DOI:10.2214/AJR.05.0086
AJR 2006; 186:1761-1770
© American Roentgen Ray Society
Relationship Between Bone Marrow Edema and Development of Symptoms in Patients With Osteonecrosis of the Femoral Head
Hiroshi Ito1,
Takeo Matsuno1 and
Akio Minami2
1 Department of Orthopedic Surgery, Asahikawa Medical College, Midorigaoka
Higashi 2-1-1-1, Asahikawa 078-8510, Japan.
2 Hokkaido University School of Medicine, Hokkaido, Japan.
Received January 18, 2005;
accepted after revision July 6, 2005.
Address correspondence to H. Ito
(itobiro{at}asahikawa-med.ac.jp).
Abstract
OBJECTIVE. The purpose of this study was to determine the
significance of risk factors on MR images for predicting the outcome of
patients with osteonecrosis of the femoral head.
SUBJECTS AND METHODS. Eighty-three asymptomatic or minimally
symptomatic hips in 61 consecutive patients were followed up prospectively.
Inclusion criteria included osteonecrosis of the femoral head identified by
typical MRI findings, with no radiographic evidence of progression of
collapse, and Harris hip score of 85 points or more. Every 3 months, the
patients underwent clinical and radiographic examination, and MR images were
obtained at 6- to 12-month intervals. Follow-up continued until worsening of
hip pain with Harris hip score of less than 70 points or surgery. Asymptomatic
hips were followed up for at least 24 months. The mean clinical and
radiographic follow-up period after the initial diagnosis was 60 months
(range, 3-168 months).
RESULTS. Thirty-six (43%) of the 83 hips were symptomatic at the
last follow-up. Bone marrow edema was present in 28 hips (34%) during the
follow-up period. Twenty-seven (96%) of the 28 hips were symptomatic, and bone
marrow edema significantly correlated with worsening of hip pain (p
< 0.0001). The necrotic volume of hips with bone marrow edema was
significantly larger than those without bone marrow edema (p <
0.0001). Bone marrow edema was found to be the most significant risk factor
for worsening of pain (p < 0.0001).
CONCLUSION. Bone marrow edema strongly correlated with necrotic
volume and was the most significant risk factor for worsening of hip pain. A
large necrotic volume of 30% or more may be the second useful indicator for
predicting future worsening of hip pain.
Keywords: bone marrow • edema • ischemia • MRI • orthopedic surgery • osteonecrosis
Introduction
Disabling hip pain with progressive collapse is the main
complication of nontraumatic osteonecrosis of the femoral head. In studies
using MRI, poor results of worsening pain were reported in 23-62% of patients
[1-7].
Many factors should be evaluated to predict the outcome of patients with
osteonecrosis. The extent and location of the necrotic lesion have been
reported to affect the prognosis
[1,
2,
5-10].
Several studies have shown that bone marrow edema at the femoral head and neck
on MR images was observed in patients with osteonecrosis, and this signal
abnormality has been reported to have a strong association with hip pain and
risk of radiographic collapse
[2-4,
7,
11-18].
Recently, the necrotic volume was reported to be one of the most
significant factors for worsening of hip pain
[5,
6,
10]. This study prospectively
investigated the risk factors on MR images to predict the outcome in patients
with osteonecrosis of the femoral head, particularly assessing bone marrow
edema, necrotic volume, and their relationship to the occurrence of
symptoms.
Subjects and Methods
Between February 1990 and July 2002, 97 asymptomatic or minimally
symptomatic hips in 61 consecutive patients with early-stage nontraumatic
osteonecrosis of the femoral head were followed prospectively. These patients
were identified by random screening of high-risk groups of patients who
underwent corticosteroid therapy or those with asymptomatic or minimally
symptomatic hips with contralateral symptomatic hips.
The diagnosis was made from conventional radiographs, MR images, or both.
The necrotic lesions of all patients were confirmed by crescent signs of low
signal intensity in the weight-bearing portion of the femoral head, bandlike
lesions (Figs. 1A,
1B,
1C,
1D and
1E), ringlike lesions with
homogeneous or inhomogeneous central areas, diffuse decreased signal intensity
with a dark band (Figs. 2A,
2B,
2C,
2D and
2E), or collapse of the
femoral head on MR images [11,
18,
19].
