Abstract
Background: Lesions of fibrous dysplasia involving the spine and
scoliosis are thought to be uncommon entities in patients with polyostotic
fibrous dysplasia and McCune-Albright syndrome. By examining bone scans of a
relatively large cohort of patients with these disorders, we sought to
determine the prevalence of spinal involvement and any association with
scoliosis.
Methods: Sixty-two patients with polyostotic fibrous dysplasia were
studied. There were twenty-three male and thirty-nine female patients, and
they had a mean age of twenty-five years (range, four to eighty years).
Technetium-99m-methylene diphosphonate (MDP) bone scans of the patients were
evaluated for evidence of increased uptake in the spine. The presence or
absence of scoliosis or a level pelvis and the distribution of other lesions
in the skeleton were noted.
Results: Thirty-nine (63%) of sixty-two patients were found to have
seventy-six lesions of fibrous dysplasia in the spine. Fifty-four lesions
(71%) demonstrated increased uptake in the posterior aspects of the spine.
Most lesions were located in the lumbar spine (thirty-two lesions) and the
thoracic spine (twenty-seven), with less frequent involvement in the sacrum
(ten) and cervical spine (six). Twenty-five (40%) of the sixty-two patients
had scoliosis; seventeen had a thoracolumbar curve; six, a lumbar curve; and
two, a thoracic curve. Seven patients had curves that could not be accurately
measured by bone scanning and, therefore, could not be classified. Thirty
patients (48%) had no evidence of scoliosis. Thus, the prevalence of scoliosis
in patients with polyostotic fibrous dysplasia was between 40% and 52%. There
was a strong correlation between spinal lesions and scoliosis (p < 0.001)
and pelvic asymmetry (p < 0.05). Back pain was an uncommon symptom. Two
patients had a neurologic abnormality; neither abnormality was related to the
location of the lesions or the curve.
Conclusions: Spinal lesions and scoliosis may be more common in
patients with polyostotic fibrous dysplasia than has been previously reported.
Since there is a strong correlation between a spinal lesion and scoliosis,
these patients should be screened clinically for scoliosis.
Level of Evidence: Prognostic study, Level II-1
(retrospective study). See Instructions to Authors for a complete description
of levels of evidence.
Fibrous dysplasia of bone is characterized by the replacement of bone and
marrow with poorly organized spicules of immature bone in a fibrous connective
tissue. The disorder, first described in 1938 by
Lichtenstein1, was
subsequently found to have monostotic and polyostotic varieties, the latter of
which may be coupled with cutaneous and/or endocrine abnormalities
(McCune-Albright
syndrome)2,3.
Overall, monostotic fibrous dysplasia is more common (80%) than polyostotic
fibrous dysplasia
(20%)4. Although
fibrous dysplasia accounts for approximately 7% of all benign bone lesions,
the prevalence of spinal involvement is thought to be very low, particularly
in the monostotic
form3,5-21.
Fibrous dysplasia involving the spine has also been seen in individuals with
the McCune-Albright syndrome, and the prevalence has been reported to range
from 7% in the cervical spine to 14% (five of thirty-seven patients) in the
lumbar spine22.
Isolated cases of patients who had malignant transformation of spinal lesions
have also been
reported4,23.
The prevalence of scoliosis in patients with polyostotic fibrous dysplasia
and the McCune-Albright syndrome is not known. Guille and
Bowen8 reported on
three patients with polyostotic fibrous dysplasia and scoliosis, two of whom
required posterior arthrodesis. Janus et
al.11 reported the
case of one such patient, who was also treated with posterior arthrodesis.
Fibrous dysplasia both expands and weakens bone, resulting in structural
changes. Although fibrous dysplasia involvement in the spine has been reported
to cause changes in vertebral
shape9, no
correlation between the lesion location and the development of scoliosis has
been described. Other anatomic considerations, such as limb-length discrepancy
of the lower extremities or pelvic obliquity, which are also common in
patients with fibrous dysplasia, may play a role in the etiology of
scoliosis.
Patients
All patients provided written informed consent before enrollment in our
long-term study of the natural history of polyostotic fibrous dysplasia or
McCune-Albright syndrome, which was approved by the institutional review
board. Enrollment in the study began in 1998, and two patients subsequently
withdrew. The majority of the patients were recruited through a web site that
described the natural history study and allowed individuals with polyostotic
fibrous dysplasia to apply for enrollment in the study online. Additional
patients were referred by their primary care physicians. The National
Institutes of Health, as a research institution, would be expected to attract
patients with a considerable amount of disease.
