1 From the Department of Radiology, Thomas Jefferson University Hospital, Philadelphia, Pa (H.P.L., M.E.S., W.B.M.); and Department of Radiology, University Hospital Basel, Petersgraben 4, 4031 Basel, Switzerland (H.P.L.). Received March 5, 2001; revision requested April 2; revision received May 10; accepted June 5.
PURPOSE: To evaluate the magnetic resonance (MR) signal intensity
characteristics of pelvic heterotopic ossification (HO) in various
stages of maturation.
MATERIALS AND METHODS: Thirty-six patients with HO proved at computed tomography (CT) (n = 17) or radiography (n = 19) who underwent 1.5-T pelvic MR imaging within 3 months were included. HO was defined at CT or radiography as grade 1, fluid attenuation without calcification at CT; grade 2, calcification; grade 3, immature ossification; or grade 4, mature ossification. The location and MR signal intensity of all HO sites were noted.
RESULTS: HO was determined to be grade 1 at 20 of 141 sites, grade 2 at 39, grade 3 at 30, and grade 4 at 52. With increasing HO grade, the following findings were observed: (a) decreasing T2 signal intensity (grade 1, 70%; grade 2, 58%; grade 3, 44%; grade 4, 4%), (b) increasing fat and cortical bone signal intensity at T1-weighted imaging (grade 1, 0%; grade 2, 3%; grade 3, 13%; grade 4, 86%), and (c) decreasing contrast enhancement (from 100% for grade 1 to 20% for grade 4). Fifteen (88%) patients with CT correlation had HO in the anatomic area of the trochanteric or iliopsoas bursa (55 [60%] of 91 sites).
CONCLUSION: With progressive maturity of HO, T2 signal intensity and contrast enhancement decrease, but fat and cortical bone–equivalent signal intensity increases.
Index terms: Bones, abnormalities, 44.1485, 44.82 • Bones, CT, 44.12111, 44.12112, 44.12115 • Bones, MR, 44.121411, 44.121412, 44.121413, 44.121415, 44.121416 • Bones, radiography, 44.11 • Paralysis, 44.82
Heterotopic ossification is common in paralyzed patients and
most frequently involves the hips (
1). In the setting of pelvic
ulceration, it can substantially complicate accurate radiographic
(
2) and scintigraphic (
3) evaluation of soft-tissue infection
and osteomyelitis. Magnetic resonance (MR) imaging of the pelvis
is being used increasingly to evaluate the presence and extent
of infection beneath the pressure sores in these patients. The
MR signal intensity characteristics of heterotopic bone were
described in two previously published reports but included only
mature heterotopic ossification with fat signal intensity (
4)
or decreased signal intensity at all sequences (
5) in these
two studies.
Immature heterotopic bone formation in cases of myositis ossificans of the extremities has been reported to have MR signal intensity characteristics that are equivalent to those of local infection with abscess formation (6–9). Thus, similar signal intensity characteristics of immature pelvic heterotopic bone can considerably complicate MR imaging analysis in paralyzed patients with pressure ulcers. The goal of this study was to evaluate the MR signal intensity characteristics of heterotopic ossification in various stages of maturation.
Our study was conducted after approval to review patient images
and medical charts was obtained from the institutional review
board of Thomas Jefferson University Hospital. The institutional
review board determined that this retrospective study could
be conducted without acquisition of formal signed informed consents
from the patients.
Between January 1995 and June 2000, 193 MR imaging examinations of the pelvis and hips were performed in 157 patients at Thomas Jefferson University Hospital to investigate for potential osteomyelitis in paralyzed or bedridden patients. With 105 of the 193 MR examinations, either computed tomography (CT) or conventional radiography of the pelvis and hips was performed within 3 months of the MR study. Review of these 105 CT and radiographic studies by two musculoskeletal radiologists (M.E.S., W.B.M.) in consensus revealed heterotopic ossification in 42 cases (40%). The mean time (± SD) between the MR examination and the corresponding CT (n = 17) or conventional radiographic (n = 19) study was 27.56 days ± 21.78.
We excluded from the study all foci of heterotopic bone formation that showed evidence of adjacent skin ulceration and extension of enhancing soft tissue from the ulcer base to the area of heterotopic bone to avoid mistaking these for signal intensity alterations due to infection. This resulted in the exclusion of six MR studies and 10 foci of heterotopic bone formation seen at nine other MR examinations. After MR imaging, 11 of the excluded cases had local surgical debridement of the infected ulcer, and in five cases, amputation (ie, girdlestone procedure [n = 3], ischiectomy [n = 1], and hemipelvectomy [n = 1]) was performed because of underlying osteomyelitis. Review of the patient charts revealed recent surgery (<1 year) in two areas of heterotopic bone formation, which also were excluded from analysis to prevent the misinterpretation of postoperative changes as heterotopic bone formation.
The final study group included 36 patients (three women and 33 men) aged 20–75 years (mean age, 43.4 years). For a comparative study of MR imaging findings, 19 patients underwent radiography and 17 underwent CT. Thirty-one patients had a history of traumatic spinal cord injury, two patients had spinal cord infarctions, and the remaining three patients were bedridden because of multiple sclerosis, ischemic brain damage, or spina bifida. There were 24 paraplegic patients and 12 tetraplegic patients. The mean duration (± SD) between onset of paralysis and MR imaging was 10.6 years ± 8.93 (range, 3 months to 38 years).