View larger version (134K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1A —Right femoral head in 48-year-old woman who received high-dose
steroid therapy for dermatomyositis. Coronal T1-weighted MR image (TR/TE,
450/17) obtained 12 months after initiation of therapy and 4 weeks after onset
of hip pain shows decreased signal intensity in femoral head and neck.
Necrotic volume is 29.8%.
|
|
View larger version (137K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1B —Right femoral head in 48-year-old woman who received high-dose
steroid therapy for dermatomyositis. Coronal T2-weighted MR image (3,000/102)
shows high-signal-intensity bone marrow edema in femoral head and neck.
|
|
View larger version (157K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1C —Right femoral head in 48-year-old woman who received high-dose
steroid therapy for dermatomyositis. Anteroposterior radiograph shows minimum
collapse of femoral head. Harris hip score
[20] is 93 points.
|
|
View larger version (157K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1D —Right femoral head in 48-year-old woman who received high-dose
steroid therapy for dermatomyositis. Anteroposterior radiograph 3 months after
A-C shows radiolucent zone and clear demarcation line.
|
|
View larger version (164K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1E —Right femoral head in 48-year-old woman who received high-dose
steroid therapy for dermatomyositis. Frog-leg lateral radiograph shows
progression of collapse. Patient reports right hip pain but refuses surgery.
Harris hip score is 48 points.
|
|
View larger version (142K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 2A —Left femoral head of 25-year-old woman who received high-dose
steroid therapy for systemic lupus erythematosus. Coronal T1-weighted MR image
(TR/TE, 500/17) obtained 18 months after initiation of therapy and 2 weeks
after onset of hip pain shows decreased signal intensity in femoral head.
Necrotic volume calculated on sequential T1-weighted images is 24.5%.
|
|
View larger version (139K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 2B —Left femoral head of 25-year-old woman who received high-dose
steroid therapy for systemic lupus erythematosus. Coronal T2-weighted MR image
(3,000/102) shows high-signal-intensity bone marrow edema in femoral head and
neck.
|
|
View larger version (141K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 2C —Left femoral head of 25-year-old woman who received high-dose
steroid therapy for systemic lupus erythematosus. Frog-leg lateral radiograph
appears to show normal findings. Harris hip score
[20] is 96 points.
|
|
View larger version (141K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 2D —Left femoral head of 25-year-old woman who received high-dose
steroid therapy for systemic lupus erythematosus. Anteroposterior radiograph
12 months after A-C shows radiolucent zone and sclerotic changes
in femoral head.
|
|
View larger version (137K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 2E —Left femoral head of 25-year-old woman who received high-dose
steroid therapy for systemic lupus erythematosus. Frog-leg lateral radiograph
shows progressive collapse of femoral head. Patient reports left hip pain, and
femoral transtrochanteric rotational osteotomy is being considered. Harris hip
score is 66 points.
|
|
Clinical evaluation was performed according to the Harris hip-scoring
system [20]. Minimally
symptomatic hips were defined as those showing occasional or low-grade hip
pain and a hip score of 85 points or more. Minimally symptomatic early-stage
osteonecrosis was found in 87 hips of 51 patients; 36 patients had bilateral
and 15 had unilateral involvement. Asymptomatic osteonecrosis was found in 10
hips of 19 patients with contralateral radiographic advanced osteonecrosis. Of
these, 14 hips in 14 patients with bilateral involvement could not be followed
up by MRI because the contralateral hip underwent surgery involving metal
within 2 years after diagnosis. Surgery included eight femoral osteotomies,
three bipolar hemiarthroplasties, and three total hip arthroplasties. These 14
hips were excluded from this analysis because it was not clear whether they
showed bone marrow edema over time. The remaining 83 hips in 61 patients were
included in this study; the subjects were 20 men and 41 women. Osteonecrosis
was idiopathic in six patients, alcohol-induced in seven patients, and
high-dose steroid-induced in 48 patients. The mean patient age when the MR
images initially revealed necrotic lesions was 35 years (range, 15-74
years).
All MRI examinations were performed on a 1.5-T superconductive unit
(Magnetom Vision, Siemens Medical Solutions; or Signa, GE Healthcare). Imaging
was performed using a body coil with the patient in the supine position. We
obtained the following sequences: coronal and axial T1-weighted spin-echo
images (TR range/TE range, 360-600/9-20) and coronal and axial
T2-weighted fast spin-echo images (1,800-3,800/90-102) of both
hips. The field of view was 30-38 cm. Slice thickness was 4-6 mm with
an interslice gap of 0.3-0.5 mm. An average of 8-10 slices were evaluated for
each femoral head. Most images were reconstructed using a 256 x 256 data
matrix.