The study population consisted of sixty-two patients. There were
twenty-three male and thirty-nine female patients, and the mean age was
twenty-five years (range, four to eighty years). As part of the study, all
patients had technetium-99m-methylene diphosphonate (MDP) bone scans performed
annually in order to track the location and activity of bone lesions.
Whole-body bone scan images, made in the anterior and posterior planes with
the patient in the supine position, were available for all patients. In
addition, all patients had skeletal surveys of the pelvis and upper and lower
extremities. When the protocol was first designed, radiographs of the spine
were not included in order to remain within the allowable radiation quota for
human subjects set by our institutional radiation safety committee. The
decision not to include spine radiographs was made on the basis of the data in
the available literature, which suggested that fibrous dysplasia involving the
spine was an uncommon
occurrence3,7,9,12,19-21,24.
Bone Scan Measurements
The bone scan was used to determine the location of lesions in the
cervical, thoracic, and lumbar spine and the sacrum, as well as the prevalence
of scoliosis. The orientation of the pelvis was determined by assessing the
angulation between the vertebrae in the lumbar spine and the line drawn across
the top of the pelvis, and the pelvis was described as either level or not
level with the shoulders. Lesions in the base of the skull, pelvis, long
bones, and ribs were also recorded. Two different observers reviewed the bone
scan data independently.
An estimate of the Cobb angle was performed with use of the end plates seen
on the bone scan that yielded the largest magnitude of curvature. Actual Cobb
angles were determined from spine or chest radiographs, made with the patient
standing, which were performed for eight patients for clinically indicated
reasons. The difference between the estimated and actual angles in these eight
patients was used to determine the degree of error in estimating Cobb angles
from bone scans.
Other Testing
Patients underwent a comprehensive physical examination by an
endocrinologist and a physiatrist on a yearly basis. Other specialists became
involved in patient care as necessitated by the specific needs of each study
patient. Patients were not routinely examined by either an orthopaedist or a
neurologist; however, most of the patients were seen by a local orthopaedist
whose records were incorporated into the patient's research chart. As a part
of the study protocol, all patients had clinical photographs made to document
skin lesions; blood analysis and urinalysis to assess for all of the
endocrinopathies associated with polyostotic fibrous dysplasia or
McCune-Albright syndrome, including precocious puberty, hyperthyroidism,
acromegaly, hyperparathyroidism, and Cushing syndrome; and urinalysis to
determine renal phosphate wasting (phosphaturia).
Statistical Analysis
Statistical analyses were performed with use of SAS software (version 6.12;
SAS Institute, Cary, North Carolina). All numerically continuous parameters
were summarized with use of means and standard deviations; means were compared
with use of a one-way analysis of variance. Other analyses performed with SAS
included Fisher exact tests and chi-square analyses.
The clinical characterization of the patient cohort is summarized in
Table I and
Figure 1. Of the sixty-two
study patients, forty-four (71%) had McCune-Albright syndrome with multiple
bone lesions, with or without café au lait spots, and with
endocrinopathies, including precocious puberty, hyperthyroidism, acromegaly,
and/or phosphaturia. The remaining eighteen patients had polyostotic fibrous
dysplasia without endocrinopathy or phosphaturia (see Appendix). As we
previously reported, the basic histological features of the biopsy specimens
from patients with polyostotic fibrous dysplasia and the McCune-Albright
syndrome were
similar4.
An examination of the routine bone scans identified thirty-nine patients
(63%) with seventy-six discrete lesions involving the spine. These lesions
typically demonstrated increased tracer uptake. Fifty-four lesions (71%) were
seen best in the scans made in the posterior plane for twenty-seven (69%) of
the thirty-nine patients, while those made in the anterior plane demonstrated
increased uptake in five patients (13%). Seven patients (18%) demonstrated
extensive lesions in the scans made in both the anterior and posterior planes.
When divided into groups on the basis of the different locations of the spine
(cervical, thoracic, lumbar, and sacrum), twenty-eight patients (72%) had
lesions that were localized to one or two anatomic sites, whereas the
remaining nine patients (23%) had more extensive disease with three or all
four anatomic locations involved (Fig.
2).
Patients were asked to report body pain at the time of initial enrollment,
and few reported back pain. A complete neurologic examination was performed on
all patients. One patient had neurologic changes attributable to basilar
invagination, and another patient had a diagnosis of a peripheral neuropathy;
the remaining sixty patients were neurologically intact. The spine was
examined for scoliosis specifically by the physiatrist, but spinal deformity
was not diagnosed in one-fifth of the patients with scoliosis. Physical
examination is particularly hard in these patients secondary to multiple
deformities from involvement of fibrous dysplasia in the proximal part of the
femora and the pelvis, in addition to that in the spine.
