Radiology

Genitourinary Yasuyuki Koji

MD #{149}Mutsumasa Hitoshi Okamura,

Yamashita,

Miyazaki,

MD

Carcinoma Dynamic

ofthe

magnetic

Takahashi, MD

#{149}

resonance

imaging was performed eight normal uteri and carcinomas. After rapid

(MR)

to evaluate 29 cervical injection

Cervix:

H

Takamine

Dynamic

been

to be

higher signal intensity than mat cervical tissue with this

the

quence.

plane.

In the

normal

uterus,

because

the

MR

tumor

in

imaging,

generates

a

does pulse

however,

norsecan-

initial slight enhancement of the junction between endometrium and

not demonstrate superficial carcinoma (including carcinoma in situ and mi-

myometrium

croinvasive

was

followed

by

hancement

of the uterine Cervical carcinoma

trium. readily distinguished from cal stroma and myometrium

en-

myomewas

the cernin the early dynamic phase (30-60 seconds). Tumor-cervix contrast in the early dynamic phase was significantly

more

marked

technique contrast weighted

with

the dynamic

SE

than with T2-weighted or material-enhanced Tiimaging (P < .01). For the

evaluation of parametrial invasion, the dynamic FLASH study showed better contrast than did T2-weighted images. derwent

Among surgery,

of the degree

18 patients accurate

who unassessment

of stromal

invasion and was possible with dyimages in i4 patients

tumor

size

namic

MR

(78%), (6i%), images

T2-weighted images in ii and postcontrast Ti-weighted in seven (39%).

Index

terms:

parative rapid

Magnetic Magnetic

studies

#{149}

imaging

854.32

Uterine

Uterine

neoplasms,

#{149}

diagnosis,

MR.

854.1214,

854.32

1992; 182:643-648

Radiology

I

From

MT.,

the Departments

T.S.)

and

of Radiology

Gynecology

(KM., HO.), Kumamoto Medicine, 1-1-1 Honjo,

From

the

ceived

16; final November (

1991

July

RSNA

1, 1991;

revision

scientific

received

School of 860, Japan.

assembly. requested

November reprint

(Y.Y.,

Obstetrics

University Kumamoto

revision

6. Address

RSNA, 1992

and

requests

ReAugust

5; accepted to Y.Y.

because

of

poor contrast between the cervical epithelium and tumor (4,5). Dynamic MR imaging with contrast enhancement obtained with gadopentetate dimegtumine is reported to be sensitive in the evaluation of hemodynamics (6) and in demonstrating small tumors (7). To enhance the detectabitity and accuracy of staging in cervical carcinoma, we prospectively performed dynamic MR imaging with

gadopentetate For this

dimeglumine. we evaluated

report,

hemodynamics vical carcinoma,

MR imaging findings,

ability imaging images

of the uterus correlated

with and

and

the and cerdynamic

histopathotogic

compared

accuracy

the

detect-

of dynamic

after

the

contrast

material.

administration

of

AND

METHODS

carcinoma

patients,

were

evaluated

within

nod of 7 days before operation, and roscopic and microscopic specimens obtained. The remaining 11 patients

therapy

(three

patients

classification

of the

International

Fed-

eration of Gynecology and Obstetrics and pathological classification of the tumor (TNM) (8). The depth of stromal invasion at the

midsagittal

as follows: sion (depth of the

level

was

no invasion, of invasion

stroma),

also

classified

superficial less than

or deep

invasion

invaone-half (greater

of the stroma). MR imaging was performed with a 1.5-T superconductive unit (Magnetom; Sicmens, Erlangen, Germany). Ti-weighted spin-echo (SE) 600/15 (repetition time msec/echo time msec) images were obtamed before and after contrast enhancement in the axial or sagittal plane with a than

one-half

192

x 256

matrix

two

and

excitations.

