A. David MacVicar, C. Ross Pinkerton,

MA, MRCP, MD, FRCPI

FRCR

#{149} Julie F. C. Olliff, MRCP, FRCR2 E. S. Husband, FRCP, FRCR

#{149} Janet

#{149} Jean

Pringle,

MB,

ChB

Ewing Sarcoma: MR Imaging of Chemotherapy-induced Changes with Histologic Correlation’

I

In a study

group of 18 consecutive with Ewing sarcoma proved by means of biopsy, the signal intensity characteristics of tumor on magnetic resonance (MR) images were assessed before and after chemotherapy. Sixteen patients underwent MR imaging at 1.5 T before chemotherapy; all 18 patients underwent MR imaging at 1.5 T within 10 days after chemotherapy. Standard spin-echo sequences were used with Ti and T2

years, magnetic resonance (MR) imaging has become the modabity of choice in the study of primary bone tumors (1,2). It is sensitive in definition of the extent of tumor in bone marrow and soft tissue in a variety of malignant diseases, including Ewing sarcoma and osteosarcorna (2-5). The information that MR irnaging provides is important to the surgeon who plans surgery for limb salvage (4), particularly because preoperative chemotherapy has increased the feasibility of such procedures. The tendency of Ewing sarcoma to recur locally has persuaded many oncobogists and surgeons that resection of the primary tumor provides the best chance of cure after chemotherapy (6,7), which is used to reduce the bulk of the softtissue tumors and has the potential to

patients

weighting in all patients. mary tumor was visualized patients who underwent

The

pri-

in all 16 MR imaging before chemotherapy. Histologic correlation, obtained in 14 patients, showed that areas of high T2weighted signal intensity on MR images obtained after chemotherapy may represent tumor necrosis, cystic hemorrhagic areas, and fibroblastic repair tissue. In 10 patients (71%), microscopic clusters of viable tumor cells were depicted in areas of both low and high signal intensity after treatment. It is concluded that MR imaging is unreliable for exclusion of active disease, although a pattern of change in signal intensity is qualitafive evidence of chemotherapeutic effect. Index

terms:

Bone neoplasms,

neoplasms, MR, 40.1214 Ewing sarcoma, 40.3281 (MR), tissue characterization,

Radiology

40.32

eliminate bone and

tumor cells entirely from soft tissue. It is important

to

define extent of remaining abnormalities accurately in both bone marrow and soft tissue and attempt to identify areas of residual active tumor. Although MR imaging has become the most sensitive modality with which to demonstrate response to treatment, no clear consensus exists on the importance of observed changes in signab intensity (8-10). The purpose of

Bone

#{149}

our

Chemotherapy Magnetic resonance 40.3281



N recent

study

was

to provide

further

in-

formation on the importance of changes in signal intensity after chemotherapy and thereby establish the role of MR imaging in the management of Ewing sarcoma.

#{149}

1992; 184:859-864

In a study group of 18 patients with Ewing sarcoma who underwent chemotherapy, the appearances of the primary tumor on MR images were reviewed. In 14 patients in whom resection was possible, detailed histologic correlation was obtained, and the four other patients have been fobbowed up with clinical and radiobogic

examination. MATERIALS Patients Our

AND

in This study

METHODS

Study

group

comprises

tive patients with Ewing by means of biopsy who

18 consecu-

sarcoma

proved

were referred to the Royal Marsden Hospital, Sutton, Surrey, England, for chemotherapy between January 1987 and June 1990: 11 male and seven female patients aged 2 years 11 months to 17 years 5 months (mean age at diagnosis, 11 years 9 months). Fourteen patients underwent surgery after chemotherapy; 12 of these patients underwent MR imaging before and after chemotherapy,

whereas

imaging

two

only

patients

after

underwent

MR

chemotherapy.

MR

contraindicated

due

to the

extent of the disease. Chemotherapy was ing

to standard

patients apy that famide,

site

or massive

performed

accord-

treatment

underwent consisted

regimes.

