Musculoskeletal Kenneth R. Kaplan, MD #{149}Donald G. Mitchell, MD #{149}Robert M. Steiner, MD Simon Vinitski, PhD #{149}Vijay M. Rao, MD #{149}D. Lawrence Burk, MD #{149}Matthew

Radiology

Scott

Murphy,

#{149}

D. Rifkin,

MD

MD

Polycythemia Vera and Myeloflbrosis: Correlation ofMR Imaging, Clinical, and Laboratory Findings’ Magnetic resonance (MR) imaging was performed in 14 patients with biopsy-proved polycythemia vera (n 4) or myelofibrosis (n = 10) to determine whether MR imaging findings can be correlated with the clinicopathologic diagnosis and established (serum

clinical

lactate and cholesterol (spleen size). in the proximal patients

parameters

of severity

dehydrogenase [LDH] levels) and chronicity Evaluation of marrow femurs showed that

could

be categorized

into

three distinct groups based on anatomic patterns of normal fatty and abnormal low-signal-intensity (nonfatty) marrow in the femoral capital epiphysis

(FCE)

and

greater

trochan-

ter (GT). Patients with nonfatty marrow in both the FCE and GT (n = 8) had significantly higher serum LDH (P < .02) and lower serum cholesterol (P < .02) levels than patients with fatty marrow in at least the GT (n = 6). Splenic volume, as measured from MR images, was significantly greater in the myelofibrosis group than in the polycythemia vera group (P < .001). MR imaging provided a better understanding of these hematologic

disorders

and

ters for classification ent from conventional laboratory data.

novel

parame-

that are differhistologic and

Index terms: Bone marrow, 443.657, 443.659 Bone marrow, MR, 443.1214 #{149}Femur, MR. 443.1214 #{149}Polycythemia, 40.659 #{149} Spleen, abnormalities, 775.657, 775.659

Radiology

1992;

183:329-334

C

HRONIC

(CMD)

myeboproliferative comprises a group

hematopoietic

acterized laboratory

disorders

by distinctive features and

that

disease of are

clinical a wide

char-

and spec-

trum of bone marrow histologic changes, ranging from hypencellularity to fibrosis or sclerosis (1-7). This

study

focuses

on two such

polycythemia Polycythemia

vera vera

a proliferative

erythrocytic

disorders,

and myelofibrosis. is characterized

phase.

in bone

marrow

to cellular

bone CMD

a retrospec-

of 14 patients

with

study

biopsy-

proved CMD in which the MR imaging findings were compared with established clinical parameters of severity (elevated serum lactate dehydrogenase [LDH] level, low serum cholesterol level) (1,2) and chronicity (degree of splenomegaly) (1,15) of the disease.

Ap-

reticulin

and is a histologic feature common both postpolycythemic and agnogenic rnyeloid metaplasia (1,4,6,7). Magnetic resonance (MR) imaging can depict conversion or reconversion

of fatty

we performed

tive

by

proximately 15% of cases of polycythemia vera evolve into a spent phase, or postpolycythernic rnyeloid metaplasia (4-11). Agnogenic rnyeloid metaplasia is a disorder in which an antecedent proliferative phase is either nonexistent or clinically unrecognized (1,4). Myelofibrosis is defined as

an increase

direction,

to

marrow (12-14).

in Global would be

individuals with assessment of bone marrow preferable to the current practice of nonguided biopsy of the iliac crest, a procedure that is subject to sampling error and may not be representative of marrow changes in the remainder of the axial and appendicular skeleton. If the spectrum of MR imaging findings in CMD and the correlates of these findings can be elucidated, MR imaging may play an important role in characterizing the extent and/or phase of disease, evaluating disease progression, and monitoring the effects of treatment. As a step in this

I From the Department of Radiology, Thomas Jefferson University Hospital and Jefferson Medical College, 10th Floor, Main Bldg. 11th and Sansom Sts, Philadelphia, PA 19107 (K.R.K., D.C.M., R.M.S., SM., S.V., V.M.R.); Department of Radiology, DePaul Medical Center, Norfolk, Va (D.L.B.); and Department of Radiology, Albany Medical College, Albany, NY (M.D.R.). Received June 11, 1991; revision requested July 23; revision received December 13; accepted December 19. Address reprint requests to K.R.K. ‘ RSNA, 1992 See also the editorial by Jones (pp 321-322) in this issue.