Bone marrow edema was defined as an ill-defined area of low signal
intensity on T1-weighted images with corresponding high signal intensity on
T2-weighted images that involved the femoral head and neck beyond the necrotic
zone and extending to intertrochanteric regions
[2-4,
7,
11-18].
A well-demarcated arcuate zone of high signal intensity on T2-weighted images
surrounding the necrotic area of the femoral head was considered granulation
tissue, which was distinguished from bone marrow edema
[16,
18]. Hips with reactive edema
completely surrounding the necrotic fragment (an extensive double-line sign)
that did not extend into the femoral neck were classified as not having bone
marrow edema. A subchondral insufficiency fracture of the femoral head was
distinguished from osteonecrosis as follows: The shape of the low-intensity
band in osteonecrosis on T1-weighted images is usually concave to the
articular surface, whereas in a subchondral insufficiency fracture the
low-intensity band often parallels the articular surface and has a serpiginous
shape that does not circumscribe lesions; patients with a subchondral
insufficiency fracture generally include elderly women with osteoporosis who
are often obese; and the patient's pain usually has an acute onset
[21].
The necrotic volume was measured on the first abnormal MR images and the
percentage of involvement of the femoral head was measured. MR images were
digitized with a scanner using image software (Photoshop, Adobe Systems). To
calculate the necrotic lesions of the entire femoral head on each image, an
image analysis software package was used (NIH Image, National Institutes of
Health). Initially, the circumferences of the femoral head and of the necrotic
area as seen on sequential T1-weighted coronal images were outlined using a
nonpermanent fine-tip marker (Figs.
3A,
3B). The software program
measures the area of the outlined structure on each particular slice. The
necrotic volume was calculated by the integration of each sequential coronal
image. The measured areas were multiplied by the sum of the slice and
interslice thicknesses of the MR image to determine the volume of the outlined
structures. After the volume of the necrotic segment and the volume of the
entire femoral head were determined, the percentage of femoral head
involvement was calculated.
View larger version (134K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 3A —Measuring necrotic volume of femoral head using NIH Image software
(National Institutes of Health). Initially, circumference of femoral head
(dotted line) as seen on each sequential T1-weighted coronal image is
outlined using a nonpermanent fine-tip marker.
|
|
The necrotic locations on T1-weighted central coronal images were
classified as type A, type B, or type C using the criteria described by Sugano
et al. [1]. Type A lesions
occupy the medial one third or less of the weight-bearing portion of the
acetabulum; type B, the medial two thirds or less of the weight-bearing
portion; and type C, more than two thirds of the weight-bearing portion
(Fig. 4).
View larger version (7K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 4 —Classification of necrotic locations according to criteria of Sugano
et al. [1]: lesions occupying
medial one third or less (type A), two thirds or less (type B), or more than
two thirds (type C) of weight-bearing portion.
|
|
Radiographically, anteroposterior and Lauenstein lateral (lateral view of
the femoral head and neck) views were obtained at each examination.
Independent of the MR images, the radiographic stage was determined according
to the classification of Steinberg et al.
[9]: stage I, normal findings
on radiography; stage II, cystic and sclerotic changes; stage III, subchondral
radiolucency or crescent sign; stage IV, flattening of the femoral head; stage
V, joint space narrowing; and stage VI, advanced degenerative change.
Initially, 55 hips were classified as stage I, 15 as stage II, and 13 as stage
III. No collapse was found in stage I and stage II hips, but all 13 stage III
hips presented collapse on initial MR images. Radiographic factors include the
necrotic stage, progression of staging, and initial 2D percentage of
involvement of necrotic lesions seen on the anteroposterior view. The 2D
percentage of necrotic lesions was calculated using the same image analysis
software as described previously. The MR images and radiographs were reviewed
and interpreted by consensus of the experts on diagnosis of this disease,
excluding residents, in a clinical conference. The observers were not blinded
to clinical and radiographic assessments, and intra- or interobserver
variability studies were not performed.