T2-

weighted 2,000/70 SE images were obtamed in sagittal and axial sections with a 256 x 256 matrix and one excitation. In 25 patients, dynamic MR imaging employed the SE technique in the sagittal plane with 200/15, a 128 x 256 rectangular matrix, and one excitation. In patients with large tumors suspected of parametrial or pelvic extension (n = 8; four patients underwent imaging with both techniques separately), dynamic MR imaging the

technique ifip angle

fast

low-angle

in the axial of 40#{176}, a 256

shot

plane x 256

(FLASH)

with 50/10, matrix, and

a

one excitation. Actual sampling time per image 25.6 seconds for the SE technique three planar

aged 33-88 (average, 59.8) years, with cervical carcinoma of the uterus, and eight healthy subjects (two in the proliferative phase, two in the secretory phase, one during menstruation, and three in menopause), aged 34-70 (average, 54.8) years, were prospectively evaluated with dynamic MR imaging. In all patients, cervical carcinoma was diagnosed with colposcopy and biopsy before MR imaging. Eighteen patients with cervical

ing

seconds

MATERIALS Twenty-nine

radiation

with stage IIB, five with stage IIIB, and three with stage IVA). Clinical staging was based on the stag-

employed

MR

with routine T2-weighted and with Ti-weighted images

obtained

resonance (MR), comresonance (MR), neoplasms,

#{149}

carcinoma),

Imaging’

derwent

useful

the evaluation of cervical carcinoma of the uterus. A T2-weighted pulse sequence has been reported to be optimal for imaging cervical carcinomas (1-3)

MD

MR

MR imaging

shown

gadopentetate dimeglumine, dynamic images were obtained every 30 seconds with the spin-echo (SE) technique in the sagittal plane or the fast low-angle shot (FLASH) technique in axial

Sawada,

#{149}

IGH-FIELD-STRENGTH

has of

MD

for

the

chosen for a sufficient

T2-weighted section

images.

with

all pulse

Two

were

thickness

The

to ratio.

obtained

with

to obtain

field

to

with a multithickness of 7

dynamic studies signal-to-noise

images

same

parable cm

technique.

sections were obtained technique. A section

mm was maintain the

FLASH

was and 10.2

of view

corn-

was

23

sequences.

Before the administration of gadopentetate dimeglumine (Magnevist; Nihon Shering, Osaka, Japan), axial or sagittal Ti-weighted and axial and sagittal T2weighted

images

were

obtained.

After

a

a pe-

macwere un-

Abbreviations: FLASH

=

C/N

fast low-angle

=

contrast-to-noise shot, SE = spin

ratio, echo.

643

rapid

hand

injection

opentetate

(2 rnL/sec)

dimeglurnine

dynamic

MR

images

were

each pulse sequence, for 4 minutes. After

routine obtained

of gad-

(0.1 rnmol/kg), obtained

every dynamic

with

30 seconds imaging,

TI-weighted in the same

weighted

utes after dimeglummne).

Image

SE images were plane as the T2and dynamic images (5-6 mminjection of gadopentetate

Analysis carcinoma

Cervical

that

surgery was evaluated T2-weighted images, images (obtained 0-I tion of gadopentetate phase dynamic images

utes

after

injection

dimeglumine),

undergone

dimeglummne), (obtained

late-

mm-

2-3

of gadopentetate

and

weighted

had

with nonenhanced early-phase dynamic minute after injec-

postcontrast

images.

The

TI-

MR imaging

find-

ings were analyzed by two of the radiologists (Y.Y., T.S.) in conference, and the findings were determined by consensus. MR imaging with each sequence was assessed

independently,

without

edge of pathologic (a) tumor visualization mor

delineation,

results,

These

macroscopic

(b)

(c) degree

and

witi

regard

(contrast),

invasion and tumor the imaging criteria

ble).

knowl-

with

to tu-

of stromal

volume (the details of are shown in the Ta-

findings

were and

then

microscopic

correlated find-

ings.

The optimal

assessment in sagittal

of stromal sections,

invasion and that

was of

parametrial invasion was optimal in axial sections. The depth of stromal invasion was interpreted as (a) no stromal invasion (the

cervical

stroma

thickness

of the

was

intact,

stroma

was

and

invasion

(abnormal

intensity

observed

in the

confined (c) deep sity

to the inner half invasion (abnormal

was

half

observed

signal outer

metrial

Figure

but

extending

to the

outer

was the

beyond stroma).

