fortune DW. Arteriovenous malformations of the uterus. Obstet Gynecol 1989;73(2):209–214.

30. Müngen E. Vascular abnormalities of the uterus: have we recently over-diagnosed them? Ultrasound Obstet Gynecol 2003;21(6):529–531.

31. Timmerman D, Wauters J, Van Calenbergh S, et al. Color Doppler imaging is a valuable tool for the diagnosis and management of uterine vascular mal-formations. Ultrasound Obstet Gynecol 2003;21 (6):570–577.

32. Syla BH, Fetiu SS, Tafarshiku SS. Transabdominal two- and three-dimensional color Doppler imaging of a uterine arteriovenous malformation. Ultra-sound Obstet Gynecol 2011;37(3):376–378.

33. Kitahara T, Sato Y, Kakui K, Tatsumi K, Fujiwara H, Konishi I. Management of retained products of conception with marked vascularity. J Obstet Gyn-aecol Res 2011;37(5):458–464.

34. Al-Mehaisen L, Al-Kuran O, Amarin ZO, Matalka I, Beitawi S, Muhtaseb A. Secondary postpartum hemorrhage following placental site vessel subin-volution: a case report. Arch Gynecol Obstet 2008; 278(6):585–587.

35. Weydert JA, Benda JA. Subinvolution of the pla-cental site as an anatomic cause of postpartum uterine bleeding: a review. Arch Pathol Lab Med 2006;130(10):1538–1542.

36. Allen SD, Lim AK, Seckl MJ, Blunt DM, Mitchell AW. Radiology of gestational trophoblastic neopla-sia. Clin Radiol 2006;61(4):301–313.

37. Vang R, Gupta M, Wu LS, et al. Diagnostic repro-ducibility of hydatidiform moles: ancillary tech-niques (p57 immunohistochemistry and molecular genotyping) improve morphologic diagnosis. Am J Surg Pathol 2012;36(3):443–453.

This journal-based SA-CME activity has been approved for AMA PRA Category 1 CreditTM. See www.rsna.org/education/search/RG.

Teaching Points May-June Issue 2013

Physiologic, Histologic, and Imaging Features of Retained Products of ConceptionMark A. Sellmyer, MD, PhD • Terry S. Desser, MD • Katherine E. • Maturen, MD • R. Brooke Jeffrey, Jr, MD • Aya Kamaya, MD

RadioGraphics 2013; 33:781–796 • Published online 10.1148/rg.333125177 • Content Codes:

Page 783Patients with RPOC often present with postpartum hemorrhage (PPH), which is divided into primary and secondary types. Primary PPH is defined as blood loss greater than 500 mL in the first 24 hours, whereas secondary PPH is excessive blood loss between 24 hours and 6 weeks postpartum.

Page 784The key to the microscopic diagnosis of RPOC is the presence of chorionic villi, which indicates the persistence of placental tissue.

Page 784The reported incidence of RPOC seems to depend on the gestational age of the pregnancy, with RPOC oc-curring most frequently after second-trimester delivery or termination of pregnancy.

Page 787We have found that the most sensitive finding of RPOC at gray-scale US is a thickened endometrial echo complex (EEC). The exact definition of “thickened” varies in the literature, ranging from 8 to 13 mm.

Page 788Color Doppler US further enhances diagnostic confidence in identifying RPOC. For example, blood clots will appear avascular at color Doppler US, whereas the detection of vascularity in a thickened EEC or en-dometrial mass is likely to represent RPOC.

WOMEN’S IMAGING 781 Physiologic, Histologic, and ......During the second trimester, the placenta...

Documents

Transcript of WOMEN’S IMAGING 781 Physiologic, Histologic, and ......During the second trimester, the placenta...