PATIENTS

AND

METHODS

MR imaging of the abdomen, lumbosacral spine, pelvis, and lower extremities was performed in 14 patients with a 1.5-T system

(Signa;

GE

Medical

Systems,

Mil-

waukee). All patients had biopsy-proved CMD. Thirteen patients were in the age range of 50-85 years, with a mean age of 67 years. One Two patients

patient (patients

was 37 years old. 4 and I 1 in the Ta-

ble)

underwent

follow-up

and

33 months,

respectively,

initial

imaging

after

24

their

examinations.

The the

MR

MR imaging

following:

protocol

Initially,

consisted

contiguous

in-phase TI-weighted time [msec}/echo time

imaging Emseci

20)

spine

of the

lumbosacral

of sagittal

(repetition 250-500/ per-

=

was

formed with a section thickness of 5 mm. This was followed by coronal T2-weighted imaging (2,000/40-80) of the abdomen and lumbar spine, with a 10-mm section thickness and Ti-weighted the

2-mm

pelvis

obtained, and

2-mm

gaps. A coronal study (400-700/12,

and

proximal

with

a 7-mm

gaps.

femurs

section

In nine

in-phase 20, 22) was

of

then

thickness

patients,

addi-

tional coronal images of the pelvis and proximal femurs were obtained with chemical

shift

weighted

echo

times

opposed-phase

sequences,

identical

TI-

with

with

repetition

those

used

and

in the

in-phase examination. On opposed-phase images, the phases of the echoes from water and fat are opposed such that the net signal intensity is the absolute difference between the water and triglyceride signals

Abbreviations: ative disease, CT = greater

CMD

=

chronic

myeloprolifer-

femoral capital epiphysis, trochanter, LDH = lactate dehyFCE

=

drogenase.

329

Laboratory,

Clinical,

Patient/ Sex/Age (y)

and MR Imaging

Findings

Clinicopathologic Diagnosis

RBC Transfusion(s)

2/M/70 3/M/68

PV PV AMM

No No No

4/M/59

AMM

Yes

251 134 287 454*

1/M/84

5/M/62 6/M/37 7/M/64 8/M/70

164 222 143 180

(4.25) (5.75) (3.70) (4.65) 127 (3.30)i 178 (4.60)

No

672

460 595

PV

Yes No Yes No

350k 8771 766

77 (2.00) 122 (3.15)

1,096 517

89 (2.30) 114 (2.95)

AMM PPMM

Yes

12/F/85 13/F/71 14/M/55

PPMM PPMM AMM

Yes No No

in parentheses are SI units (millimoles per liter). AMM = agnogenic myeloid metaplasia, PV = polycythemia vera, RBC = red blood cell, XRT = irradiation. value at time of MR examination. Normal range for LDH level, 110-220 U/U; normal range

NM

=

1 1

1 1* 311

1,545i 464

(1.75) (1.90) (3.95) (2.75) (3.15) 86 (2.20*

380

Groups

424 155 1,740 NM

67 74 153 107 121

298

Marrow

(cm3)t

Yes

PV AMM PPMM AMM

Note-Values

Spleen Volume

66711 215

9/M/68 10/F/SO 11/M/65

myeloid

Serum Cholesterol Level(mg/dL)*

LDH Level (U/U)* Serum

2

Splenectomy Splenectomy Splenectomy Splenic XRT

2 3 3 3 3 3 311 3 3 3

540

1,714 3,2001 934

2,308 2,063

not measured,

PPMM

= postpolycythemic

metaplasia,

* Uaboratory mmol/L). t Measured from coronal MR images Marrow group 1 = normal femoral group 3 = abnormal FCE and CT. § Initial MR examination. I Follow-up MR examination.

of the abdomen. capital epiphysis

Normal range, 120-480 c&. (FCE) and greater trochanter

a.