Para-

stroma,

there

was

diffuse

or localized

abnormal signal intensity in the parametrium. Involvement of the pelvic side wall or invasion of the bladder or rectum was defined

by

loss

of normal

signal

of these structures. A small (1-4-mm2)

region

placed

myometrium,

in the

tumor,

endometrium,

and

On sagittal SE dynamic trast material enhancement obtained for each region lows: (S - S,) x I00/S,,

the

tumor

and

644

#{149} Radiology

completely ratio

the

cervical

slight

enhancement

myometrium. with that

.s

postcontrast;

The

initially

junction

lower

between

(arrowheads),

The enhancement of myometrium.

of the

which

cervical

left,

J

the

endometrium

seconds

is followed

stroma

by

and

en-

endome-

fat

tissue.

studies, the conratio was of interest as fol-

where

S, mdi-

destroyed), (C/N)

stroma

ratio

U

normal

8).

Sequential dynamic

uteri

in the

Ranges

mean.

MR

sagittal

indicate

within

seconds

after

The en-

administration

dimeglumine. of

the

uterine

The

a) E a)

C a) U) ‘5

of

en-

myometrium

C

U C ‘5

myo-

of the junction between the and myometrium is marked

gadopentetate significant enhancement endometrium that of the

of

cervix,

and

#{149} = endometrium.

hancement

error =

of endometrium

0

for

plane

standard

metrium, hancement endometrium 60

enhanceimages

#{149} = myometrium,

junction

=

contrast

of SE

C 0

0

is

in the late dynamic phase. of the cervical stroma was gradual compared myometrium.

The and with

was

cervix,

cates signal intensity at each dynamic phase and S0, signal intensity on the image obtained before administration of contrast material. In patients with cervical carcinoma (except for those patients whose cervix had been the contrast-to-noise

shows

of the uterine slight compared

right, 30 postcontrast).

intensity

of interest

parametrial

2.

ment

the (z the

invasion was diagnosed when, in to the loss of low-intensity cervi-

addition cal

signal stroma,

of the stroma), signal inten-

intensity was observed stripe of the cervical

myometrium

hancement trium were

upper seconds

the

of the stroma, but the outer stripe or (it) serosal invasion (abnormal

intact),

and

left, precontrast; lower right, 210

uniform),

(b) superficial was

phase (upper postcontrast;

as follows: (S - S)/noise SD, S indicates signal intensity of the S, signal intensity of the cervical stroma; and noise SD, the standard deviation of the intensity of the background noise in the phase-encode direction (9). C/N between the tumor and parametrial fat

tissue

dynamic

ages

was obtained images and as follows: (S

ob-

-

for FLASH conventional S1)/noise SD,

and SE SE imwhere

Si indicates

signal intensity of the paramefat. The averages for the two pulse

trial sequences

between

was

tamed where tumor;

dent

t test.

were

compared

with

the

Stu-

RESULTS In normal the

endometrium

uteri,

the and

junction

between

myometrium

was

slightly enhanced initially; enhancement of the uterine myometrium followed. Initial enhancement of the junclion was not observed in two of eight healthy volunteers (one in the secretory phase and the other in menstruation). The endometrium and cervical epithelium showed little enhancement during the dynamic phase, regardless of the phase of the menstrual cycle or the age March

1992

of Pathologic

Summary

and MR Findings

in Operated-on

Carcinoma

of the Cervix Tumor

Case

Age

FIGO

No.

(y)

Stage

Visualization*/

Delineationt/Interpretation*

.