Sixteen

induction chemotherof four courses of ifos-

vincristine

sulfate,

and

doxorubi-

cm hydrochloride (Adriamycin; Farmitalia Carlo Erba, St Albans, Hertfordshire, England). Two patients underwent treatment with cycles of vincristine sulfate, methotrexate, doxorubicin hydrochloride, and cyclophosphamide alternating with bleomycin

and

actinomycin

D. The

courses

were I

From

the

Cancer

ology (A.D.M., Sutton, Surrey Brockley Hill,

Research

5, 1992; revision

Research

Campaign,

2 C

requests Current

RSNA,

Magnetic

Resonance

Research

J.F.C.O., J.E.S.H.), and Paediatric Unit (C.R.P.), SM2 5PT, England; and Department of Morbid Stanmore, Middlesex, England (J.P.). Received

February print

Campaign

received

London,

and

March

23; accepted

the Daniel

Marks

March Fund,

Group,

Department

of Radi-

Royal Marsden Hospital, Downs Rd, Anatomy, Institute of Orthopaedics, December 19, 1991; revision requested

27. Supported

Royal

im-

aging was performed before and after chemotherapy in four patients who did not undergo surgery because surgery was

administered at 21-day intervals, though delays of 7 days or less were

sionably

necessary

to allow

blood

alocca-

counts

to

recover.

in part by the Cancer

Marsden

Hospital.

Hospital,

Birmingham,

Address

re-

to A.D.M. address:

1992

Department

of Radiology,

Queen

Elizabeth

England.

Abbreviations: SE time, TR = repetition

=

spin echo, TE time.

=

echo

859

a.

b.

c.

Figure 1. MR images of Ewing sarcoma of right iliac bone in a girl aged 16 years. (a) Ti-weighted (500/17) MR image obtained before chemotherapy shows large soft-tissue tumor (both anterior and posterior to the iliac wing) with the same signal intensity as that of muscle. Vertical bar at right in a-c indicates 5 cm. (b) T2-weighted (2,100/70) MR image obtained before chemotherapy shows tumor with area of inhomogeneous high signal intensity that originates in the posterior part of the iliac bone and spreads into soft tissue both anterior and posterior to the iliac wing. (c) T2-weighted (2,100/70) MR image, obtained 7 days after completion scopic foci of active tumor were found in an area of low signal intensity medial dense reactive bone and reduction in soft-tissue tumor. The areas of high signal

no active

MR

of treatment, to the iliac bone intensity in the

was obtained 4 cm (small arrow). Note posterior iliac bone

caudal to a and b. Microthe signal void of the (large arrow) contained

tumor.

Imaging

Fifteen patients underwent MR imaging during the week before the start of induction chemotherapy; one patient began to undergo induction chemotherapy 3 weeks before the first MR examination. All 18 patients underwent MR imaging within 10 days

:

after

completion

of induction

chemo-

therapy. The delay before surgery varied between 21 and 40 days, depending on the patient’s recovery from chemotherapy. The MR examinations were performed on a 1.5-T system (Magnetom; Siemens, Erlangen, Germany). MR images of the primary tumor were obtained in at least two planes to optimally depict extent of the tumor. Standard spin-echo (SE) Sequences were used with TI weighting (repetition time [TR] msec/echo time [TE] msec = 500-700/17-20) or T2 weighting (1,800-2,500/25-30, T2-weighted

70-90). sequences

were

Both TI- and performed

in all patients. Section thickness depended on the size of the tumor but did not exceed 10 mm; the corresponding gap did not exceed 20% of section thickness. MR images were reviewed by two radiologists (A.D.M., J.E.S.H.). Measurements of bone marrow abnormality were made in the

extent. tissue

plane

that

Maximum lesions

were

best

demonstrated

The

grades

areas

measured

on

of signal

#{149} Radiology

b. 2.

treatment

MR images in a boy

obtained

aged

left fifth rib. The signal image. found.

The area Minimal

of high soft-tissue

(a) before

13 years.

chemotherapy

(a) T2-weighted

intensity

is high

and

and

(2,100/90)

(b) 10 days MR

inhomogeneous.

signal intensity corresponded tumor remains.

image

after shows

(b) T2-weighted

to necrotic

bone.

completion large

of

tumor

(2,100/90)

No

active

of

MR

tumor

was

of softaxial

MR

intensity

of

tumor were as follows: low, equivalent to or lower than that of muscle; intermediate, higher than that of muscle and lower than that of fat; high, equivalent to or higher than that of fat and bower than that of water; and very high, equivalent to that of water. Areas of high or very high signal intensity were seen on T2-weighted images 860

Figure

its

transverse

images. The signal intensity of tumor on both Ti- and T2-weighted images was assigned (by consensus) a nominal grade by means of comparison with the signal intensities of other tissues on hard-copied images.

a.

only.