All patients were followed up prospectively by clinical evaluation,
conventional radiography, and MRI. Every 3 months, the patients underwent
clinical and radiographic examination, and MR images were obtained at 6- to
12-month intervals. Follow-up continued until the patient became symptomatic
or underwent surgery. Symptomatic patients were defined as those complaining
of mild or more severe pain with a hip score of less than 70 points to
simplify the clinical outcome. Asymptomatic hips were followed up for at least
24 months. The mean clinical and radiographic follow-up period after the
initial diagnosis was 60 months (range, 3-168 months), and the mean MRI
follow-up period was 48 months (range, 0-168 months). Informed consent
was obtained before the patients were enrolled in this study.
Statistical analyses of the radiographic data were performed using StatView
statistical software (SAS Institute). Sex, age, etiologic factors, and various
factors on radiographs and MR images were analyzed. Multivariate regression
analysis was performed using the Cox proportional hazards model to evaluate
various factors. The Cox model was applied to each variable to separately
screen for possible risk factors affecting survival, and then the selected
possible risk factors were examined together to determine their contribution
to survival. Based on the final model, a p value was calculated for
each parameter. A probability value of less than 0.05 was considered
significant. The cumulative rates of survival were calculated, with a hip
score of less than 70 points or the need for surgery as the end point.
Results
Thirty-six (43%) of the 83 hips were symptomatic at the latest follow-up
examination. The average Harris hip score
[20] was initially 96 points
(range, 85-100 points) and finally 78 points (range, 28-100
points). Correlation of the initial and final radiographic stages is
summarized in Table 1. All 36
symptomatic hips showed radiographic collapse. Progression of staging was
found in 42 (51%) hips. New collapse developed in 26 (52%) of 50 hips that
were initially classified as stage I or stage II without collapse. At the time
of follow-up, none of 44 hips without collapse (final stage, I or II) were
symptomatic, whereas 36 of 39 hips with collapse (final stage, III, IV, or V)
were symptomatic (chi-square test, p < 0.0001). Twenty-two of
these 36 symptomatic hips had undergone surgery: intertrochanteric varus
osteotomy (n = 5), transtrochanteric anterior rotational osteotomy
(n = 4), bipolar prosthesis (n = 5), and total hip
arthroplasty (n = 8).
Bone marrow edema was present in 28 (34%) hips during the follow-up period.
Twenty-one of these 28 hips showed bone marrow edema on the diagnostic MR
images in which osteonecrosis was initially identified. Seven hips did not
show bone marrow edema on those images. The average period between initial
diagnosis of osteonecrosis and development of bone marrow edema in these seven
hips was 16 months (range, 6-24 months). Bone marrow edema was
typically identified around the time of onset of hip pain. Bone marrow edema
was seen at stage I in three (11%) hips, stage II in five (18%) hips, stage
III in 18 (64%) hips, and stage IV in two (7%) hips. Twenty-seven (96%) of
these 28 hips were symptomatic at the final follow-up evaluation
(Table 2), and bone marrow
edema significantly correlated with the symptoms (chi-square test, p
< 0.0001). All 21 hips that showed bone marrow edema on initial diagnostic
MR images for osteonecrosis were symptomatic at the final follow-up, which was
significantly higher than the 62 hips without bone marrow edema on those
images, of which 15 (24%) hips were symptomatic (chi-square test, p
< 0.0001). The radiographic stage at the time of follow-up of the 28 hips
with bone marrow edema was stage III in 18 hips, stage IV in seven hips, and
stage V in three hips, which was significantly more advanced than the 55 hips
without bone marrow edema during follow-up (chi-square test, p <
0.0001).
The average necrotic volume was 28.5% (range, 3.0-70.2%): 39.5%
± 13.3% (SD) in the 36 symptomatic hips and 19.9% ± 16.5% in the
47 asymptomatic hips. In 35 hips with a large necrotic volume of 30% or more,
25 (71%) hips were symptomatic at the final follow-up, which was significantly
higher than 48 hips with a small necrotic volume of less than 30%; only 11
(23%) of these hips were symptomatic (chi-square test, p <
0.0001). The necrotic volume of hips with bone marrow edema was significantly
larger than in those hips without bone marrow edema (Mann-Whitney U
test, p < 0.0001). The radiographic stage at the time of follow-up
had significantly progressed in hips with bone marrow edema (chi-square test,
p < 0.0001).