Histologic

Pathologic

Findings

Stage

Stromal Invasion

T2WSE

Dynamic

Post-CE

pTis

Insitu

-/N/N

pTla

Superficial

-/N/N

-IN/N +/C/+

-IN/N -/N/N

Adenocarcinoma SCCK SCCKS

pTib pT2b pTlb

Superficial Serosa Superficial

pTlb pI2a

Superficial

++/C/+ ++/C/+ ++/R/+ ++/C/+

+/U/+ ++/C/+ +/U/+/U/-

lID

SCC NK L SCC NK L

+/U/+ ++/C/+ +/U/+ -/N/N

Deep

+ +/C/+

+ +/R/+

+

CIS

N/A

pus

In situ

-IN/N

-/N/N

1113

5CC NK L SCCKS

pT2b

Deep

++/C/+

++/R/+

+ +/C/+

SCCNKL

pT2a pT2a

Deep Deep

+/U/++/C/+

#{247}+/R/+ ++/R/+

++/U/+ ++/U/+

N/A

pus

Superficial

-/N/N

-/N/N

SCCNKL

pTla

Superficial

-/N/N

+/U/-

lB

5CC NK L

pTib

IIB

SCCNKL

pTib

Deep Deep

++/C/+ ++/C/+

+#{247}/R/+ ++/C/+

66

IIB

pT2b pub

++/C/+

++/R/+

+/U/-

lB lB

SCC NK L SCC K S

Deep

34 73

Superficial

+/C/+

+ +/R/+

-/N/N

SCC

pTlb

Superficial

+/C/+

++/C/+

+/U/-

I

40

lB

2

51

hA

3 4 5 6

58 62 36 78

IIA IIB

7

70

8 9 10

33 55 36

11 12

49 50

hA 1113 CIS

13

36

IA

14

59

15

68

16 17 18

SCCNKL Verrucose carcinoma

lB lB

NK L

Note-Dynamic = SE dynamic image, FIGO = International Federation ofGynecology NK = nonkeratinizing, post-CE = postcontrast Ti-weighted SE image, S = small, SCC = *Tumor visualization on MR images: + + = tumor was excellently demonstrated with observed with fairly good contrast; = tumor was not visualized, or the contrast between Tumor delineation on MR images: R = tumor was clearly delineated with a rim, c = delineated, N = tumor was not visualized. 1Thterpretation of the stromal invasion and the tumor size on MR images: + = stromal defined, but degree of stromal invasion or tumor size was not correctly interpreted, N =

and Obstetrics,

K

keratinizing,

L

=

+/C/+ -/N/N

-/N/N -IN/N ++/C/+ +/U/-

large cell, N/A

=

not available,

squamous cell carcinoma, T2WSE = 12-weighted SE image. remarkable constrast between the tumor and stroma; + = tumor was the tumor and the stroma was poor. tumor was dearly delineated without a rim, U = tumor was unclearly invasion and tumor size was correctly tumor was not visible.

interpreted,

-

=

tumor

was

Figure 3. Case 15. MR images and contrast enhancement ratios obtamed in a 68-year-old woman with a cervical carcinoma. (a) SE dynamic MR images (200/15) (upper left, precontrast; upper right, 30 seconds postcontrast; lower left, 90 seconds postcontrast; lower right, 210 seconds postcontrast). (b) Sequential contrast enhancement ratio of SE dynamic MR images. 0 = tumor, #{149} = myometrium, 0 = cervix, U = junction of endometrium and myometrium, #{149} = endometrium. (c) T2-weighted SE image (2,000/70). (d) Postcontrast Ti-weighted image (600/15). (a, b) In the early dynamic phase, the tumor is seen as a well-enhanced mass in the cervix. The enhancement of the junction of uterine endometrium and myometrium is seen (arrowheads). In the late dynamic phase, the tumor became isointense with myometrium. Gadopentetate dimeglumine is observed in the bladder at 210 seconds (arrow). (c) T2-weighted image obtained at the same sagittal plane demonstrates a hypenntense mass corresponding to the enhanced area on dynamic MR images. (d) Postcontrast Ti-weighted image also shows a hypenntense that on dynamic and

mass, but T2-weighted

the

contrast images.

is not

so remarkable

as

a. 0 Ia) C 0

E 0 0 C C 0 I0 Ia)