(CT),

marrow

group

for cholesterol

2

=

abnormal

level,

150-250

FCE and normal

mg/dL

(3.90-6.45

CT, and marrow

b.

Figure gional

1.

MR

images

marrow

demonstrating

patterns

re-

in the proximal

fe-

murs are shown. In-phase (a) and opposedphase (b) Ti-weighted (700/20) coronal MR images of the proximal femurs in an 84-year-

old man with polycythemia vera demonstrate fatty marrow in the FCEs and GTs and nonfatty metaphyses

marrow that

in the femoral is predominantly

necks and hypercel-

lular (marrow group 1). (c) In-phase Tiweighted (600/20) coronal MR image of the femurs in a 62-year-old man with polycythemia vera demonstrates fatty marrow in the GTs and nonfatty marrow in the FCEs, femoral necks, row group

phase

metaphyses, 2). In-phase

(e) Ti-weighted

images

of the

old woman nonfatty hypercellular

femoral dominantly

(400/12)

proximal

with marrow and

necks

and cellular

and diaphyses (mar(d) and opposedfemurs

myelofibrosis

coronal

(marrow

MR

in a 71-year-

demonstrate

in the FCEs and nonfatty marrow

metaphyses

d C

that group

GTs that in the

is

is pre3).

e. 330

#{149} Radiology

May

1992

in each voxel T2-weighted 2,000/20,

40,

pelvis

and

(16-20). four-echo 60,

80)

Finally, a coronal sequence (1,500was

proximal

used

femurs,

to image

with

the

fat sup-

I I patients, the tibias in eight the feet in one patient. Bone marrow was classified conventional

Ti -weighted

patients,

and

as fatty

images

if

showed

technique. The hybrid fat-suppression technique involved selective presaturation of the triglyceride signal followed by real-

high signal intensity comparable with that of subcutaneous fat. Marrow with a signal intensity lower than that of subcutaneous fat on conventional Ti-weighted images

time

was

pression

achieved

subtraction

by

means

of raw

of a hybrid

data

phase and opposed-phase Additional examinations weighted imaging of the

from

the

in-

images (21). included TIdistal femurs in

designated

as nonfatty.

In some

cases,

than that of muscle on conventional inphase Ti-weighted images but lower than that of muscle on opposed-phase TIweighted images, owing to chemical shift signal cancellation (18,22); (b) hypercellular marrow, with signal intensity equal to or higher than that of muscle on both conventional in-phase and opposed-phase Ti-weighted images (19,20); and (c) fi-

when chemical shift opposed-phase weighted images were also obtained, was thought that nonfatty marrow be further characterized as follows:

Tiit could (a) cel-

brotic

lular

higher

groups on the basis of the signal intensity of the marrow in the femoral capital epiphysis (FCE) and greater trochanter

marrow,

with

signal

intensity

and/or

nal intensity all images The

siderotic lower (TI and

patients

marrow, than that T2 weighted)

were

with

sig-

of muscle (23-25).

divided

into

on

three

(CT) at the initial MR imaging examination (Table). Marrow group I patients = 4) had normal fatty marrow in both the FCE and CT. Marrow group 2 patients (n = 2) had nonfatty marrow in the FCE and normal fatty marrow in the CT. Marrow group 3 patients (n = 8) had nonfatty marrow in both the FCE and CT. Spleen volumes were derived from area measurements on sequential coronal MR images of the abdomen in 10 patients; three patients had undergone splenectomy, and one patient had received splenic irradiation. Two patients (patients

(n

4 and

Figure

2.

Ti-weighted

(600/20)

image of the proximal femurs old woman with myelofibrosis

coronal

MR

in an 85-yeardemonstrates

homogeneous nonfatty marrow in the right FCE and CT, with a mixture of fatty and nonfatty marrow in the right femoral neck,

Figure 4. TI-weighted (600/20) coronal MR image of the knees in a 71-year-old woman with myelofibrosis shows inhomogeneous

metaphysis,

marrow

and

diaphysis.