The average necrotic lesion on T1-weighted central coronal images was 28.2%
(range, 0-75.9%): 38.6% ± 13.4% in the 36 symptomatic hips and
20.2% ± 13.7% in the 47 asymptomatic hips. According to the necrotic
location, 18 hips were classified as type A, 13 hips as type B, and 52 hips as
type C lesions. In relation to the classification of MR images, none of the 18
type A hips, five (38%) of the 13 type B hips, and 31 (60%) of the 52 type C
hips were symptomatic at the final follow-up. In radiographic assessment, the
average involvement of lesions was 19.4% (range, 0-60.0%) at final
follow-up.
With a hip score of less than 70 points or the need for surgery as the end
point, the cumulative rates of survival were 72.3% (95% confidence interval
[CI], 67.4-77.2%) at 1 year, 61.4% (56.1-66.7%) at 2 years,
57.3% (51.8-62.8%) at 5 years, and 54.7% (48.9-60.5%) at 10
years. Figure 5 shows the
cumulative rate of survival comparing hips with and those without bone marrow
edema on diagnostic MR images in which osteonecrosis was initially identified.
No significant factors were found with application of the Cox model in age
(p = 0.5564). Male sex (p = 0.0054), alcohol-related cause
(p = 0.0086), initial stage at radiography (p = 0.0005), and
percentage of necrotic lesions (p < 0.0001) were possible risk
factors. On MR images, the presence of bone marrow edema (p <
0.0001), the percentage of necrotic volume (p < 0.0001), the
percentage on coronal images of central necrotic lesions (p <
0.0001), and the necrotic location type (p = 0.0008) were also
possible risk factors. When these eight possible factors were analyzed
together using the Cox model, the presence of bone marrow edema (p
< 0.0001) was found to be the most significant risk factor.
View larger version (5K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 5 —Line graph shows cumulative rate of survival comparing hips with
(dotted line) and those without (solid line) bone marrow
edema on diagnostic MR images in which osteonecrosis was initially
identified.
|
|
Discussion
Previous studies reported that diffuse signal abnormalities in the marrow
of the femoral head and neck were found in early-stage osteonecrosis
[2-4,
7,
11-18].
These abnormal signals consisted of decreased signal intensity on T1-weighted
images and a matching increase in signal intensity on T2-weighted images,
indicating bone marrow edema of elevated free-water content. Some authors have
proposed that bone marrow edema might be the initial MRI finding in early
osteonecrosis [12,
13]. However, recent studies
have reported that the initial MRI finding of early osteonecrosis is a
bandlike pattern, that bone marrow edema is not the initial MRI finding after
the onset of hip pain, and that bone marrow edema correlated highly with the
subsequent collapse of the femoral head
[2,
4,
15-18].
Bandlike lesions, ringlike lesions, or diffuse decreased signal intensity with
a dark band, with or without collapse on MR images, was the initial MR finding
in most cases in our study, which was consistent with those recent studies.
Separately, the most significant factor for clinical worsening of hip pain in
our study was the occurrence of bone marrow edema during the follow-up
period.
Using computer software, we measured the necrotic volume on sequential
coronal MR images and found that the necrotic volume was one possible risk
factors to predict the outcome. The necrotic volume and bone marrow edema
might be important factors; however, some hips with a large necrotic volume
remained asymptomatic without progressive collapse, and some hips with a
moderate necrotic volume became symptomatic with progressive collapse.
Therefore, the necrotic volume measured on initial diagnostic MR images may be
the second useful indicator to predict the outcome.
The mechanism that produces bone marrow edema remains unknown, although
some investigators have speculated that bone marrow edema could be a secondary
reaction to subchondral fractures arising from mechanical stress
[2,
4,
15-18].
Koo et al. [3] proposed that
edema adjacent to necrosis is a secondary reaction to ischemia and that
localized edema is a secondary phenomenon after an ischemic attack in the area
surrounding the necrotic region, such as brain ischemia. In this explanation,
however, it is not clear why all necrotic femoral heads do not always show
bone marrow edema in the course of the disease, and why the time discrepancy
between the occurrence of necrosis and the occurrence of bone marrow edema is
not consistent.
Kubo et al. [2] suggested
that bone marrow edema may be an inflammatory change in the reparative process
after collapse of the femoral head because it was detected most frequently in
stage III hips (classification of Steinberg et al.
[9]) and because two stage II
hips with bone marrow edema collapsed and progressed to stage III within 6
months after the MR examination. Sakai et al.