C 0

0 30

60

90

120

Time after administration

150

180

210

of Gd-DTPA(seconds)

b. Volume

182

#{149} Number

3

Radiology

#{149} 645

of

the

vical that

patient. of

(Figs In

the

of

slight

was

the

compared

myometrium

in

surgery are shown in the Table. On T2weighted SE images, dynamic images, and postcontrast Ti-weighted images, tumors demonstrated high signal intensity compared with the low signal intensity of the surrounding cervical stroma. Abnormal signal intensity was not observed in two patients with carcinoma in situ and one with superficial stromal invasion on T2-weighted SE and dynamic images. In three patients, only dynamic MR images depicted the tumor, and in five patients, tumors were seen more clearly on dynamic MR images than on T2-weighted images. In

cer-

with

most

cases

1, 2). patients

tumors

than

Enhancement

stroma

with

showed

carcinoma, earlier

myometrium and cervical earlier than did the junction between the endometrium and myometrium (Fig 3). The C/N between the tumor and the cervical stroma could be evaluated on SE dynamic studies obtamed in 17 patients. Contrast was stroma

did

cervical enhancement

the

slightly

and

marked

in

the

early

dynamic

phase

seconds), but decreased with time (Fig 4). The MR imaging and histopathologic (30-60

findings

in patients

who

these

eight

tween was

underwent

patients,

the not

the

tumor

and

contrast

cervical

remarkable

on

S a

Time after administration

Figure

be-

T2-weighted

seconds)

a. Figure trast;

upper

2. right,

Cervical 30

carcinoma

seconds

of the

postcontrast;

verrucose lower

left,

Sequential the cervical

and

b. 5. Case

of Gd-DTPA(seconds)

C/Ns between the tumor and stroma on SE dynamic MR images. There are various degrees and patterns of enhancement in dynamic MR imaging. In most patients, the contrast between the tumor and the cervical stroma was substantial in the early dynamic phase (30-60

stroma

4.

decreased

with

time.

C.

type

in a 51-year-old

90 seconds

postcontrast;

woman. lower

(a) SE dynamic right,

210

MR

images

seconds

(200/15)

postcontrast).

(upper (b)

Surgical

left,

precon-

specimen

of

the uterus. (c) Histologic specimen of the uterus (hematoxylin-eosin stain; original magnification, x 1). (a) Dynamic MR images obtained in the early dynamic phase show early enhancement of the tumor in the anterior and posterior cervical lips (arrowheads). Enhancement of the endocervix is also seen. (b) Tumor is located from the anterior lip through the lateral wall to the posterior lip on the surgical specimen (arrowheads). (c) Microscopic observation shows a papillary tumor extending into two-thirds of the cervical stroma, correlating very well with dynamic MR images. The depth of invasion was less than 5 mm (arrowheads). The tumor was not visualized on T2-weighted images.

a. Figure onds

b. 6. Case

5.

postcontrast;

Cervical lower

carcinoma left,

90

in

seconds

c.

a 36-year-old

woman.

postcontrast;

lower

(a) right,

SE 210

dynamic seconds

MR

images

postcontrast).

(200/15) (b)

(upper T2-weighted

logic specimen (hematoxylmn-eosin stain; original magnification, x40). (a) In the dynamic phase, enhancement the peripheral portion of the tumor (arrowheads). The junction of uterine endometrium and myometrium row). (b) T2-weighted image in the same sagittal plane shows slight hypermntensity of the tumor compared margin of the tumor is unclear. (c) Microscopic observation shows well-demarcated squamous cell carcinoma. served

646

at

the

#{149} Radiology

margin

of

the

tumor

left,

precontrast; SE

image

upper (2,000/70).

right, (c)

30

5cc-

Histo-

of the tumor is seen especially at is slightly seen at 90 seconds (arwith the surrounding tissue, but the Infiltration of lymphocytes is ob-

(arrowheads).

March

1992

images.

This

contrast was due to of fibrous tissue in patients and a relaextension of the tumor

lack

in

six

with well-demarcated tumors and were usually consistent with compression of normal tissue or an inflammatory reaction (Fig 6). In 18 patients who underwent surgery, accurate assessment of the degree of stromal invasion and tumor size was possible in ii patients (61%) with T2-weighted images, 14 patients (78%) with dynamic MR images, and seven patients (39%) with postcontrast Ti-weighted images. When C/Ns were compared by testing various pulse sequences, the earlyphase SE dynamic images elicited a better C/N (P < .01) than did T2-weighted images and late-phase dynamic images. Ti-weighted images obtained after the administration of contrast material showed poor contrast (Fig 7). FLASH axial images showed parametrial extension more clearly than did T2-weighted images in all patients (Fig 8). Dynamic FLASH images showed hyperintensity in the tumor and relative hypointensity in the parametrial fat (Fig 9). C/N between the tumor and parametrial fat was better on Ti-weighted images with or without contrast enhancement and on postcontrast FLASH images than on T2-weighted or precontrast FLASH images (Fig 10). Contrast on Ti-weighted images obtained with or without contrast enhancement was much superior to that on other images (negative contrast), but there was little contrast between the tumor and normal tissue. FLASH images showed good contrast between the tumor and parametriat fat tissue or the surrounding organs (Figs 8, 9).

of

the predominance the tumor in two tively superficial patients.