The

left

proximal

femur (out of plane in this image) showed the same regional marrow changes.

signal

metaphyses,

intensity and

mixture

of fatty

in the epiphyses,

diaphyses,

and

representing

nonfatty

ii)

was

a

underwent

follow-up

MR

imag-

ing. The spleen volume was not obtained during the initial examination of patient 4; therefore, the spleen volume obtained during the follow-up examination was used in the calculation of the average spleen volume in the myelofibrosis group. For patient 1 1, the initial spleen volume used

in the

calculations.

Serum UDH and cholesterol levels were obtained by means of venipuncture within a 3-week period either before or after each MR imaging examination.

marrow.

Statistical

with

significance

the Student

was

determined

t test.

RESULTS The clinical, laboratory, and MR imaging findings in the 14 patients in the study group are summarized in the Table.

Proximal

Femurs

Marrow signal intensity and regional anatomic patterns in the proximal femurs were bilaterally symmetric in all patients. Representative images from each of the three marrow groups are shown in Figure 1. All 14 patients had nonfatty marrow in the femoral neck and intertrochanteric region. Of the 10 patients who had nonfatty marrow in the CT and/or FCE (marrow groups 2 and 3), Figure

3.

image

of the

Ti-weighted knees

(600/20) coronal MR in an 85-year-old woman

with myeloflbrosis demonstrates homogeneous nonfatty marrow taphyses and diaphyses tibias, with fatty marrow

Volume

183

#{149} Number

relatively in the me-

of the femurs and in the epiphyses.

2

Figure

5.

image

of the

man

Ti-weighted

with

of fatty

the tibias

legs

(600/20)

and

feet

myelofibrosis

and

and

nonfatty

coronal

demonstrates marrow

the bones

MR

in a 70-year-old interspersed

of the hindfoot.

eight tensity

mal

areas in

ses

had similar marrow in the remainder

femur; and

marrow

two

patients

apophyses

was

signal of the

more

had

in which

cellular

inproxi-

epiphythe

than

Radiology

the #{149} 331

marrow in the femoral neck and proximal femonal metaphysis (Fig 2). Marrow group 3 patients (n = 8) had significantly higher serum LDH (505 U/L ± 164 vs 268 U/L ± 118, P < .02) and lower serum cholesterol (106 mg/dL ± 28 [2.75 mmol/ L ± 0.70] vs 177 mg/dL ± 29 [4.60 mmol/L ± 0.75], P < .02) levels at the time of the initial MR imaging examination when compared with the six patients in marrow groups 1 and 2

(b) marked decrease tensity of the marrow

level

(from

180 to i27

(fatty

3.30

mmol/L]),

and

CT)

Lower

combined

(Table).

Extremities

Images

of the

lower

extremities

showed a spectrum of regional marrow abnormalities, ranging from replacement of the marrow in the femoral and tibial diaphyses with sparing of the marrow in the epiphyses of the

knee

(Fig

3) to replacement

row in the knee ment of marrow lower extremities

of all mar-

(Fig 4) and in all parts (Fig 5).

to replaceof the

Spine

Marrow weighted spine

signal images

and

pulse

tient erately to markedly decreased in the other 13 patients, as seen on a representative image in Figure 7. Pelvic marrow in all 14 patients was nonfatty and in each case had more heterogeneous signal intensity than vertebral marrow.

serum U/L),

significantly

patients

larger

with

spleens

polycythemia

(i,717 cm3 P < .001).