The two musculoskeletal radiologists jointly compared the MR images with the corresponding CT or radiographic images. First, the maturity of heterotopic bone was graded at CT or radiography, according to categories derived from prior reports (10,11), as follows: grade 1, fluid attenuation without evidence of calcification at CT (Only the images obtained in patients who underwent CT could be classified as grade 1.); grade 2, calcification of soft tissues without evidence of bone formation; grade 3, immature bone formation; and grade 4, mature bone with cortical differentiation. Then, the anatomic location of the heterotopic bone was noted. In the patients who underwent CT for comparison, the reviewers determined whether the heterotopic bone formation was located in the anatomic area of a bursa (especially the trochanteric and iliopsoas bursae). The size of each area of heterotopic bone formation was measured on transverse MR or CT images. Last, the signal intensity characteristics on nonenhanced T1- and T2-weighted MR images and on contrast material–enhanced T1-weighted MR images were noted.
On the nonenhanced T1-weighted MR images, the signal intensity of heterotopic bone was defined as one of the following: hyperintense to muscle, isointense to muscle, hypointense to muscle, and signal intensity equivalent to that of fat and cortical bone. On the nonenhanced T2-weighted images, the following signal intensity categories were applied: signal intensity equivalent to that of fluid, hyperintense to muscle, isointense to muscle, hypointense to muscle, and signal intensity equivalent to that of fat and cortical bone. On the contrast-enhanced T1-weighted MR images, the following signal intensity categories were applied: rim enhancement, enhancement, and no enhancement. The sites of heterotopic ossification that had areas of distinctly different maturity on CT or radiographic images were analyzed separately for size and MR signal intensity characteristics.
MR imaging was performed with a 1.5-T unit (Signa; GE Medical Systems, Milwaukee, Wis) and either a body coil (n = 7) or a phased-array multicoil (n = 29). At each study, at least two orthogonal planes were imaged. A 256 x 256 matrix was typically used with a section thickness of 5–8 mm, intersection gap of 1 mm, and field of view of 24–40 cm. T1-weighted spin-echo MR images were obtained with two acquired signals and 400–600/10–20 (repetition time msec/echo time msec). In 32 cases, T2-weighted MR images were obtained by using a fast spin-echo technique with an echo train length of eight, two signals acquired, 4,000–6,000/70–100 (effective), and fat suppression. In six cases, T2-weighted MR images were obtained by using a conventional spin-echo technique with 2,000–3,000/60–85. In 29 studies, fat-suppressed T2-weighted MR images were obtained by using a fast spin-echo short inversion time inversion-recovery technique with 5,000–7,000/40–60 (effective) and an inversion time of 150 msec. T1-weighted fat-saturated MR images were obtained following the intravenous administration of 0.1 mmol/kg of gadopentetate dimeglumine (Magnevist; Berlex Laboratories, Wayne, NJ) in all but two patients. The fat-saturated images were obtained with either a spin-echo (with parameters identical to those used for T1-weighted imaging [n = 5]) or a fast multiplanar spoiled gradient-echo (with 150–340/2.0–3.3 and 90° flip angle [n = 29]) technique.
CT of the pelvis was performed in 17 patients with a spiral CT unit (HighSpeed Advantage; GE Medical Systems). The section thickness ranged from 3 to 7 mm, and the protocol included the use of 120 kV and 200–280 mAs. The contrast agent iothalamate meglumine (Conray; Mallinckrodt, St Louis, Mo) was administered intravenously in 14 patients. Conventional radiographs were obtained for comparison in 19 patients and included an anteroposterior projection of the pelvis plus one additional view of the affected hip (pelvis projection plus additional view, n = 12), two views of a hip (n = 5), and scout images (at least two views) of specimens obtained at fluoroscopically guided bone biopsy (n = 2).
The 2 test for trend was performed to statistically establish potential trends in MR signal intensity alterations with increasing maturity of heterotopic ossification. A trend of MR signal intensity alterations was considered to be statistically significant when a P value of .05 or lower was determined.
At analysis of CT and radiographic images, we identified 101
areas of heterotopic ossification. Thirty-seven of these 101
areas contained heterotopic bone of distinctly different maturity
within one area. Thirty-four (92%) of these 37 areas had heterotopic
bone formation that corresponded to two different grades of
maturity, and three areas had heterotopic bone that corresponded
to three different maturity grades. These areas yielded a total
of 141 distinct sites of heterotopic ossification. At grading
of these sites, there were 20 sites with grade 1 heterotopic
bone maturity, 39 with grade 2, 30 with grade 3, and 52 with
grade 4.
The mean size (± SD) of the evaluated sites at MR imaging was 3.87 cm ± 1.98 x 2.48 cm ± 1.25. The relative frequencies of signal intensity alterations on T1-weighted images obtained before and after gadolinium-based contrast material administration and on nonenhanced T2-weighted images are summarized in the Table.
| fig.ommitted |
MR Signal Intensity and Contrast Enhancement Characteristics of 141 Heterotopic Ossification Sites
| |
On nonenhanced T1-weighted images, with increasing maturity
of heterotopic ossification, the frequency of low signal intensity
relative to muscle steadily decreased and the percentage of
heterotopic ossification sites with signal isointense relative
to muscle increased (with the exception of grade 4 sites). With
increasing maturity of heterotopic ossification, an increasing
proportion of heterotopic ossification areas had signal intensity
that was equivalent to that of cortical bone and fat (
Fig 1).
Eighty-six percent of areas that were graded as mature heterotopic
ossification (grade 4) had signal intensity that was equivalent
to fat and cortex on nonenhanced T1-weighted images.