The

estimations

of maximum

trans-

verse area and grading of signal intensity were performed before and after chemotherapy. Data from region-of-interest analyses were not obtained because the signal from

at the tumor

cientby

inhomogeneous

to render

cedure

at least

as subjective

console was suffias visual

the

prograd-

ing.

Pathologic

Middlesex,

England;

were

performed

by local

geons.

Pathologic

formed

at the

Institute

Hill,

or the

Brockley

and flat bones National OrHill, Stan-

rib

resections

thoracic

examination

surwas

per-

of Orthopaedics,

Royal

Marsden

Hos-

pital. All specimens were examined macroscopically and with fine-detail radiography before processing and were histobogically examined Sections of bone plane considered

Examination

Surgical resection of long was performed at the Royal thopaedic Hospital Brockley

more,

est extent bone were ping.

after decalcification. were obtained likely to cross

of tumor-bearing numbered

for

in the the great-

bone. histologic

September

Blocks map-

of

1992

Sites

of viable

tumor

and

other

fled.

histo-

logic abnormalities were then correlated with findings on MR images obtained after chemotherapy. Wherever MR imaging planes did not coincide with histologic sections, an attempt was made to identify the various abnormalities by means of norma! anatomic landmarks and MR imaging coordinates.

In two

patients

tumor was too The Ti-weighted

unchanged tients

in whom

intensity

on MR Images Chemotherapy

Findings

Obtained

Soft-tissue tumor-The primary tumor was visualized in all 16 patients who underwent MR imaging before chemotherapy. Maximum measurable diameters of the soft-tissue tumors were 2-18 cm. The soft-tissue tumor had low signal intensity on Tiweighted images in all patients except one in whom intermediate signal intensity was observed. On T2weighted images, soft-tissue tumors had intermediate or high signal intensity in 14 patients. Tumors in six of the 16 patients were assigned a mixed grade, which reflects the observation that the signal was frequently inhomogeneous (Figs ia, lb. 2). Two patients had mixed high and very high signal intensity in the untreated softtissue tumor, a finding believed to have been caused by prominent tumor edema.

Changes

in bone.-Signal

intensity

within the bone marrow was assessed in 14 patients. In two patients with rib lesions, the bone marrow abnormality was too small to be accurately assessed. In 12 of the 14 patients, the signal intensity of the tumor was low on Ti-weighted images; in the two

other

patients

(both

of whom

had rib On T2-

lesions), it was intermediate. weighted images, the signal intensity was intermediate or high in all tumors (intermediate in one tumor, high in seven tumors, and mixed in six tumors). The pretreatment MR images of long bones also had areas of very low signal intensity, caused by marked shortening of Ti and T2 in areas of cortical thickening and periosteal reaction associated with the tumor.

Findings on MR after Chemotherapy

Images

could be evaluated, sity (similar to that

before

Soft-tissue tumor-In whom soft-tissue tumor ated

after

mum

all patients

was

chemotherapy,

transverse

area

the

evalumaxi-

of soft-tissue

tumors was reduced by over 50% (Figs 1, 2). In three patients no residual soft-tissue tumor could be identi-

Volume

184

#{149} Number

3

intenimage

chemotherapy)

higher

was

in two

and

periosteal

sur-

patients,

Review of MR images obtained after chemotherapy enabled areas of abnormal signal intensity to be correlated with histologic appearances. Areas in which active tumor was found correlated with areas of bow or intermediate signal intensity on T2weighted images in eight patients.

lower in five patients, and unchanged in four patients. Changes in bone.-The extent of bone marrow abnormalities did not change markedly. Ti-weighted signal intensity remained low in all tumors except two, where it changed from

Areas

low

nal intensity in the area because of limitations in image clarity. The frequentby seen areas of high and very high signal intensity on T2-weighted

to mixed

low

and

hand,

the intensity

T2-weighted

signal

from

row nine

was assigned of 13 patients

comparison

common

pattern

signal

grade direct

possible.