Although scoliosis is best determined radiographically, the high prevalence
of spinal lesions in these patients prompted a closer evaluation of the
patients for the presence of scoliosis. The possibility of using the bone
scans to detect scoliosis was evaluated by a comparison of the estimated Cobb
angles in the eight patients for whom both bone scans and plain radiographs
were available (Fig. 3). By
comparing the values, we found that, in every case, the bone scan measurement
underestimated the magnitude of the curve. The difference between the bone
scan and the radiographic measurement ranged from 4.4% to 50.0%, with a
percentage of error of 27.7% ± 15.0%
(Fig. 4). Thus, although bone
scan estimations are not very accurate in the measurement of scoliosis, they
can be useful in the detection of spinal lesions that in turn correlate with
the presence of a curve.
With use of the bone scan estimations, twenty-five (40%) of the sixty-two
patients were identified as having a definitive scoliosis of >15°;
thirty patients (48%) did not have scoliosis. The findings in the remaining
seven patients (12%) could not be classified with regard to the presence or
absence of scoliosis on the basis of the bone scan. These patients had an
estimated Cobb angle measurement of between 6° and 12°, which is
indeterminate because of the high percentage of error inherent in the bone
scan measurement. Despite the imprecision of the estimated Cobb angle derived
from a bone scan, these data demonstrated that the prevalence of scoliosis in
this population of patients with polyostotic fibrous dysplasia and
McCune-Albright syndrome is between 40% and 52% (if the indeterminate curves
are all actual scoliotic curves).
The most common site of scoliosis was in the thoracolumbar spine (seventeen
patients; 68%), followed by the lumbar spine (six patients; 24%) and the
thoracic spine (two patients; 8%). Twenty-three (92%) of the twenty-five
patients with definitive scoliosis had lesions of fibrous dysplasia contained
within the confines of the curve, and only two patients (8%) had lesions
outside the curve; the correlation between a spinal lesion and a curve was
significant (Fisher exact test, p < 0.001). Fourteen patients had spinal
lesions without any evidence of curvature of the spine, and only two patients
had scoliosis in the absence of fibrous dysplasia in the spine.
Four patients included in the analysis had posterior spinal arthrodesis and
instrumentation with a variety of devices. All procedures were performed at
other institutions. All of these patients underwent an uneventful arthrodesis
with no evidence of loss of fixation or nonunion, on the basis of a review of
their spine radiographs, at an average of twelve years (range, four to
twenty-three years) of follow-up. While they were enrolled in the study, the
four patients were monitored and showed no progression of the scoliosis after
the spinal arthrodesis.
In the total population, we identified forty-six café au
lait skin lesions (in patients both with and without endocrinopathy). We
further classified a midline lesion as one that occurs centrally along the
spine and a nonmidline lesion as one that occurs lateral to the central axis
of the body (Fig. 5). Nineteen
patients (31%) had midline lesions, twenty-seven patients (44%) had lateral
skin lesions, and the remaining sixteen patients (26%) did not have skin
involvement. No correlation was found between a midline skin lesion and
scoliosis.
Only one of the endocrinopathies, precocious puberty (thirty patients),
correlated with the presence of lesions in the spine (Fisher exact test, p
< 0.02). Since precocious puberty can increase growth velocity, one would
expect to see an increase in scoliosis in these patients, but no significant
relationship was detected between precocious puberty and scoliosis. Instead,
the presence of scoliosis appeared to be related to the presence of
phosphaturia (twenty-two patients; Fisher exact test, p < 0.005) and
hyperparathyroidism (six patients; Fisher exact test, p < 0.01). None of
the other endocrinopathies appeared to be related either to the presence of
lesions or to scoliosis; however, the number of patients in each group may
have been too small to identify interactions.
For each patient, we also compared the presence of lesions in the long
bones, pelvis, ribs, or base of the skull with the presence of lesions in the
spine. However, with the numbers available, we found no correlation between
lesion location in the spine and lesions in the long bones (p = 0.99), pelvis
(p = 0.26), ribs (p = 0.35), or base of the skull (p = 0.13).
Since bone scans are performed with the patient in the supine position,
they were also evaluated for the presence of a level pelvis. Nine patients
(15%) had pelvic angulation from the midline of the spine, while the remaining
fifty-three had a level pelvis. There was a significant correlation between
the absence of a level pelvis and scoliosis (p < 0.05).