[16] reported that
extralesional reactive changes with diffuse enhancement were detected along
the boundary most frequently in stage III hips in the form of dilated vessels
and bone marrow edema, which may represent an inflammatory change in the
reactive process after collapse of the femoral head. They proposed that even
if a hip has not radiologically collapsed, the detection of bone marrow edema
on MRI might indicate that the femoral head has begun to collapse at the
microscopic level. Kim et al.
[15] reported that bone marrow
edema on MR images was only detectable when the subchondral fracture was just
beginning. They believed that a bone marrow edema pattern appeared as a
secondary reaction to subchondral fracture.
View larger version (133K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 6A —Left femoral head of 16-year-old girl who received steroid therapy
for systemic lupus erythematosus. Coronal T1-weighted MR image (TR/TE, 360/9)
obtained 12 months after initiation of therapy and 4 weeks after onset of hip
pain shows a bandlike hypointense zone in femoral head. Necrotic volume is
38.5%.
|
|
View larger version (132K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 6B —Left femoral head of 16-year-old girl who received steroid therapy
for systemic lupus erythematosus. Coronal T2-weighted MR image (3,800/102)
shows bandlike hypointense zone in femoral head. Bone marrow edema is not
found. Harris hip score [20]
is 88 points.
|
|
View larger version (135K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 6C —Left femoral head of 16-year-old girl who received steroid therapy
for systemic lupus erythematosus. Coronal T1-weighted MR image (500/20)
obtained 5 years after A and B shows almost same necrotic area
in femoral head. No collapse is found. Bone marrow edema was not found during
this period.
|
|
View larger version (137K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 6D —Left femoral head of 16-year-old girl who received steroid therapy
for systemic lupus erythematosus. Frog-leg lateral radiograph 5 years after
A and B appears to show normal findings. Patient reports no
worsening of hip pain during this period. Harris hip score is 91 points.
|
|
Iida et al. [4] reported
that initial abnormal findings on radiography were slight in 13 hips with bone
marrow edema, but 11 (85%) of 13 hips subsequently progressed to advanced
osteonecrosis. They concluded that bone marrow edema should be considered a
marker for potential progression to advanced osteonecrosis. These studies
coincide with our study. The final radiographic stage of the 28 hips that
showed bone marrow edema was significantly advanced compared with those
without bone marrow edema. Bone marrow edema might represent an inflammatory
change in the reactive process during the progression of collapse of the
femoral head and might be an unfavorable sign for the prognosis.
On the other hand, Koo et al.
[3] reported no significant
difference between the collapse rate in hips associated with edema (10/12) and
the remaining hips not associated with edema (19/25). They also reported that
bone marrow edema of the proximal femur was strongly associated with pain;
however, several cases did not show collapse after edema for a considerable
period, and pain was alleviated with the resolution of bone marrow edema
despite radiographic evidence of collapse. Ficat and Arlet
[22] reported that bone marrow
edema can occur from intramedullary venous stasis without collapse in patients
with osteonecrosis, although the reason for this difference is not clear.
One possible reason is the different treatments indicated for osteonecrosis
of the femoral head: We did not perform core decompression, preferring femoral
osteotomy. We believe that the most important concept in the treatment of
osteonecrosis of the femoral head is to remove the necrotic lesions from the
weight-bearing portions of the hip joint
[8]. The lesions of the
weight-bearing portions should be replaced with normal articular cartilage and
subchondral bone by femoral osteotomy. Although core decompression is
reportedly an effective procedure
[22], studies have questioned
its effectiveness because of its high failure rate
[23,
24].
Subchondral focal lesions are likely to be overlooked because bone marrow
edema develops around the surrounding living bone and may obscure subchondral
focal abnormalities on MR images (Fig.
2B). Mirowitz et al.
[25] reported that bone marrow
abnormalities can be accurately evaluated on fat-suppressed T2-weighted and
STIR images. Iida et al. [4]
reported that the high signal intensity of normal fat is decreased on STIR
images, and small focal lesions of osteonecrosis associated with bone marrow
edema can easily be detected. It is often difficult to detect focal
abnormalities with bone marrow edema on T2-weighted images; STIR images are
more useful for evaluating bone marrow edema with osteonecrosis. We agree with
these comments that fat-suppressed T2-weighted or STIR images should be
evaluated for bone marrow edema. We now routinely obtain STIR images for
evaluation, although we did not do so during this study, for which we
evaluated bone abnormalities only on T1- and T2-weighted images at 6- to
12-month intervals, which was one weakness of this study. It is possible that
we underestimated the incidence of bone marrow edema.