With dynamic MR imaging, carcinoma in situ was not visualized, but even a preclinical invasive tumor less than 5 mm in depth demonstrated high signal intensity, which correlated well with histologic findings (Fig 5). In large tumors, any extension to the vagina and uterine body was more clearly demonstrated with dynamic MR imaging. Sharp demarcation of the margin of the tumor was demonstrated in 14 patients (78%) on the dynamic images, nine (50%) on T2-weighted images, and four (22%) on postcontrast Ti-weighted images.

Hyperintense

rims

were

ob-

served around the tumor on dynamic MR images in eight patients (44%). Histologically, the rims were correlated

0

0

0 0

7.

Figure vical

Contrast

stroma

between

with

each

tumor

pulse

and

sequence

cer(the

averages in 18 measurable cases). The contrast on FLASH dynamic images was not available because the cervix was almost destroyed in patients who were imaged with those ages

pulse in

sequences.

the

early

phase

The

SE

dynamic

had

the

best

ages

(TIWSE)

Figure

8. radiation

right,

30

therapy.

seconds

(c)

ment. Invasion clearly than on (b). These areas internal

Volume

poor

MR images

went

(2,000/70).

showed

182

obtained in a 51-year-old (a) Dynamic MR images lower TI-weighted

to the left internal the T2-weighted probably represent

obturator

muscle

Number

#{149}

High-field-strength, imaging of invasive

T2-weighted cervical carcinoma

supply

of the uterus

is de-

at dynamic MR imaging. The contrast resolution of MR im-

aging demonstrates the tumor definitely better than do other imaging techniques. Dynamic MR imaging reveals the hemodynamics of the uterus in vivo. Patterns of enhancement at dynamic MR imaging reflect the vascularity of each part of the uterus. In the normal uterus, enhancement occurred initially at the junction between the endometrium and myometrium, followed by enhancement of the uterine myometnum. Initial enhancement of the junction, however,

MR

contrast.

postcontrast; Postcontrast

markable excellent

DISCUSSION

(T2WSE) images images (P < .01). CE) Ti-weighted SE im-

(Post

The blood

rived from the uterine and ovarian arteries anastomosing with each other to form a uteroovarian vascular arch with many arterial branches supplying the uterine wall (10). As shown with histologic and microangiographic observation, all parts of the myometrium of the uterus are vascularized, especially at the junction of the stratum basale of the endometnum and subjacent myometrium, while vessels of the cervical stroma and endometrium are relatively sparse (1 1 ,i2,13). Cervical carcinoma had been considered to be hypovascutar at angiography (14); the degree of contrast enhancement of the cervical carcinoma, however, was shown to be re-

imcontrast,

followed by T2-weighted and late-phase dynamic Postcontrast

has been shown to have an accuracy rate as high as 9i% (3) and 95% (5). Superficial lesions were underestimated, however, because of poor contrast between the high signal intensity of the tumor and that of the cervical epithehum and because of the presence of surface blood clotting (4,5). Although our study is preliminary, the early phase of dynamic imaging demonstrated significantly better contrast than did T2weighted images.

with

3

left,

woman obtained

90 seconds image

postcontrast;

(600/15).

obturator muscle (right image (b). Unenhanced cystic necrosis in the good

contrast,

but

with cervical carcinoma extending to adjacent with the FLASH technique (50/10; flip angle,

there

(a, b)

lower Images

right,

210

of dynamic

seconds study

postcontrast). show

invasive

organs (stage IVA). The patient under40#{176}) (upper left, precontrast; upper (b) tumor

T2-weighted with

marked

spin-echo

image

contrast

enhance-

arrowhead), bladder (left arrowhead), and rectum (arrow) is visualized more areas in the tumor in a correspond to hypenntense areas on the T2-weighted tumor. (c) Postcontrast Ti-weighted image also shows invasion to the rectum is no

contrast

between

the

tumor

and

the

image and left

bladder.