472 vs 396 In fact, there

lap

range

of spleen

those cm3)

with myelofibrosis and in those with

vera

(155-540

cm3)

Progression

volumes

laboratory

(Table).

with

conversion

332

#{149}

and

Radiology

LDH

level

4 and ii) imaging 24 after Both demtheir disease,

MR imaging

and

of the

fatty

454

in serum

to 667 cholesterol

mg/dL

an

[4.65

increase

3,200

cm3);

and

LDH

(from

350 to 877 U/L)

in serum

(c) increase

[2.20

to

in the

in imde-

to

in serum

and

cholesterol

MR imaging sights

into

de-

(from

to 2.00

performed in nine patients, three with polycythemia vera and six with myelofibrosis. On the basis of a cornparison of marrow signal intensity relative to that of muscle on conventional in-phase and chemical shift opposed-phase Ti-weighted images of the proximal femurs, it was thought that nonfatty marrow could, in some instances, be further characterized as cellular, hypercellular, or fibrotic/siderotic. In our limited experience, fibrotic/siderotic marrow was the easiest to identify, as it manifests as strikingly low signal intensity with all pulse sequences. We do not, however, have histologic proof of proximal femoral marrow composition in any patient. Conventional and fat-suppressed T2-weighted images were not

86

mmol/L])

marrow

marrow;

of

has

bone

provided

marrow

new

in-

pathophysi-

obogy with respect to normal age-related changes (22,24,26-30) and in diffuse hematopoietic disorders (1214,24,28). Our study shows that MR imaging can depict a spectrum of bone marrow abnormalities in patients with polycythemia vera and myebofibrosis. All patients in our underwent

iliac

crest

bone

mar-

row biopsy, which is the current standard of practice for characterizing the phase and severity of disease. We do not have histologic proof of regional marrow composition in the lumbar spine and lower extremities. This does not, however, invalidate our study, which indicates that MR imaging findings in patients with polycythemia vera and myelofibrosis correlate with established laboratory (serum LDH and cholesterol levels) and clinisize)

parameters

of dis-

Marrow cellularity in polycythemia vera varies from nonmocellular to hypercellular (4,7). In most cases of polycythemia vera, there is no increase in marrow reticulin before development of the spent phase (7). In myelofibrosis, marrow

study in patient 4 three striking changes: CT to nonfatty

(from

a decrease

cal (spleen ease.

testing.

The follow-up demonstrated FCE

in

of Disease

as measured

the

± 168; no over-

(934-2,308 polycythemia

Two patients (patients underwent follow-up MR and 33 months, respectively, their initial examinations. onstrated progression of

(a)

had the

vera cm3 was

±

in the

than

consistent

red blood cell transfusion requirement. Patient 11 had nonfatty marrow the FCE and CT at both examinations. However, a similar pattern of MR aging and laboratory findings was noted on the follow-up study: (a) creased signal intensity of the bone marrow, liver, and spleen, indicating transfusional siderosis; (b) marked increase in spleen size (from 1,714

study

Size with myelofibrosis or agnogenic)

findings

as all

Conventional Ti-weighted images were found to be the most useful overall in characterizing normal ageappropriate versus abnormal marrow changes in the axial and appendicular skeleton. Chemical shift imaging was

DISCUSSION

normal for age in one pa(patient 2) (Fig 6) but was mod-

The 10 patients (postpolycythemic

femurs liver with

transfusional siderosis; and (c) marked subjective increase in spleen size (Fig 8). Furthermore, this patient had an interval increase in

intensity on Tiof the lumbosacral

was

Spleen

sequences,

in-

lumbar

with

to 77 mg/dL levels.

Pelvis

signal

in the

spine, pelvis, and proximal well as of the spleen and

crease

Lumbosacral

in the

cellularity

ranges

from

virtually iOO%, with a slight increase in reticulin, to virtual depletion of the hematopoietic elements, with dense fibrosis (4,i5).

Figure

6. TI-weighted (400/20) sagittal MR image of the lumbosacral spine in a 70-yearold man with polvcythemia vera demonstrates predominantly fatty marrow, which is normal for his age hut unusual for this patient population. May

1992

particularly useful in characterizing marrow composition. Three patients in the myelofibrosis population demonstrated MR imaging findings compatible with marrow and parenchymal (hepatic and splenic) siderosis, presumably second-

of 67 years. Histologic and MR imaging studies have shown that hematologically normal individuals in this age range have cellular (hematopoietic) marrow in the axial skeleton (skull, ribs, sternum, vertebrae, pelvis) and proximal appendicular skeleton,

tam cellular marrow in younger mdividuals (such as the femoral neck and intertnochanteric region).