Use of a body coil instead of a local coil was another potential limitation
of our study. The resolution achievable with adequate signal-to-noise ratio
might not have been enough to identify accurately the volume of involvement or
the presence of subtle focal subchondral lesions or collapse. Other
limitations were that the observers were not blinded for clinical and
radiographic assessments, and intra- or interobserver variability studies were
not done.
The presence or absence of bone marrow edema on initial diagnostic MR
images, not just the development of bone marrow edema during follow-up, is
important. Ideally, prognostication should be based on the earliest available
data. If the necrotic volume measured on the initial MR images can almost
predict the outcome and the development of bone marrow edema 1 or 2 years
later, then the volume indicator is better clinically because it gives an
earlier indication and is cheaper. Measuring the necrotic volume requires only
one MRI sequence, not sequential sequences. However, if at the time of onset
the patient's hip pain is not severe (Figs.
1A,
1B,
1C,
1D and
1E and
2A,
2B,
2C,
2D and
2E), we can prognosticate
worsening of hip pain by the presence of bone marrow edema on simultaneous MR
images. If bone marrow edema is not present at that time, the hip pain may
possibly not progressively deteriorate, at least for a while (Figs.
6A,
6B,
6C and
6D). Follow-up with MRI at
6-month intervals would be recommended for such patients to detect the
occurrence of bone marrow edema, which is the most significant risk factor for
worsening of hip pain. If the patient refuses follow-up MRI because of its
cost, the presence or absence of bone marrow edema and the necrotic volume
measured on the initial diagnostic MR images may be the second useful
indicators for prognostication. A necrotic volume of 30% or more indicates a
high likelihood of future worsening of hip pain.
In conclusion, bone marrow edema strongly correlates with necrotic volume
and worsening of hip pain in patients with osteonecrosis of the femoral head.
We believe that bone marrow edema represents the occurrence of collapse of the
femoral head and is an unfavorable sign in disease prognosis. Assessment of
the presence or absence of bone marrow edema may be most useful to foretell
worsening of hip pain for patients at the time of onset of hip pain. A large
necrotic volume of 30% or more measured on initial diagnostic MR images may be
the second useful indicator to predict future worsening of hip pain.
References
- Sugano N, Ohzono K, Masuhara K, Takaoka K, Ono K. Prognostication
of osteonecrosis of the femoral head in patients with systemic lupus
erythematosus by magnetic resonance imaging. Clin
Orthop 1994; 305:190
-199
- Kubo T, Yamazoe S, Sugano N, et al. Initial MRI findings of
non-traumatic osteonecrosis of the femoral head in renal allograft recipients.
Magn Reson Imaging 1997;15
: 1017-1023[CrossRef][Medline]
- Koo KH, Ahn IO, Kim R, et al. Bone marrow edema and associated pain
in early stage osteonecrosis of the femoral head: prospective study with
serial MR images. Radiology 1999;213
: 715-722[Abstract/Free Full Text]
- Iida S, Harada Y, Shimizu K, et al. Correlation between bone marrow
edema and collapse of the femoral head in steroid-induced osteonecrosis.
AJR 2000; 174:735
-743[Abstract/Free Full Text]
- Nishii T, Sugano N, Ohzono K, Sakai T, Sato Y, Yoshikawa H.
Significance of lesion size and location in the prediction of collapse of
osteonecrosis of the femoral head: a new three-dimensional quantification
using magnetic resonance imaging. J Orthop Res2002; 20:130
-136[CrossRef][Medline]
- Ito H, Matsuno T, Omizu N, Aoki Y, Minami A. Mid-term prognosis of
non-traumatic osteonecrosis of the femoral head. J Bone Joint Surg
Br 2003; 85:796
-801
- Radke S, Kirschner S, Seipel V, Rader C, Eulert J. Magnetic
resonance imaging criteria of successful core decompression in avascular
necrosis of the hip. Skeletal Radiol2004; 33:519
-523[Medline]
- Merle D'Aubign R, Postel M, Mazabraud A, Massias P, Gueguen J.