The

of change

increase in signal these nine patients,

intensity

most

was

intensity. areas

an

In eight of of very high

developed

where

had previously 3, 4). The signal

been present intensity became

more

heterogeneous

after

a change

which was

the

also

nominally

Two

chemother-

in tumors

of signal who

of the iliac of very high

had bone signal

stable during and radiation

with

of 24 and

follow-up

Features Correlation

that

further therapy,

30 months.

with

patients,

in eight small

MR

patients.

lesions

in diame-

ter) were seen in the resected soft tissue, which revealed tumor at histologic examination. Clusters of active tumor

cells

were

quently mens. blastic

in bone

seen Delicate, reparative

marrow,

related weighted isolated cells

scattered

through-

often between layers of cortical tumor

were

no

in all 14 resected highly tissue

cortical

sped-

vascular fibrowas present

bone,

and

soft tissue and was associated with varying degrees of edema. New bone formation caused cortical thickening

cor-

patients

with

in rib lesions), conclusion

ac-

we

about

sig-

with a high intensity of T2signal (Fig 3a, 3b), although microscopic clusters of active

were

identified

with

in such

high

frequently

areas

in

patients, localfound that cor-

signal

images.

intensity

on

A signal

seen

void

in association

with such hemorrhagic

cystic areas. The finding of debris within such cystic areas suggests that hemosiderin may have caused the signal void, although dense bone bordering cystic areas was also seen. It is also possible that chemical shift artifact contributed to the void

(Fig

3c, 3d).

DISCUSSION Examination with MR imaging is now firmly established as a key procedure in management of bone tumors because it reveals the extent of invasion of bone marrow and soft-tissue masses, although plain radiographs

remain the single most reliable predictor of the histologic nature of a bone tumor at initial assessment (ii) and are considerably useful in follow-up (12). Thus, MR images and plain radiographs should be considered complementary studies in initial assessment.

fre-

other

two patients. In some ized cystic areas were

signal

In two

(1 cm

intensity

images were proved to represent regions rich in the delicate, highly vascular fibrobbastic repair tissue. Areas with extensive bone necrosis also cor-

was

Active tumor was found in 10 of the 14 patients with detailed histologic findings (71%). In six patients it was found in bone alone, three patients had soft-tissue tumor only, and in one patient active tumor was present in bone and soft tissue. Tumor cells were found in small clumps, visible only at microscopy,

make

related

unresectabbe developed intensity

signal

(both

T2-weighted

have remained chemotherapy

Pathologic Imaging

in

intensity

unchanged.

patients

tumors areas

seen

grading

bow

two

tumor

could

in

none (Figs apy,

tive

mar-

a different in whom

was

6). In the

of

bone

of very

responded to dense cortical bone, within which small tumor nests could be found only at microscopy (Figs 5,

intermediate.

On the other

bone (Figs 5, 6). Areas of necrotic in

was

pa-

tumor

the signal on the MR

endosteal

faces.

signal

soft-tissue

out the specimens, thickened, dense

Obtained

on both

In ii

T2-weighted

of the

considered

before

soft-tissue

in all patients.

obtained RESULTS

the

small to be graded. signal intensity

In our

group

of patients,

MR images obtained before therapy consistently showed ing sarcoma had low signal on

Ti-weighted

diate

or high

T2-weighted

marrow finding reported

images

signal images

chemothat Ewintensity

and

intensity in both

and soft tissue (Figs that confirms previously findings (3,9).

interme-

on bone 1, 2), a

Radiology

#{149} 861

b.

C.

Figure

3.

after

Images

completion

obtained of chemotherapy

(b, d, e) 3 weeks

after

d.

(a, c) 10 days

surgery

and

in a boy aged

12 years. (a) T2-weighted (2,100/80) MR image obtained after chemotherapy shows areas of very high signal intensity in anterior left iliac bone. Vertical bar in a and b indicates

5 cm. (b) Fine-detail specimen

from

cavities

radiograph

the

patient

that contained

of resected

in a shows

extensive

large

necrosis

of

bone at histologic examination and that correspond to areas of high signal intensity in a. Small clusters of active tumor cells were not found in these cavities but were found in areas of bone necrosis with high signal intensity in other patients. (c) T2-weighted (2,100/80) MR image obtained 20 mm caudal to a shows an area of high signal intensity (small arrow) that represents a cystic area and an area of low signal intensity (large arrow) that contamed active tumor at histologic examination. (d) Low-power view of cystic area corresponds

to area

in c. is present, and hemosiderin within the cyst may cause the very low signal intensity in c, at the medial border of the cyst. (Hematoxylin-eo-

A small

of high

amount

sin stain; original power

signal

intensity

of hemorrhagic

magnification,

debris

X

25.) (e) Low-

view

corresponds to area of low signal intensity in C. A sheet of small, dark-stained active tumor cells appears between sclerotic

bone

and

lin-eosin

adjacent stain;

muscle

original

fibers.

magnification,

(Hematoxyx 100.)