Radiology

#{149} 647

was

not

always

observed.

The

zone

of

initial enhancement the junction of the

probably reflects stratum basale of the

endometrium

subjacent

and

myome-

trium. There may be some relationship between the hormonal change and pattern of enhancement. Further study is required to clarify this point. Enhancement of the cervical stroma appeared slight and gradual compared with that of the myometrium. On the other hand, cervical carcinoma demonstrated enhancement earlier and more intensely

than

uterus. therefore, visual

The

any

tissue

in the

dynamic MR images, the best contrast at

analysis,

as

before

any

of C/N

other

early provided

well

as quantification

enhancement

cervical

stroma

curred.

On

postcontrast

and

images,

the

contrast

of the

myometrium

oc-

TI-weighted between

the

tumor

and stroma was decreased because of the delay that followed administration of gadopentetate dimeglumine. Our histologic correlation showed that the early enhanced lesions were consistent

with

the

presence

prove

or tissue.

to be

tion

of the

invasive

sign

degree

well demarrim was the tumors, with reac-

compression This rim might

a useful

in

the

of

evalua-

of infiltration

of the

tumor.

both the tumor fat show hyperintensity dynamic SE and T2-weighted

and

Because

trial

niques,

it is difficult

trial

invasion.

ters

of 50/10

showed

to

tumor

tensity

in

addition, nor

to the

two

averages

and

shorter

paramewith the SE tech-

evaluate

low

signal

displayed the

early

this

FLASH

signal

in-

phase.

In

dynamic technique

is supe-

SE parameters 600/15 in signal-to-noise time.

On

technique

for

tumors

will

be

gas

SE

usefulness

required

of the

parametrial

We ity

weighted

imaging,

648

to

FLASH

have

(15).

and

but

#{149} Radiology

MR

First,

evaluating

that

of the

this

enhanced

shows

parametrial

the

3.

4.

pulse

tumor.

a relatively

sequence,

Invasion

to the

hyperintense

mass

with

fat.

6.

can

be

per-

routine

SE

techniques

With improvement and acquisition will

become

carcinomas

are

em-

of the

useful

7.

in

superiorT2-

draw-

0

0 0#{149}

T2WSE

Figure

10.

averages trast

in

cervix.

U

FLASH

p..

Contrast

parametrial

of spatial time, this more

025

fat

in five

Ti-weighted

FLASH

CE

between

with

each

TIWSE

tumor

pulse

CE

PS1

dy”*#{232}

and

sequence

(the

measurable

cases). PreconSE (T1WSE) images, images, and postcontrast

FLASH dynamic weighted (Post CE T1WSE) contrast, while precontrast and T2 weighted (T2WSE) poor contrast (P < .01).

10.

images

had

Ti good

(pre CE) FLASH images

Togashi K, Nishimura K, Sagoh T, et al. Carcinoma of the cervix: staging with MR imaging. Radiology 1989; 171:245-251. Choyke PL, Frank JA, Girton ME, et al. Dynamic Gd-DTPA-enhanced MR imaging of the kidney: experimental results. Radiology 1989; Sakamoto Y, Takahashi M, Korogi Y, Bussaka H, Ushio Y. Normal and abnormal pituitary glands: gadopentetate dimeglumine-enhanced MR imaging. Radiology 1991; 178:441445. American Joint Committee on Cancer. Gynecologic sites. In: Beahrs OH, Henson DE, Mutter RVP, Myes MH, eds. Manual for staging of cancer. 3rd ed. Philadelphia: Lippin-

Pritchard JA, MacDonald the female reproductive stetncs.

11.

16th

tiiry-Crofts, Farrer-Brown blood ture.

showed

supply

J Obstet

ed.

New

PC. tract. York:

1980; 11-39. G, Beilby JOW, of the

uterus.