2 testing
for trend resulted in the following values for the observed
MR signal intensity alterations on T1-weighted images: hyperintense
to muscle,
P value of .29
; hypointense to muscle,
P value less
than .001; isointense to muscle,
P value less than .001; and
signal intensity equivalent to that of fat and cortical bone,
P value less than .001.
| fig.ommitted |
Figure 1a. Extensive heterotopic ossification in a 28-year-old paraplegic man who had a gunshot injury of the thoracic spinal cord 7 years ago. (a) Anteroposterior radiograph of the left hip shows extensive mature heterotopic ossification in the left thigh and around the left acetabulum. Note the region of immature (grade 3) heterotopic ossification (arrow) in the upper lateral area. (b) Transverse CT image of immature (grade 3) rim-shaped heterotopic bone formation (arrows) lateral to the left acetabulum. (c) Coronal nonenhanced T1-weighted MR image (400/10) of the pelvis. The marrow signal intensity of mature heterotopic bone (arrowheads) corresponds to fatty marrow with cortical inclusions. The signal of immature heterotopic bone (arrows), however, is isointense to muscle. (d) Coronal nonenhanced T2-weighted fast spin-echo short inversion time inversion-recovery MR image (6,067/45; inversion time, 150 msec) shows a hyperintense area (arrow) in the region of immature bone. Note the low signal intensity (arrowheads) corresponding to suppressed fat in mature heterotopic ossification. (e) Coronal T1-weighted contrast-enhanced fast multiplanar spoiled gradient-echo MR image (340.0/2.5, 90° flip angle) shows rim enhancement (arrow) and diffuse enhancement more medially in the area of immature heterotopic bone. Note the nonenhancing mature heterotopic bone (arrowheads).
| |
| fig.ommitted |
Figure 1b. Extensive heterotopic ossification in a 28-year-old paraplegic man who had a gunshot injury of the thoracic spinal cord 7 years ago. (a) Anteroposterior radiograph of the left hip shows extensive mature heterotopic ossification in the left thigh and around the left acetabulum. Note the region of immature (grade 3) heterotopic ossification (arrow) in the upper lateral area. (b) Transverse CT image of immature (grade 3) rim-shaped heterotopic bone formation (arrows) lateral to the left acetabulum. (c) Coronal nonenhanced T1-weighted MR image (400/10) of the pelvis. The marrow signal intensity of mature heterotopic bone (arrowheads) corresponds to fatty marrow with cortical inclusions. The signal of immature heterotopic bone (arrows), however, is isointense to muscle. (d) Coronal nonenhanced T2-weighted fast spin-echo short inversion time inversion-recovery MR image (6,067/45; inversion time, 150 msec) shows a hyperintense area (arrow) in the region of immature bone. Note the low signal intensity (arrowheads) corresponding to suppressed fat in mature heterotopic ossification. (e) Coronal T1-weighted contrast-enhanced fast multiplanar spoiled gradient-echo MR image (340.0/2.5, 90° flip angle) shows rim enhancement (arrow) and diffuse enhancement more medially in the area of immature heterotopic bone. Note the nonenhancing mature heterotopic bone (arrowheads).
| |
| fig.ommitted |
Figure 1c. Extensive heterotopic ossification in a 28-year-old paraplegic man who had a gunshot injury of the thoracic spinal cord 7 years ago. (a) Anteroposterior radiograph of the left hip shows extensive mature heterotopic ossification in the left thigh and around the left acetabulum. Note the region of immature (grade 3) heterotopic ossification (arrow) in the upper lateral area. (b) Transverse CT image of immature (grade 3) rim-shaped heterotopic bone formation (arrows) lateral to the left acetabulum. (c) Coronal nonenhanced T1-weighted MR image (400/10) of the pelvis. The marrow signal intensity of mature heterotopic bone (arrowheads) corresponds to fatty marrow with cortical inclusions. The signal of immature heterotopic bone (arrows), however, is isointense to muscle. (d) Coronal nonenhanced T2-weighted fast spin-echo short inversion time inversion-recovery MR image (6,067/45; inversion time, 150 msec) shows a hyperintense area (arrow) in the region of immature bone. Note the low signal intensity (arrowheads) corresponding to suppressed fat in mature heterotopic ossification. (e) Coronal T1-weighted contrast-enhanced fast multiplanar spoiled gradient-echo MR image (340.0/2.5, 90° flip angle) shows rim enhancement (arrow) and diffuse enhancement more medially in the area of immature heterotopic bone. Note the nonenhancing mature heterotopic bone (arrowheads).
| |
| fig.ommitted |
Figure 1d. Extensive heterotopic ossification in a 28-year-old paraplegic man who had a gunshot injury of the thoracic spinal cord 7 years ago. (a) Anteroposterior radiograph of the left hip shows extensive mature heterotopic ossification in the left thigh and around the left acetabulum. Note the region of immature (grade 3) heterotopic ossification (arrow) in the upper lateral area. (b) Transverse CT image of immature (grade 3) rim-shaped heterotopic bone formation (arrows) lateral to the left acetabulum. (c) Coronal nonenhanced T1-weighted MR image (400/10) of the pelvis. The marrow signal intensity of mature heterotopic bone (arrowheads) corresponds to fatty marrow with cortical inclusions. The signal of immature heterotopic bone (arrows), however, is isointense to muscle. (d) Coronal nonenhanced T2-weighted fast spin-echo short inversion time inversion-recovery MR image (6,067/45; inversion time, 150 msec) shows a hyperintense area (arrow) in the region of immature bone. Note the low signal intensity (arrowheads) corresponding to suppressed fat in mature heterotopic ossification. (e) Coronal T1-weighted contrast-enhanced fast multiplanar spoiled gradient-echo MR image (340.0/2.5, 90° flip angle) shows rim enhancement (arrow) and diffuse enhancement more medially in the area of immature heterotopic bone. Note the nonenhancing mature heterotopic bone (arrowheads).