Although the expected appearance of untreated Ewing sarcoma on MR images is established, less agreement exists about the importance of changes in signal intensity after chemotherapy. The importance of such changes after chemotherapy in Ewing sarcoma has not yet been fully evalu-

862

#{149} Radiology

e.

ated, although attention has been drawn to the increasing importance of MR images in follow-up evaluation of malignant bone tumor (13). Ideally, sites of active tumor would be reliably identified from patterns of change in signal intensity. Our study shows that this is impossible. Much of the radiology literature has tended to report appearances of Ewing sarcoma and osteosarcoma on MR images together (3,5,14) and may thus have caused

confusion.

These

two

tumors

are

dif-

ferent in their cell of origin, clinical course, radiologic appearance, and histologic characteristics. The most valuable data with histologic correbation come from the studies by Lemmi et ab (9) and Frouge et ab (15), which deal with Ewing sarcoma alone. Certain findings in our study deserve emphasis. The signal from tumor-bearing bone and soft tissue remained low in intensity on Ti-

September

1992

MR imaging before surgery, the same researchers have reported high-intensity T2-weighted

radiobogic practice at present. MR imaging was useful in confirming reduction in the volume of softtissue tumor. In many of our patients, chemotherapy reduced the soft-tissue tumor to a size at which evaluation of change in signal intensity became difficult. It has been reported that decrease in tumor volume correlates, albeit weakly, with a favorable histologic response (8). In our study, how-

signal

ever,

confirm intensity

that high T2-weighted signal may be seen in responsive

tumors tive

in areas tumor.

without

A similar

residual

ac-

appearance

has

been described after radiation therapy in a patient without histologic evidence of disease (16). In a group of patients underwent

tory mor,

with

Ewing

sarcoma

who

in association with inflammachange, necrosis, and active tualthough it is not clear whether

all their

patients

therapy

alone

confirm

that

areas tirely

underwent (15).

cherno-

While

our

Frouge Figure

4.

T2-weighted

of the humerus

obtained

(2,100/70)

after

MR

completion

image

of

chemotherapy in a girl aged ii years shows areas of very high signal intensity that were rich in delicate, edematous, highly vascular fibroblastic repair tissue. No active tumor was found in these areas, but malignant cells were found in the capitellum.

weighted images, but the intensities of T2-weighted signal became more inhomogeneous, and T2-weighted images contained localized areas of high and very high signal intensity in the bone marrow. These changes were observed in patients in whom histologic correlation was available as well as in those who have been fobbowed up with serial MR imaging. Once established, these changes in signal intensity have been shown to be stable on follow-up MR images in our four patients who did not undergo surgery. Pathologic correlation revealed several noteworthy findings. Areas of high signal intensity on T2-weighted images may represent tumor necrosis in bone, cystic hemorrhagic areas, and regions rich in highly vascular fibroblastic repair tissue. Lemmi et al (9) reported a similar increase in intensity of T2-weighted signal in the bone marrow component of Ewing sarcomas that had responded to chemotherapy. In their seven patients with histologic correlation (two patients who underwent resection and five patients who underwent multiple needle biopsies of the bone marrow), the high signal intensity was attributed to the presence of watery, myxoid connective tissue that contained no residual malignancy. Our findings

data

Volume

184

#{149} Number

3

all con-

in some to such

found

at microscopy

in 71%

tients.

The

of microscopic

tumor

despite

development

patients

supports

et al (15) that

the

areas

view

images.

Small

(or of

signal

soft-tissue

both

tumor

chemotherapy

is used

disease

in use)

exclusion

activity.