ary to multiple transfusions (23-25). Again, we do not have histologic proof of these findings. Polycythemia vera and myelofibrosis affect older adults, usually in the 6th-8th decade of life. Thirteen patients in our study were in the age range of 50-85 years, with a mean age

excluding

disorder as determined with use of established laboratory parameters, namely, increased LDH and decreased cholesterol levels (1,2). Patients demonstrating nonfatty marrow in both the FCE and CT (n = 8) (marrow group 3, the most abnormal

the

head)

and

epiphyses

apophyses

(27-29,31-33).

of the

(eg,

An

proximal

(eg, MR

femoral

the

CT)

imaging

femur

has

study

shown

that

intermediate-signal-intensity (cellular) marrow is present in the intertrochanteric region of the femur in 95% of hematologically normal individuals less than 50 years of age but in only 12.5% of those older than 50 years (22). Furthermore, a recent study of normal age-related bone marrow patterns showed that when fatty marrow

occupied

all of the

proximal

femur,

group.

Ten

marrow

abnormal patients

in the

epiphyses, age after

in this had

abnormal the first few

to reconversion

One especially in our study was two patients had

apophyses more

in

cellular

remainder Figure

7. TI-weighted (600/20) sagittal MR image of the thoracolumbar spine in a 64year-old man with myelofibrosis demonstrates homogeneous nonfatty marrow in the vertebral bodies. The liver has abnormally low signal intensity, indicating parenchymal siderosis secondary to numerous transfusions.

This

fatty to cellular

at any of life

were

more

than

the

resisepiphy-

surprising finding the observation that epiphyses and

which the (less fatty)

of the

suggests

and/or findings months

(34). The apophyses tant ses.

age

nonfatty

apophyses

marrow was than in the

proximal

that

femur.

reconversion

marrow

does

of

not

relate

al-

patterns

with

of the

in the proximal polycythemia appear to cor-

the clinical

MR marrow

severity

classifications

marrow in the sion-dependent study,

four

epiphyses

had

abnormal at the

and apophyses

MR examination and veloped this marrow

244 months.

This

imaging

of the

suggests

that

proximal

MR

femurs

may

be useful in both staging and evaluating the progression of disease in patients with myelofibrosis (postpolycythemic or agnogenic), thereby

circumventing “blind” bone

the need for multiple marrow biopsies unless

malignancy

is suspected.

Our study relationship

also examined the of MR bone marrow size, and chronicity

ings, spleen

interfindof

disease in myelofibrosis. Splenomegaly is a major diagnostic criterion for both polycythemia vera and myelofibrosis (4,5). The cause of splenic enlargement has been debated; it has

occurs

and increase in spleen size reported, based on physical

normally

initial

one patient depattern within

been

that

in

CT. Of the six transfumyelofibrotic patients

ways follow a strict order-as has been suggested (26-28,31,33)whereby epiphyses and apophyses are involved only after reconversion in regions

of the

our study) had significantly higher serum LDH and lower serum cholesterol levels than patients with fatty

in the

89% of the patients were older than 50 years (26). All patients in our study had nonfatty marrow in the femoral neck and intertrochanteric region, findings that

are generally

Marrow

femurs of patients with vera and myebofibrosis

shown

that

the

enlargement

is

not due solely to extramedullary hematopoiesis (1,2,15). A positive comelation between duration of disease

con-

tion with

and splenic myelofibrosis

weight who

has been examina-

in individuals have under-

gone splenectomy (1,15). The results of volumetric analysis of the spleen in our study agree with those in the din-

ical literature: elofibrosis spleens

vera, in two

The patients had

than

and

significantly those with

follow-up

patients

with

my-

larger polycythemia

MR examinations

with

myebofibrosis

showed a large interval increase in spleen size in both cases. Histopathologic studies of patients with myebofibrosis initially suggested that increasing marrow fibrosis correlated with increasing spleen size (1,35). Howa. Figure

8.