| |
| fig.ommitted |
Figure 1e. Extensive heterotopic ossification in a 28-year-old paraplegic man who had a gunshot injury of the thoracic spinal cord 7 years ago. (a) Anteroposterior radiograph of the left hip shows extensive mature heterotopic ossification in the left thigh and around the left acetabulum. Note the region of immature (grade 3) heterotopic ossification (arrow) in the upper lateral area. (b) Transverse CT image of immature (grade 3) rim-shaped heterotopic bone formation (arrows) lateral to the left acetabulum. (c) Coronal nonenhanced T1-weighted MR image (400/10) of the pelvis. The marrow signal intensity of mature heterotopic bone (arrowheads) corresponds to fatty marrow with cortical inclusions. The signal of immature heterotopic bone (arrows), however, is isointense to muscle. (d) Coronal nonenhanced T2-weighted fast spin-echo short inversion time inversion-recovery MR image (6,067/45; inversion time, 150 msec) shows a hyperintense area (arrow) in the region of immature bone. Note the low signal intensity (arrowheads) corresponding to suppressed fat in mature heterotopic ossification. (e) Coronal T1-weighted contrast-enhanced fast multiplanar spoiled gradient-echo MR image (340.0/2.5, 90° flip angle) shows rim enhancement (arrow) and diffuse enhancement more medially in the area of immature heterotopic bone. Note the nonenhancing mature heterotopic bone (arrowheads).
| |
At nonenhanced T2-weighted sequences (
Table), a similar tendency
was recognized: Fluid-equivalent signal intensity steadily decreased
in frequency with increasing maturity of heterotopic ossification
(
Figs 1,
2). With increasing maturity, a relative increase in
isointense and hyperintense signal (grades 1–3) was observed.
With mature heterotopic ossification, the majority of sites
had signal intensity that was equivalent to that of fat and
cortical bone. Only 3% of grade 2 areas and 13% and 7% of grade
3 areas on T1- and T2-weighted MR images, respectively, had
signal intensity characteristics that were equivalent to those
of fat and cortical bone; however, calcification or ossification
was seen in all of these cases at CT or radiography.
2 testing
for trend resulted in the following values for the observed
MR signal intensity alterations on T2-weighted images: fluid-equivalent
signal intensity,
P value less than .001; hyperintense to muscle,
P value less than .001; isointense to muscle,
P value of .002;
hypointense to muscle,
P value of .52; and signal intensity
equivalent to that of fat and cortical bone,
P value less than
.001.
| fig.ommitted |
Figure 2a. Heterotopic ossification in the right upper thigh of a 23-year-old paraplegic man with a history of multiple gunshot wounds resulting in thoracic spinal cord injury 6 years ago. (a) "Frog-leg" lateral radiograph of the right hip shows multiple foci of immature heterotopic ossification (arrowheads) (grades 2 and 3) in the right upper thigh and around the lesser trochanter. Note the bullet (arrow) lateral to the right acetabulum. (b) Coronal nonenhanced T1-weighted MR image (400/11) of the right hip shows a large mass in the region of the immature heterotopic ossification, which is isointense to muscle and contains hypointense inclusions (arrowheads) corresponding to immature bone. A hypointense collection in the inferior aspect of the mass (short arrow) and an isointense mass effect (long arrows) in the right trochanteric bursa also are seen. (c) Coronal nonenhanced T2-weighted fast spin-echo short inversion time inversion-recovery MR image (6,000/45; inversion time, 150 msec) of the right hip shows a hyperintense collection (short arrow) and rim-shaped high signal intensity (arrowheads) around immature heterotopic bone. Note the fluid collection (long arrows) in the right trochanteric bursa. (d) Coronal T1-weighted contrast-enhanced fast multiplanar spoiled gradient-echo MR image (200/2, 90° flip angle) of the right hip shows rim enhancement around immature heterotopic bone (arrowheads) and rim enhancement (short arrow) around the collection inferiorly. Also note the rim enhancement of the right trochanteric bursa (long arrows).
| |
| fig.ommitted |
Figure 2b. Heterotopic ossification in the right upper thigh of a 23-year-old paraplegic man with a history of multiple gunshot wounds resulting in thoracic spinal cord injury 6 years ago. (a) "Frog-leg" lateral radiograph of the right hip shows multiple foci of immature heterotopic ossification (arrowheads) (grades 2 and 3) in the right upper thigh and around the lesser trochanter. Note the bullet (arrow) lateral to the right acetabulum. (b) Coronal nonenhanced T1-weighted MR image (400/11) of the right hip shows a large mass in the region of the immature heterotopic ossification, which is isointense to muscle and contains hypointense inclusions (arrowheads) corresponding to immature bone. A hypointense collection in the inferior aspect of the mass (short arrow) and an isointense mass effect (long arrows) in the right trochanteric bursa also are seen. (c) Coronal nonenhanced T2-weighted fast spin-echo short inversion time inversion-recovery MR image (6,000/45; inversion time, 150 msec) of the right hip shows a hyperintense collection (short arrow) and rim-shaped high signal intensity (arrowheads) around immature heterotopic bone. Note the fluid collection (long arrows) in the right trochanteric bursa. (d) Coronal T1-weighted contrast-enhanced fast multiplanar spoiled gradient-echo MR image (200/2, 90° flip angle) of the right hip shows rim enhancement around immature heterotopic bone (arrowheads) and rim enhancement (short arrow) around the collection inferiorly. Also note the rim enhancement of the right trochanteric bursa (long arrows).