On the basis of the findings of our study, involvement of fibrous dysplasia
in the spine is more common than previously reported, with a prevalence of 63%
in this group of patients with polyostotic fibrous dysplasia or
McCune-Albright syndrome. The bone scan data in the present study may, in
fact, underestimate the true extent of disease since some areas of the spine
can be obscured on the bone scan. For example, the anterior aspect of the
cephalad thoracic spine is difficult to visualize under the sternum, and the
mandible can cover the anterior aspect of the cervical spine. In addition, it
is difficult to assign the precise location of lesions found at the rib-spine
junction, which were designated rib lesions in this study but may actually
represent involvement of fibrous dysplasia in the transverse processes.
While the prevalence of scoliosis could be estimated with use of bone
scans, it was apparent that, for the eight patients who had both a bone scan
and a plain radiograph, the bone scan consistently underestimated the degree
of curvature. In addition, without radiographs, sagittal plane deformity such
as kyphosis was missed. Anterior vertebral body disease may result in kyphotic
deformity, which could not be measured on the bone scans, as no lateral scans
were made. On the basis of the data from the bone scans and plain radiographs
of the long bones of our cohort, we noted that all lesions of fibrous
dysplasia on radiographs showed some increased uptake on bone scan. Thus, we
are confident that we did not undercount lesions in the spine by examining the
bone scan data. However, we intend to confirm this assumption with
radiographs, computed tomography, and magnetic resonance imaging of our
patients, a project currently underway. Nonetheless, the degree of spinal
involvement detected by bone scanning is much higher than had been previously
suspected, and this finding provides valuable information to clinicians
concerning the need for evaluation of the spine in patients with polyostotic
fibrous dysplasia and McCune-Albright syndrome.
For each patient, we noted lesions in the long bones, pelvis, ribs, or base
of the skull to determine whether multiple sites of involvement in various
anatomic areas are based on proximity. For example, do rib lesions and
thoracic spine lesions, or a lesion in the sacrum and one in the pelvis, occur
in combination? However, no association was found between spinal disease and
lesions in other areas of the skeleton. The apparent random distribution of
lesions in the body is thought to be related to the temporal and anatomic
point during embryologic development when the missense mutation in the GNAS1
gene is introduced. Other studies have found that there also seems to be no
relationship between the location of café au lait spots and
the location of skeletal
lesions1,22.
Interestingly, we noted from the clinical photographs of our patients that
very distinct café au lait spots were consistently located in
the midline of the spine. These lesions were different from other lesions that
stretched across the body laterally. However, separating midline skin lesions
from all skin lesions did not reveal a correlation between midline lesions and
spinal lesions in bone. We therefore cannot report any significant difference
between the two types of skin lesions that we observed in terms of correlation
of their presence with the presence of fibrous dysplasia in the spine.
The prevalence of both spine involvement and scoliosis was higher than
expected. Our findings indicate that bone scan data may be useful in
determining when plain radiographs of the spine are needed for clinical
management, especially for patients who are difficult to examine clinically
because of deformities about the pelvis and lower extremities.
Also, as our study patients return for yearly follow-up examinations, they
will undergo magnetic resonance imaging through the lesions in order to more
accurately describe the anatomic location of the lesions, as well as the
morphological characteristics of the vertebrae. Our cohort, who sought out a
research protocol, may represent patients with a larger disease burden than
those who are treated in the community at large, but this group will allow us
to monitor the progression of scoliosis after skeletal maturity in these
conditions.
The natural history of fibrous dysplasia of the spine is unknown. Many
patients with polyostotic fibrous dysplasia have precocious puberty, which can
result in the adolescent growth spurt occurring at an earlier age. Since
scoliosis progresses during periods of rapid growth, it is possible that
scoliosis occurs at an earlier age in patients with polyostotic fibrous
dysplasia who have precocious puberty. Spinal deformity caused by vertebral
deformity from replacement of bone with fibrous tissue may continue to
progress into adulthood, and we hope that further study of fibrous dysplasia
of the spine in our population will allow us to be able to determine the
natural history of curve progression.
We believe that patients with polyostotic fibrous dysplasia need to be
surveyed for spinal involvement, and that this is best accomplished by a bone
scan. If lesions are found in the spine, then radiographs of the spine should
be made, a detailed physical examination should be performed, and the
individual should be closely monitored, and, if necessary, treated for
scoliosis.
A table showing specific data on all sixty-two patients is available with
the electronic versions of this article, on our web site at
(go to the article citation and click on "Supplementary Material")
and on our quarterly CD-ROM (call our subscription department, at
781-449-9780, to order the CD-ROM).