MR images

demonstrating

parameters

b. of disease

progression

in a patient

with

my-

elofibrosis are shown. Initial (a) and follow-up (b) conventional T2-weighted (1,500/80 and 2,000/80, respectively) coronal MR images of the abdomen in a 59-year-old man with myelofibrosis demonstrate progressive enlargement of the spleen and development of abnormally low signal intensity in the liver, spleen, and spine (b), secondary to transfusional siderosis.

Volume

183

Number

#{149}

2

ever, more recent not substantiated (15,36). phasizes usually

investigations those reports

have

In any event, our study emthat massive splenomegaly associated with myelofibrosis

Radiology

is

333

#{149}

as opposed

to polycythemia

eds. Hematology.

vera.

Imaging of the lower extremities (distal to the proximal femur) showed a wide spectrum of regional marrow abnormalities, demonstrating that the appendicular skeleton is often affected in patients with polycythemia vera and myeloflbrosis. However, the clinical utility of routinely including the lower extremities in an imaging protocol for such patients remains to be determined. In conclusion, MR imaging is a noninvasive means of evaluating bone marrow in patients with myeboproliferative disorders such as polycythemia vera and myelofibrosis. The information obtained with MR imaging is different from and complementary to that obtained with iliac crest biopsy. MR imaging can be used to assess the entire

marrow

compartment;

fore, in conjunction and clinical data, help characterize and progression with polycythemia brosis. U

Craw-Hill,

6.

H. Studies themia vera

brosis

9.

10.

11.

12.

there-

era Von script.

The authors for preparation

Colclough

13.

14.

thank UaUivof the manu-

16.

References 1.

Ward HP, Block MH. of agnogenic myeloid and

2.

a critical

The natural history metaplasia (AMM)

evaluation

of its relationship

with the myeloproliferative Medicine 1971; 50:357-420. Uaszlo J. Myeloproliferative

syndrome.

4.

stem Wolf BC, Neiman

5.

P, Levitt U. Hematopoietic cells. N Engl J Med 1979; 301:868-872.

RS.

The bone

in myeloproliferative

and

syndromes. Hematol 1988; 2:669-694.

Oncol

Murphy

S.

Polycythemia E, Erslev

liams W, Beutler

marrow

P. Celler

Annu

i979;

14:

SA, Rappaport

23.

of the bone marrow in polycyand the evolution of myelofi-

hematologic malignan1986; 23:144-155. Minot CR, Buckman TE. Erythremia (polycythemia rubra vera). Am J Med Sd 1923; 166:469-489. Wasserman LR. Polycythemia vera-its course and treatment: relation to myeloid metaplasia and leukemia. Bull N Y Acad Med 1954; 30:343-375. Ikkala E, Rapola J, Kotilainen M. Polycythemia vera and myelofibrosis. Scand Haematol 1%7; 4:453-464. Silverstein MN. Postpolycythemia myeloid metaplasia. Arch Intern Med 1974; 134:113-115. Steiner RM, Mitchell DC, Rao VM, et al. Magnetic resonance imaging of bone marrow: diagnostic value in diffuse hematologic disorders. Magn Reson Q 1990; 6:17-

vera. In: WilA, Uichtman M,

Porter BA, Shields AF, Olson DO. Magnetic resonance imaging of bone marrow disorders. Radiol Clin North Am 1986; 24: 269-289. Lanir A, Aghai E, Simon JS, Uee RCL, Clouse ME. MR imaging in myelofibrosis. J Comput Assist Tomogr 1986; 10:634-636.

Wolf BC, Neiman RS. Myelofibrosis with myeloid metaplasia: pathophysiologic implications of the correlation between bone marrow changes and progression of splenomegaly. Blood 1985; 65:803-809. Dixon WT. Simple proton spectroscopic imaging. Radiology 1984; 153:189-194. Brateman U. Chemical shift imaging: a review. AJR 1986; i46:97i-980.

24. 25.

fatty

M, et al.

marrow

distribu-

27.

CS, Siegel

MJ, Donaldson

musculoskeletal

MR imag-

thalassemia

JS.

Pediatric

ing.