| |
| fig.ommitted |
Figure 2c. Heterotopic ossification in the right upper thigh of a 23-year-old paraplegic man with a history of multiple gunshot wounds resulting in thoracic spinal cord injury 6 years ago. (a) "Frog-leg" lateral radiograph of the right hip shows multiple foci of immature heterotopic ossification (arrowheads) (grades 2 and 3) in the right upper thigh and around the lesser trochanter. Note the bullet (arrow) lateral to the right acetabulum. (b) Coronal nonenhanced T1-weighted MR image (400/11) of the right hip shows a large mass in the region of the immature heterotopic ossification, which is isointense to muscle and contains hypointense inclusions (arrowheads) corresponding to immature bone. A hypointense collection in the inferior aspect of the mass (short arrow) and an isointense mass effect (long arrows) in the right trochanteric bursa also are seen. (c) Coronal nonenhanced T2-weighted fast spin-echo short inversion time inversion-recovery MR image (6,000/45; inversion time, 150 msec) of the right hip shows a hyperintense collection (short arrow) and rim-shaped high signal intensity (arrowheads) around immature heterotopic bone. Note the fluid collection (long arrows) in the right trochanteric bursa. (d) Coronal T1-weighted contrast-enhanced fast multiplanar spoiled gradient-echo MR image (200/2, 90° flip angle) of the right hip shows rim enhancement around immature heterotopic bone (arrowheads) and rim enhancement (short arrow) around the collection inferiorly. Also note the rim enhancement of the right trochanteric bursa (long arrows).
| |
| fig.ommitted |
Figure 2d. Heterotopic ossification in the right upper thigh of a 23-year-old paraplegic man with a history of multiple gunshot wounds resulting in thoracic spinal cord injury 6 years ago. (a) "Frog-leg" lateral radiograph of the right hip shows multiple foci of immature heterotopic ossification (arrowheads) (grades 2 and 3) in the right upper thigh and around the lesser trochanter. Note the bullet (arrow) lateral to the right acetabulum. (b) Coronal nonenhanced T1-weighted MR image (400/11) of the right hip shows a large mass in the region of the immature heterotopic ossification, which is isointense to muscle and contains hypointense inclusions (arrowheads) corresponding to immature bone. A hypointense collection in the inferior aspect of the mass (short arrow) and an isointense mass effect (long arrows) in the right trochanteric bursa also are seen. (c) Coronal nonenhanced T2-weighted fast spin-echo short inversion time inversion-recovery MR image (6,000/45; inversion time, 150 msec) of the right hip shows a hyperintense collection (short arrow) and rim-shaped high signal intensity (arrowheads) around immature heterotopic bone. Note the fluid collection (long arrows) in the right trochanteric bursa. (d) Coronal T1-weighted contrast-enhanced fast multiplanar spoiled gradient-echo MR image (200/2, 90° flip angle) of the right hip shows rim enhancement around immature heterotopic bone (arrowheads) and rim enhancement (short arrow) around the collection inferiorly. Also note the rim enhancement of the right trochanteric bursa (long arrows).
| |
After administration of gadolinium-based contrast material (
Table),
88% of the grade 1 sites had strong ring enhancement. Sixty-seven
percent of sites with calcification (grade 2) (
Figs 2,
3) and
66% with immature ossification (grade 3) (
Fig 1) still had ring
enhancement, whereas the percentage of nonenhancing areas was
only 10% at grade 3. Although most areas with grade 4 maturity
did not have enhancement (80%), 14% of these sites had ring
enhancement. Diffuse enhancement increased in frequency from
grade 1 (12%) to grade 2 (30%) maturity, but it decreased in
frequency from grade 3 (24%) to grade 4 (6%).
2 testing for
trend resulted in the following values for the observed MR signal
intensity alterations on contrast-enhanced T1-weighted MR images:
rim enhancement,
P value less than .001; enhancement,
P value
of .016; and no enhancement,
P value less than .001.
| fig.ommitted |
Figure 3a. Bilateral immature (grade 2) heterotopic ossification ventral to the femoral neck in a 36-year-old tetraplegic man who had a gunshot injury of the cervical spinal cord 18 months prior to admission. (a) Transverse CT image of the pelvis shows rimlike arrangement of fine calcifications (arrowheads) ventral to both femoral necks. Note the broad-based calcification (arrow) adjacent to the left femoral neck; this calcification corresponds to the margins of the iliopsoas bursa. (b) Transverse nonenhanced T1-weighted spin-echo MR image (450/11) shows an isointense soft-tissue mass anterior to the femoral necks bilaterally but predominantly on the left (white arrow). Compared with the calcifications depicted in a, the low-signal-intensity foci (black arrows) representing fine calcifications on this image are less apparent. (c) Transverse contrast-enhanced fast multiplanar spoiled gradient-echo MR image (200.0/2.5, 90° flip angle) shows hypointense collections in the region of the iliopsoas bursa bilaterally, with rim enhancement (arrowheads).
| |
| fig.ommitted |
Figure 3b. Bilateral immature (grade 2) heterotopic ossification ventral to the femoral neck in a 36-year-old tetraplegic man who had a gunshot injury of the cervical spinal cord 18 months prior to admission. (a) Transverse CT image of the pelvis shows rimlike arrangement of fine calcifications (arrowheads) ventral to both femoral necks. Note the broad-based calcification (arrow) adjacent to the left femoral neck; this calcification corresponds to the margins of the iliopsoas bursa. (b) Transverse nonenhanced T1-weighted spin-echo MR image (450/11) shows an isointense soft-tissue mass anterior to the femoral necks bilaterally but predominantly on the left (white arrow). Compared with the calcifications depicted in a, the low-signal-intensity foci (black arrows) representing fine calcifications on this image are less apparent. (c) Transverse contrast-enhanced fast multiplanar spoiled gradient-echo MR image (200.0/2.5, 90° flip angle) shows hypointense collections in the region of the iliopsoas bursa bilaterally, with rim enhancement (arrowheads).