Idiopathic necrosis of the femoral head in adults. J Bone Joint
Surg Br 1965; 47:612
-633
- Steinberg ME, Hayken GD, Steinberg DR. A quantitative system for
staging avascular necrosis. J Bone Joint Surg Br1995; 77:34
-41
- Steinberg ME, Bands RE, Parry S, Hoffman E, Chan T, Hartman KM.
Does lesion size affect the outcome in avascular necrosis? Clin
Orthop 1999; 367:262
-271
- Mitchell DG, Rao VM, Dalinka MK, et al. Femoral head avascular
necrosis: correlation of MR imaging, radiographic staging, radionuclide
imaging, and clinical findings. Radiology1987; 162:709
-715[Abstract/Free Full Text]
- Turner DA, Templeton AC, Selzer PM, Rosenberg AG, Petasnick JP.
Femoral capital osteonecrosis: MR finding of diffuse marrow abnormalities
without focal lesions. Radiology 1989;171
: 135-140[Abstract/Free Full Text]
- Li KC, Hiette P. Contrast-enhanced fat saturation magnetic
resonance imaging for studying the pathophysiology of osteonecrosis of the
hips. Skeletal Radiol 1992;21
: 375-379[Medline]
- Vande Berg BC, Malghem JJ, Lecourvet FE, Jamart J, Maldague BE.
Idiopathic bone marrow edema lesions of the femoral head: predictive value of
MR imaging findings. Radiology 1999;212
: 527-535[Abstract/Free Full Text]
- Kim YM, Oh HC, Kim HJ. The pattern of bone marrow oedema on MRI in
osteonecrosis of the femoral head. J Bone Joint Surg
Br 2000; 82:837
-841
- Sakai T, Sugano N, Nishii T, Haraguchi K, Ochi T, Ohzono K. MR
findings of necrotic lesions and the extralesional area of osteonecrosis of
the femoral head. Skeletal Radiol 2000;29
: 133-141[CrossRef][Medline]
- Fujioka M, Kubo T, Nakamura F, et al. Initial changes of
non-traumatic osteonecrosis of femoral head in fat suppression images: bone
marrow edema was not found before the appearance of band patterns.
Magn Reson Imaging 2001;19
: 985-991[CrossRef][Medline]
- Huang GS, Chan WP, Chang YC, Chang CY, Chen CY, Yu JS. MR imaging
of bone marrow edema and joint effusion in patients with osteonecrosis of the
femoral head: relationship to pain. AJR2003; 181:545
-549[Abstract/Free Full Text]
- Totty WG, Murphy WA, Ganz WI, Kumar B, Daum WJ, Siegel BA. Magnetic
resonance imaging of the normal and ischemic femoral head.
AJR 1984; 143:1273
-1280[Abstract/Free Full Text]
- Harris WH. Traumatic arthritis of the hip after dislocation and
acetabular fracture: treatment by mold arthroplasty. J Bone Joint
Surg Am 1969; 51:737
-755[Abstract/Free Full Text]
- Motomura G, Yamamoto T, Miyanishi K, Shirasawa K, Noguchi Y,
Iwamoto Y. Subchondral insufficiency fracture of the femoral head and
acetabulum: a case report. J Bone Joint Surg Am2002; 84:1205
-1209[Free Full Text]
- Ficat RP, Arlet J. Necrosis of the femoral head. In: Hungerford DS,
ed. Ischemia and necrosis of bone. Baltimore, MD:
Williams & Wilkins, 1980:53
-74
- Koo KH, Kim R, Ko GH, Song HR, Jeong ST, Cho SH. Preventing
collapse in early osteonecrosis of the femoral head: a randomised clinical
trial of core decompression. J Bone Joint Surg Br1995; 77:870
-874
- Learmonth ID, Maloon S, Dall G. Core decompression for early
atraumatic osteonecrosis of the femoral head. J Bone Joint Surg
Br 1990; 72:387
-390
- Mirowitz SA, Apicella P, Reinus WR, Hammerman AM. MR imaging of
bone marrow lesions: relative conspicuousness on T1-weighted, fat-suppressed
T2-weighted, and STIR images. AJR 1994;162
: 215-221[Abstract/Free Full Text]
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
|
|
 
A. V. Korompilias, A. H. Karantanas, M. G. Lykissas, and A. E. Beris
Transient Osteoporosis
J. Am. Acad. Ortho. Surg.,
August 1, 2008;
16(8):
480 - 489.
[Abstract]
[Full Text]
[PDF]
|
|
|
Top
Abstract
Introduction