The

of redifficulty

is

in Ewing sarcoma by of this disease to leave

residual widely

tiny clusters of active scattered through the

tumor. tential

Any of these for subsequent

chemotherapy mor necrosis

cells area of

clusters has pogrowth; tumor

as a percentage difficult without

of tucorn-

plete tumor resection. Furthermore, the apparent differences between our results and those of other studies may be due to variations in methods of tumor sampling. Our study suggests that quantification of tumor necrosis by means of MR images alone is unlikely to be accurate, although devel-

has

opment

of the

of change qualitative

characteristic

in signal evidence

peutic effect. Confirmation ease

with

pattern

intensity produces of a chemothera-

of control

MR

imaging

of local

should

dis-

pre-

cede resection of the primary tumor, provided that one has excluded metastatic disease (in which MR imaging has no use). ity of choice

MR imaging for evaluation

is the modalof the ex-

tent of local disease (2,3) and enables the surgeon to plan a limb salvage procedure (4). Examination with MR imaging before chemotherapy is nec-

and

as

indicator of response (14). Although this method shows promise, involves relatively complex manipubalion of quantitative data and seems unlikely to be adopted as a routine

allow

imaging now

tends to recur locally late in the disease, sometimes after apparent clinical remissions of longer than 5 years. Such a pattern of residual disease makes quantification of response to

areas of reactive change, thus reducing its potential for specific tumor characterization (is). A technique of dynamic enhancement has been described whereby a change in rate of uptake of gadopentetate dimeglu-

after

MR

modality

reliably

of pa-

favorable

that

imaging

compounded the proclivity

been said to indicate a favorable response to chemotherapy (8,14). However, low T2-weighted signal intensity after chemotherapy is not a reliable indicator of tumor sterilization in Ewing sarcoma. Frouge et al (15) state that low T2-weighted signal intensity has a high sensitivity as a sign of tumoral sterilization, but our data suggest that this is not a safe assumption. Intravenous administration of gadopentetate dirneglumine has been

to enhance

any

sidual

of high

intensity

apparently

implies

cannot

tumors with bow signal intensity on T2-weighted images may contain active tumor (Fig 3c, 3e), and small clusters of active tumor cells were found in areas of dense reactive bone in 5everal patients (Figs 5, 6). Because of the characteristic scattered distribution of these tumor cells, it is difficult to imagine how these microscopic nests of cells could affect signal intensity, which will be low as a result of the surrounding reactive bone. Reduction

in T2-weighted

discovery

responses

a

evaluation was imposactive tumor was

of such

weighted

mine

were

of high signal intensity is encompatible with a favorable re-

the

signal intensity may nevertheless contam microscopic tumor. Another important observation is the presence of active tumor cells in areas of low signal intensity on T2-

proved

tumors

reduced in size; they were reduced

size that signal sible. However,

sponse to chemotherapy, the finding of isolated clusters of active tumor within areas of high signal intensity in two

soft-tissue

siderably patients,

an

it

essary

for

entire

area

such

planning

cised. surgery

It is to be hoped will improve

of tumor

because

should

that ablative the prognosis

Ewing sarcoma, a prognosis remained unchanged despite early

promise

Whenever sible,

MR

imaging

resection

in

that has the

of chemotherapy

surgical

the

be ex-

(i7).

is impos-

is important

Radiology

in fob#{149} 863

a.

Figure

6.

Images

ob-

tamed (a) 10 days after completion of chemo-

therapy and (b, c) after resection, performed 4 weeks after a was obtamed, in a girl aged 7 years. (a) Balanced T2weighted (2,100/30) MR image shows a curvilinear area of low signal intensity in the

upper

Figure

5.

surgical

Histologic

resection

completion 12 years,

section,

obtained

of the fibula

of chemotherapy shows clusters

at

5 weeks

after

in a boy of dark-stained

aged ac-

tive tumor cells in areas of dense cortical bone. MR images of this area showed a thickened signal void. (Hematoxylin-eosin stain; original

magnification,

x 100.)

humerus

was present on images regardless pulse sequence.

bone marrow cavity had predominantly high

signal

intensity

on T2-weighted

low signal of the cortex

indicates

considerable sclerotic

(c) High-power

response

because

it confirms

to chemotherapy

that

initial

tumors

Acknowledgments: line

(18).

The authors

thank

Pau-

3.

4.