| |
| fig.ommitted |
Figure 3c. Bilateral immature (grade 2) heterotopic ossification ventral to the femoral neck in a 36-year-old tetraplegic man who had a gunshot injury of the cervical spinal cord 18 months prior to admission. (a) Transverse CT image of the pelvis shows rimlike arrangement of fine calcifications (arrowheads) ventral to both femoral necks. Note the broad-based calcification (arrow) adjacent to the left femoral neck; this calcification corresponds to the margins of the iliopsoas bursa. (b) Transverse nonenhanced T1-weighted spin-echo MR image (450/11) shows an isointense soft-tissue mass anterior to the femoral necks bilaterally but predominantly on the left (white arrow). Compared with the calcifications depicted in a, the low-signal-intensity foci (black arrows) representing fine calcifications on this image are less apparent. (c) Transverse contrast-enhanced fast multiplanar spoiled gradient-echo MR image (200.0/2.5, 90° flip angle) shows hypointense collections in the region of the iliopsoas bursa bilaterally, with rim enhancement (arrowheads).
| |
Review of data for all 17 patients who underwent CT (47%) revealed
that 55 (60%) of all 91 sites of heterotopic ossification were
localized in the distribution of the iliopsoas (
Fig 3) or trochanteric
bursa. Eighteen iliopsoas bursae and 20 trochanteric bursae
were involved in 15 patients (88%) and had a mean size (±
SD) at CT of 4.14 cm ± 1.8
x 2.82 cm ± 1.7 in
the transverse plane. Additional heterotopic ossification in
areas of anatomic bursae were seen in the ischium (bursa ischiadica
musculi semimembranosi) in three patients and around the lesser
trochanter in six patients (bursa musculi pectinei).
Heterotopic ossification is one of the most common orthopedic
complications after spinal cord injury and represents bone formation
(
12,
13) in the soft tissues around paralyzed joints. In most
studies, the reported incidence after spinal cord injury is
between 20% and 35% (
1,
14–
16), although values as low
as 13% (
17) and as high as 50% (
18) have been reported. Heterotopic
ossification always occurs below the level of neurologic impairment
and is seen most frequently around the hips (
3,
14,
15), with
involvement of both hips in up to 41% of cases (
17). Complications
of heterotopic ossification include decubitus ulcers (
17,
19),
restricted joint motion, and ankylosis (
11,
20,
21), with subsequent
loss of the patient’s independence (
22).
Evaluation of MR signal intensity characteristics with corresponding radiographic or CT findings enabled us to recognize some features of MR signal intensity changes that occur with increasing heterotopic ossification maturity: On T1-weighted images, the fraction of areas with low signal intensity relative to muscle decreased steadily as the fraction of areas with signal that was isointense to muscle increased. With increasing maturity of heterotopic ossification, the signal intensity of fat and cortical bone also increased. On T2-weighted images, a steady decrease in hyperintense signal was observed. After the administration of gadolinium-based contrast material, most grade 2 and grade 3 areas showed enhancement, which was still observed in 20% of the mature sites (grade 4) of heterotopic ossification and possibly represented residual activity in seemingly mature heterotopic bone.
These observed signal intensity characteristics of heterotopic bone formation are unspecific and can easily be mistaken for infection. Grade 1–3 lesions with hyperintense signal intensity centrally and marked rim enhancement may be confounded with septic bursitis or abscess. Immature heterotopic ossification (grade 3) with diffuse enhancement can be mistaken for soft-tissue infection, and enhancing grade 4 heterotopic bone can be mistaken for osteomyelitis.
Similar MR signal intensity alterations have been described with myositis ossificans, the pathophysiologic features of which may be closely related to heterotopic ossification in experimental studies (23,24). The majority of grade 1 lesions in our study showed rim enhancement and fluid-equivalent signal intensity on contrast-enhanced T1-weighted and nonenhanced T2-weighted MR images, respectively. These findings correspond to reports of early-stage myositis ossificans lesions with rim enhancement (9) and hyperintense signal (6,7,9) on T2-weighted images. Kransdorf et al (6) observed that such areas of high signal intensity in immature myositis ossificans correspond to tissue consisting of fibroblasts and myofibroblasts with myxoid stroma, which has been shown in other investigations (25) to represent the initial stage of heterotopic ossification. On CT images, immature areas of myositis ossificans and heterotopic ossification demonstrate low attenuation compared with muscle (10, 26). Progressive accumulation of calcium density paralleling radiographic evidence of bone formation was demonstrated in pelvic heterotopic ossification at serial CT examinations of such low-attenuating tissue (10).
It is interesting to note that only approximately 10% of the grade 3 heterotopic ossification sites in our study had MR evidence of bone formation, even though this was clearly visible at CT or radiography. Mineralization of immature heterotopic ossification is clearly underestimated at MR imaging, as has been previously noted with myositis ossificans (6,7), particularly when gradient-echo imaging is not performed. In concordance with reports of mature myositis ossificans (7,9), most of the mature regions of heterotopic ossification (grade 4) had MR signal intensity that was equivalent to that of fatty marrow surrounded by cortical bone.
Although metaplasia of mesenchymal cells is the generally accepted cause of heterotopic ossification (25), the tissue of origin still is unknown. Our review of the locations of heterotopic ossification in the patients with CT correlation (n = 17) revealed evidence of calcification or ossification in the region of the trochanteric and iliopsoas bursae in 88% of patients. Also, most grade 1 lesions were observed in the region of the trochanteric and iliopsoas bursae in the typical location of trochanteric and iliopsoas bursitis. As discussed earlier, similar lesions in the same location have been described at CT (10) in paralyzed patients with findings of low attenuation and progressive calcification with increasing maturity. However, the authors of that study did not localize these immature foci of heterotopic ossification to the bursae.