Gynaecol

Anatomy of In: Williams oh-

Appleton-Cen-

Tarbit

MH.

I. Arterial

The

vascula-

Br Commonweal

1970;

77:673-681. 12.

Farrer-Brown

C, Beilby

blood

of the

supply

J Obstet

Gynaecol

JOW,

uterus.

Tarbit

MH.

II. Venous

Br Commonweal

The

pattern. 77:

1970;

682-689. 13.

14.

170:713-720.

Ti-weighted

several

study

150:135-138.

the

over

a dynamic

References

are

evaluate

imaging

after

ployed. resolution

8.

are

as

image

formed only in limited imaging planes. In our system, the number of sections available per study is limited to three at present. Second, the spatial resolution is inferior to that of routine SE sequences. Therefore, we believe this technique should be limited to selected patients with a carcinoma with superficial extension, invasion to the other organs, or poor visualization on T2-weighted im-

5.

parametrial images

postcontrast

there

backs.

intensity

With

with

1986; 159:725-730.

We

technique

demonstrated

same

T2-weighted

Togashi K, Nishimura K, Itoh K, et al. Uterinc cervical cancer: assessment with high-field MR imaging. Radiology 1986; 160:431-435. Hricak H, Lacey CG, Sandles LG, Chang YCF, Winkler ML, Stern JL. Invasive cervical carcinoma: comparison of MR imaging and surgical findings. Radiology 1988; 166:623-631. Rubens D, Thornbury JR. Angel C, et al. Stage lB cervical carcinoma: comparison of clinical, MR, and pathologic staging. AJR 1988;

invasion.

of dynamic

the

(b)

enhancement.

compared

2.

reported to be useful in the evaluation of parametrial invasion (3), but they cannot aid in differentiating the tumor from normal parametrial tissue. Further study

almost

SE dynamic techand the FLASH with

Ti-weighted

is speculated.

dark

the

induced

invasion.

parametrium

contrast

relatively

Worthington JL, Balfe DM, Lee JKT, et al. Uterine neoplasms: MR imaging. Radiology

facts

intestinal

inhomogeneous

1.

hand, the SE dynamic images superior in tissue contrast benormal uterine tissue and the and were less susceptible to artiby

with

fat appears

with ratio

other were tween tumor

therefore selected the nique for small tumors

tumor

parametrial

technique

intensity,

high

sampling

invasive

ages

parame-

With the FLASH parameand a flip angle of 40#{176}, fat

relatively

while

b.

9. Case 7. MR images obtained in a 70-year-old woman with a cervical carcinoma with parametrial invasion (stage IIB). She underwent radiation therapy. (a) Dynamic MR images obtained with the FLASH technique (50/10; flip angle, 40#{176}) (upper left, precontrast; upper right, 30 seconds postcontrast; lower left, 90 seconds postcontrast; lower right, 210 seconds postcontrast). (b) T2-weighted SE image (2,000/70). (a) Images of the dynamic study show an

Figure

of tumors

in all cases. In histologically cated lesions, a hyperintense frequently observed around which was usually consistent tive inflammation surrounding

a.

is.

Hendrickson

M, Kempson

R.

The

normal

myometrium. In: Surgical pathology of the uterine corpus. Philadelphia: Saunders, 1980; 439-451. Yamashita Y, Takahashi M, Bussaka H, et al. Balloon-occluded arterial infusion therapy in the treatment of primary and recurrent gynecologic malignancies. Cardiovasc Intervent Radiol 1989; 12:188-195. Czervionke LF, Daniels DL, Wehrli FW, et al. Magnetic susceptibility artifacts in gradientrecalled echo MR imaging. AJNR 1988; 9:11491145.

cott, 1988; 151-153. 9.

Stark DD, Wittenberg J, Edelman RR, et at. Detection of hepatic metastases: analysis of pulse sequence performance in MR imaging. Radiology 1986; 159:365-370.

March

1992

Carcinoma of the cervix: dynamic MR imaging.

Dynamic magnetic resonance (MR) imaging was performed to evaluate eight normal uteri and 29 cervical carcinomas. After rapid injection of gadopentetat...
1MB Sizes 0 Downloads 0 Views