864

Bloem JL, Bluemm RG, Taminiau AHM, van Oosterom AT, Stolk J, Doornbos J. Magnetic responce imaging of primary malignant bone tumors. RadioGraphics 1987; 7:42.5-445. Zimmer WD, Berquist TH, McCleod RA, et a!. Bone tumors: magnetic resonance imaging versus computed tomography. Radiology 1985; 155:709-718. Boyko OB, Cory DA, Cohen MD, Provisor A, Mirkin D, De Rosa PG. MR imaging of osteogenic and Ewing’s sarcoma. AJR 1987; 148:317322. Sundaram M, McGuire MH, Herbold DR. Wolverson MK, Heiberg E. Magnetic resonance imaging in planning limb salvage surgery for pnmary malignant tumors of bone. Bone Joint Surg [Am! 1986; 68:809-819.

#{149} Radiology

6.

and Jackie manuscript.

References

2.

5.

U

Kennedy, Janet MacDonald, Seckel for help in preparing the

1.

same cation,

curvilinear x250.)

bone

section area

b.

C.

thickening. that

corresponds

of a cluster of low

signal

(b) Fine-detail

radiograph

of the resected

humerus

shows

to the curvilinear area of low signal intensity in a. of active sarcoma cells in dense sclerotic bone from the intensity in a. (Hematoxylin-eosin stain; original magnifi-

is main-

tamed, and it has been proved to be the most sensitive modality in detection of local recurrence of muscuboskeletal

im-

ages. The intensity dense

low-up

that all MR of The

7.

8.

9.

10. 11.

Bloem iL Taminiau AHM, Eulkderink F, Hermans J, Pauwels EKJ. Radiologic staging of primary bone sarcoma: MR imaging, scintigraphy, angiography, and CT correlated with pathologic examination. Radiology 1988; 169: 805-810. Jurgens H, Exner U, Gadner H, et al. Multiplidisciplinary treatment of primary Ewing’s sarcoma of bone: a 6-year experience of a European co-operative trial. Cancer 1988; 61:2126. Jurgens H, Dunst J, Gobel U, et al. Improved survival in Ewing’s sarcoma with response based local therapy and intensive chemotherapy (abstr). Proc Am Soc Clin Oncol 1991; 10: 316. Holscher HC, Bloem JL, Nooy MA, Taminiau AHM, Eulderink F, Hermans J. The value of MR imaging in monitoring the effect of chemotherapy on bone sarcomas. AJR 1990; 154: 763-769. Lemmi MA, Fletcher BD, Marina NM, et al. Use of MR imagin to assess results of chemotherapy for Ewing s sarcoma. AIR 1990; 155:

12.

343-346.

18.

Pan C, Raymond AK, Carrasco CH, et al. Osteosarcoma: MR imaging after preoperative chemotherapy. Radiology 1990; 174:517-526. Sundaram M, McGuire MH. Computed tomography or magnetic resonance for evaluating the solitary tumor or tumor-like lesion of bone? Skeletal Radiol 1988; 17:393-401.

13. 14.

15.

16.

17.

Taber DS, Libshitz HI, Cohen MA. Treated Ewing sarcoma: radiographic appearance in response, recurrence and new primaries. AJR 1983; 140:753-758. Fletcher BD. Response of osteosarcoma and Ewing’s sarcoma to chemotherapy: imaging evaluation. AJR 1991; 157:825-833. Erlemann R, Sciuk J, Bosse A, et al. Response of osteosarcoma and Ewing sarcoma to preoperative chemotherapy: assessment with dynamic and static MR imaging and skeletal scmtigraphy. Radiology 1990; 175:791-796. Frouge C, Vanel D, Coifre C, Couanet D, Contesso C, Sarrazin D. The role of magnetic resonance imaging in the evaluation of Ewing sarcoma. Skeletal Radiol 1988; 17:387-392. Vanel D, Lacombe MJ, Couanet D, Kalifa C, Spielmann M, Genin J. Musculoskeletal tumors: follow-up with MR imaging after treatment with surgery and radiation therapy. Radiology 1987; 164:243-245. Stiller CA, Bunch KJ. Trends in survival for childhood cancer in Britain diagnosed 197185. Br J Cancer 1990; 62:806-815. Reuther G, Mutschler W. Detection of local recurrent

disease

in musculoskeletal

magnetic resonance imaging tomography. Skeletal Radiol

tumours:

versus computed 1990; 19:85-90.

September

1992

Ewing sarcoma: MR imaging of chemotherapy-induced changes with histologic correlation.

In a study group of 18 consecutive patients with Ewing sarcoma proved by means of biopsy, the signal intensity characteristics of tumor on magnetic re...
1MB Sizes 0 Downloads 0 Views