Several reports of intraoperative findings represent evidence of heterotopic ossification formation in anatomic bursae: "Adjacent tendons and muscles, especially the rectus femoris and the iliopsoas, lay in grooves over the ectopic mass, or occasionally, tunneled through the mass" (12). Muscle was observed to be compressed by the underlying heterotopic ossification (12), and the authors stressed that heterotopic ossification may develop without any connection with muscular tissue. In another study (27), the investigators identified heterotopic ossification "in a plane deep to the rectus femoris and superior to the hip capsule" or "deep to the gluteus maximus and posterior to the hip capsule." Approximately 14–21 bursae have been described in the hip region (28–30), and these structures may account for the heterotopic ossification at locations other than regions of the just mentioned bursae. Although our study data support the hypothesis of heterotopic ossification formation in anatomic bursae, further investigations are needed to definitively prove it.
The previously reported time of heterotopic ossification onset after spinal cord injury is not supported by our study data and the results of an earlier CT study (10): There is general agreement among prior studies that the peak incidence of heterotopic ossification after spinal cord injury occurs at 4 to 12 weeks (31) but may be seen at 2–5 months (3,32). The development of heterotopic ossification with chronic spinal cord injury generally is assumed to be considerably less frequent (33). In our study, however, only approximately one-third of all heterotopic ossification sites were graded as mature at CT or radiography, although the mean duration of paralysis in our patients was 10 years. More than one-third of the heterotopic ossification sites were graded as having early maturity (grade 1 and 2) at radiography or CT.
The presence of grade 1 heterotopic ossification in long-standing paralysis has been previously confirmed at CT: Low-attenuating soft tissue was detected adjacent to mature heterotopic bone in six patients with long-standing paralysis for more than 5 years, and it was assumed that this tissue most likely corresponded to immature noncalcified connective tissue, which may have the potential for ossification (10). In earlier descriptions of resected heterotopic ossification (12), the presence of such "edematous and immature connective tissue, which was found next to long-standing heterotopic ossification," also has been noted.
Our study data are also supported by earlier scintigraphic reports, in which it was observed that the development of heterotopic ossification can be very variable: Normalization of radionuclide uptake ratio can occur as early as 3.5 months after injury (34), but immature heterotopic ossification still may be observed 5 years after injury (15). Reactivation of heterotopic ossification with increasing uptake ratios (34,35) and coexistent areas of apparently mature heterotopic ossification with developing new bone (11) also have been described. Although many of the grade 1–3 sites in our population represented immature heterotopic ossification of the classic definition, we speculate that long-standing paralysis leads to pressure- or microtrauma-related bursitis with subsequent calcification and ossification.
Future applications of MR imaging in the setting of heterotopic ossification could include early diagnosis after spinal cord injury and preoperative evaluation of radiographically mature ankylosing heterotopic ossification for residual enhancement and activity. Diagnosis of early heterotopic ossification after spinal cord injury may be facilitated by the identification of the strong enhancement of these lesions and the fluid signal intensity on T2-weighted images.
There were some limitations to our study: First, the interval of up to 3 months between MR imaging and CT or radiography may have in some cases led to the progression or delay of MR signal intensity evolution relative to the CT or radiographic grading. Second, the grading of heterotopic ossification maturity on radiographs may not be as exact (even though all radiographs were obtained in two views) as that on CT images because sites of low mineralization may be obscured by well-mineralized foci. For this reason, only those grade 1 lesions that were low attenuating at CT were included in our study. Third, the formation of heterotopic bone in anatomic bursae was not proved histologically. Last, the simultaneous comparison of MR studies with CT or radiographic images may have introduced some reader bias in the evaluation because the readers may have recognized some typical patterns. It is, however, unrealistic to correlate MR imaging findings with CT or radiographic results in a blinded fashion at analysis of immature heterotopic ossification, because grade 1–3 lesions are very difficult to locate without direct comparison of CT images or radiographs.
We conclude that some features of MR signal intensity evolution exist with increasing maturity of heterotopic bone: With progressive maturity of heterotopic ossification, a relative decrease in T2 hyperintensity and enhancement and an increase in fat and cortical bone–equivalent signal intensity are seen. Areas of immature heterotopic bone lead to a nonspecific pattern of signal intensity alterations with contrast enhancement, which can easily be mistaken for infection. Misinterpretation of heterotopic ossification as infection represents a considerable pitfall in paralyzed patients, because many of them have decubitus ulcers with florid soft-tissue infection at the time of MR imaging. Correct recognition of heterotopic ossification is further impeded by the insensitivity of MR imaging in the detection of matrix calcification in immature heterotopic ossification. Our study data are evidence of the formation of heterotopic ossification in anatomic bursae, especially the iliopsoas and trochanteric bursae.
The authors acknowledge the work of Laurence Parker, PhD, who
performed the statistical analysis.
- Blane CE, Perkash I. True heterotopic bone in the paralyzed patient. Skeletal Radiol 1981; 7:21-25.
- Lewis VL, Jr, Bailey MH, Pulawski G, Kind G, Bashioum RW, Hendrix RW. The diagnosis of osteomyelitis in patients with pressure sores. Plast Reconstr Surg 1988; 81:229-232.
- Orzel JA, Rudd TG. Heterotopic bone formation: clinical, laboratory, and imaging correlation. J Nucl Med 1985; 26:125-132.