Radiology

1991;

ES, Mitchell

Parenchymal

versus

DC, Rubin

R, et al.

reticuloendothelial

in the liver: distinction Radiology

1991;

with

179:361-366.

Ricci C, Cova M, Kang YS, et al.

Normal

age-related patterns of cellular and fatty bone marrow distribution in the axial skeleton: MR imaging study. Radiology 1990; 177:83-88. Moore SC, Dawson KL. Red and yellow marrow in the femur: age-related changes in appearance at MR imaging. Radiology 1990; 175:219-223.

28.

VogIerJB, Murphy imaging. Radiology

29.

Dooms

30.

179:345-360.

Siegelman

MR imaging.

26.

major.

imagcompliRadiology 1984;

150:767-771. Moore SC, Bisset

iron overload

CC,

WA. Bone marrow 1988; 168:679-693.

Fisher

MR.

Hricak

H, Richard-

son M, Crook row imaging:

UE, Cenant HK. Bone marmagnetic resonance studies

related 429-432.

to age

and

ME.

Red-yellow

Kricun sion:

its effect

tary bone

sex.

on the

lesions.

Radiology

1985;

marrow

location

Skeletal

155:

conver-

of some

Radiol

soli-

1985; 14:

10-19.

31.

Custer RP, Ahlfeldt FE. Studies on the structure and function of bone marrow. II. Variations in cellularity in various bones with

advancing

tive response 32.

1932; 17:960-962. Hashimoto M.

years

20.

temic bone marrow disorders: characterization with proton chemical shift imaging. Comput Assist Tomogr 1990; 14:633-642.

36.

of life and

to stimuli. Pathology

row. Acta Haematol

35.

21.

Mitchell Chemical

and

tion in the normal and ischemic hip: new observations with 1.5-T MR imaging. Radiology 1986; 161:191-202. Brasch RC, Wesbey CE, Cooding CA,

cating

Hematol

Hematologic bone marrow disorders: quantitative chemical shift MR imaging. Radiology 1988; 169:799-804. Cuckel F, Brix C, Semmler W, et al. Sys-

Am

DC, Rao VM, Dalinka

Koerper MA. Magnetic resonance ing of transfusional hemosiderosis

19.

dysmyelopoietic

Clin North

Pathol

in

DG,Joseph PM, Fallon M, et al. shift MR imaging of the femoral head: an in vitro study of normal hips and hips with avascular necrosis. AJR 1987; 148: 1159-1164. Rosen BR, Fleming DM, Kushner DC, et al.

disorders

ated MPD, and hemorrhagic thrombocythemia. Semin Hematol 1975; 12:409-432. Quesenberry

i7. 18.

(MPD): myelofibrosis, myelosclerosis, extramedullary hematopoiesis, undifferenti-

3.

The bone marrow

34.

with laboratory MR imaging can the phase, severity, of disease in patients vera and myebofl-

15. Acknowledgment:

vera.

Mitchell

Hematopoietic

and second

cies. Semin

8.

P.

Effis JT, Peterson

22.

Mc-

i93-202.

Ellis JT, Peterson polycythemia 383-403.

7.

4th ed. New York:

1989;

their

rela-

J Lab Clin Med of bone

mar-

1962; 27:193-216.

33.

Piney A. The anatomy of the bone row. Br Med J 1922; 2:792-795.

34.

Jaramillo

al. MR imaging of epiphyseal marrow in infancy (abstr). Radiology 1990; 177(P):i6i. PitcockJA, Reinhard EH, Justus BW, Men-

D, Buchbinder

BR, Hoffer

mar-

FA, et

delsohn RS. A clinical and pathologic study of seventy cases of myelofibrosis. Ann Intern Med 1%2; 57:73-84. Bentley SA, Herman C). Bone marrow fiber production in myelofibrosis: a quantitative study. BrJ Haematol 1979; 42:51-59.

SzumowskiJ, Eisen JK, Vinitski 5, Haake PW, Plewes DB. Hybrid methods of chemical shift imaging. Magn Reson Med 1989; 9:379-388.

334

Radiology

#{149}

May

1992