- Huang AB, Schweitzer ME, Hume E, Batte WG. Osteomyelitis of the pelvis/hips in paralyzed patients: accuracy and clinical utility of MRI. J Comput Assist Tomogr 1998; 22:437-443.
- Hencey JY, Vermess M, van Geertruyden HH, Binard JE, Manchepalli S. Magnetic resonance imaging examinations of gluteal decubitus ulcers in spinal cord injury patients. J Spinal Cord Med 1996; 19:5-8.
- Kransdorf MJ, Meis JM, Jelinek JS. Myositis ossificans: MR appearance with radiologic-pathologic correlation. AJR Am J Roentgenol 1991; 157:1243-1248.
- De Smet AA, Norris MA, Fisher DR. Magnetic resonance imaging of myositis ossificans: analysis of seven cases. Skeletal Radiol 1992; 21:503-507.
- Ehara S, Nakasato T, Tamakawa Y, Yamataka H, Murakami H, Abe M. MRI of myositis ossificans circumscripta. Clin Imaging 1991; 15:130-134.
- Shirkhoda A, Armin AR, Bis KG, Makris J, Irwin RB, Shetty AN. MR imaging of myositis ossificans: variable patterns at different stages. J Magn Reson Imaging 1995; 5:287-292.
- Bressler EL, Marn CS, Gore RM, Hendrix RW. Evaluation of ectopic bone by CT. AJR Am J Roentgenol 1987; 148:931-935.
- Wharton GW. Heterotopic ossification. Clin Orthop 1975; 112:142-149.
- Rossier AB, Bussat P, Infante F, et al. Current facts of para-osteo-arthropathy (POA). Paraplegia 1973; 11:38-78.
- Chantraine A, Nusgens B, Lapiere CM. Biochemical analysis of heterotopic ossification in spinal cord injury patients. Paraplegia 1995; 33:398-401.
- Hernandez AM, Forner JV, de la Fuente T, Gonzalez C, Miro R. The para-articular ossifications in our paraplegics and tetraplegics: a survey of 704 patients. Paraplegia 1978; 16:272-275.
- Freed JH, Hahn H, Menter R, Dillon T. The use of the three-phase bone scan in the early diagnosis of heterotopic ossification (HO) and in the evaluation of Didronel therapy. Paraplegia 1982; 20:208-216.
- Prakash V, Lin MS, Perkash I. Detection of heterotopic calcification with 99mTc-pyrophosphate in spinal cord injury patients. Clin Nucl Med 1978; 3:167-169.
- Bravo-Payno P, Esclarin A, Arzoz T, Arroyo O, Labarta C. Incidence and risk factors in the appearance of heterotopic ossification in spinal cord injury. Paraplegia 1992; 30:740-745.
- Banovac K, Gonzalez F. Evaluation and management of heterotopic ossification in patients with spinal cord injury. Spinal Cord 1997; 35:158-162.
- Hassard GH. Heterotopic bone formation about the hip and unilateral decubitus ulcers in spinal cord injury. Arch Phys Med Rehabil 1975; 56:355-358.
- Wharton GW, Morgan TH. Ankylosis in the paralyzed patient. J Bone Joint Surg Am 1970; 52:105-112.
- Garland DE, Orwin JF. Resection of heterotopic ossification in patients with spinal cord injuries. Clin Orthop 1989; 242:169-176.
- Nicholas JJ. Ectopic bone formation in patients with spinal cord injury. Arch Phys Med Rehabil 1973; 54:354-359.
- Snoecx M, De Muynck M, Van Laere M. Association between muscle trauma and heterotopic ossification in spinal cord injured patients: reflections on their causal relationship and the diagnostic value of ultrasonography. Paraplegia 1995; 33:464-468.
- Silver JR. Heterotopic ossification: a clinical study of its possible relationship to trauma. Paraplegia 1969; 7:220-230.
- Subbarao JV, Garrison SJ. Heterotopic ossification: diagnosis and management, current concepts and controversies. J Spinal Cord Med 1999; 22:273-283.
- Amendola MA, Glazer GM, Agha FP, Francis IR, Weatherbee L, Martel W. Myositis ossificans circumscripta: computed tomographic diagnosis. Radiology 1983; 149:775-779.
- Garland DE, Blum CE, Waters RL. Periarticular heterotopic ossification in head-injured adults: incidence and location. J Bone Joint Surg Am 1980; 62:1143-1146.
- Spear IM, Lipscomb PR. Noninfectious trochanteric bursitis and peritendinitis. Surg Clin North Am 1952; 32:1217-1224.
- Bywaters EGL. The bursae of the body. Ann Rheum Dis 1965; 24:215-218.
- Nishi S. Topographical atlas of human anatomy Vol 3. Tokyo, Japan: Kaneharan Shuppan, 1974; 89-91.
- Hsu JD, Sakimura I, Stauffer ES. Heterotopic ossification around the hip joint in spinal cord injured patients. Clin Orthop 1975; 112:165-169.
- Finerman GA, Stover SL. Heterotopic ossification following hip replacement or spinal cord injury: two clinical studies with EHDP. Metab Bone Dis Relat Res 1981; 3:337-342.
- Ditunno JF, Jr, Formal CS. Chronic spinal cord injury. N Engl J Med 1994; 330:550-556.
- Muheim G, Donath A, Rossier AB. Serial scintigrams in the course of ectopic bone formation in paraplegic patients. Am J Roentgenol Radium Ther Nucl Med 1973; 118:865-869.
- Vento JA, Sziklas JJ, Spencer RP, Rosenberg RJ. "Reactivation" of Tc-99m MDP uptake in heterotopic bone. Clin Nucl Med 1985; 10:206-207.
作者：
Hans Peter Ledermann MD Mark E. Schweitzer MD a 2007-5-12