Lichtenstein L. Polyostotic
fibrous dysplasia. Arch Surg.1938;36:
874-98.36874
1938
Albright F, Butler AM, Hampton AO,
Smith P. Syndrome characterized by osteitis fibrosa disseminata, areas of
pigmentation and endocrine dysfunction, with precocious puberty in females.
Report of five cases. N Engl J Med.1937;216:
727-46.216727
1937
[CrossRef]
Nabarro MN, Giblin PE. Monostotic
fibrous dysplasia of the thoracic spine. Spine.1994;19:
463-5.19463
1994
[PubMed][CrossRef]
Riminucci M, Liu B, Corsi A, Shenker
A, Spiegel AM, Robey PG, Bianco P. The histopathology of fibrous dysplasia
of bone in patients with activating mutations of the Gs alpha gene:
site-specific patterns and recurrent histological hallmarks. J
Pathol.1999;187:
249-58.187249
1999
[CrossRef]
Avimadje AM, Goupille P, Zerkak D,
Begnard G, Brunais-Besse J, Valat JP. Monostotic fibrous dysplasia of the
lumbar spine. Joint Bone Spine.2000;
67: 65-70.6765
2000
[PubMed]
Chow LT, Griffith J, Chow WH, Kumta
SM. Monostotic fibrous dysplasia of the spine: report of a case involving
the lumbar transverse process and review of the literature. Arch
Orthop Trauma Surg.2000;120:
460-4.120460
2000
[CrossRef]
Ehara S, Kattapuram SV, Rosenberg
AE. Fibrous dysplasia of the spine. Spine.1992;17:
977-9.17977
1992
[PubMed][CrossRef]
Guille JT, Bowen JR. Scoliosis
and fibrous dysplasia of the spine. Spine.1995;20:
248-51.20248
1995
[PubMed][CrossRef]
Hoffman KL, Bergman AG, Kohler S.
Polyostotic fibrous dysplasia with severe pathologic compression fracture of
L2. Skeletal Radiol.1995;
24: 160-2.24160
1995
[PubMed]
Hu SS, Healey JH, Huvos AG.
Fibrous dysplasia of the second cervical vertebra. A case report. J
Bone Joint Surg Am.1990;72:
781-3.72781
1990
Janus GJ, Engelbert RH, Pruijs
JE. Instrumentation for correction and fixation of scoliosis in fibrous
dysplasia of the thoracolumbar spine. Eur Spine J.1998;7:
260-2.7260
1998
[PubMed][CrossRef]
Kahn A 3rd, Rosenberg PS.
Monostotic fibrous dysplasia of the lumbar spine.
Spine.1988;13:
592-3.13592
1988
[PubMed][CrossRef]
Nigrisoli M. Monostotic fibrous
dysplasia of the spine: case report. Ital J Ortho
Traumatol.1987;13:
273-8.13273
1987
Oba M, Nakagami W, Maeda M, Kobayashi
K. Symptomatic monostotic fibrous dysplasia of the thoracic spine.
Spine.1998;23:
741-3.23741
1998
[PubMed][CrossRef]
Penrod BJ, Haramati N, Frager D,
Dorfman HD. A 40-year-old woman with back pain. Clin
Imaging.1996;20:
143-6.20143
1996
[CrossRef]
Przybylski GJ, Pollack IF, Ward
WT. Monostotic fibrous dysplasia of the thoracic spine. A case report.
Spine.1996;21:
860-5.21860
1996
[PubMed][CrossRef]
Resnik CS, Lininger JR.
Monostotic fibrous dysplasia of the cervical spine: case report.
Radiology.1984;151:
49-50.15149
1984
[PubMed]
Rodenberg J, Jensen OM, Keller JN,
Nielsen OS, Bunger C, Jurik AG. Fibrous dysplasia of the spine, costae and
hemipelvis with sarcomatous transformation. Skeletal
Radiol.1996;25:
682-4.25682
1996
[CrossRef]
Rosenblum B, Overby C, Levine M,
Handler M, Sprecher S. Monostotic fibrous dysplasia of the thoracic spine.
Spine.1987;12:
939-42.12939
1987
[PubMed][CrossRef]
Troop JK, Herring JA. Monostotic
fibrous dysplasia of the lumbar spine: case report and review of the
literature. J Pediatr Orthop.1988;8:
599-601.8599
1988
[PubMed][CrossRef]
Wright JF, Stoker DJ. Fibrous
dysplasia of the spine. Clin Radiol.1988;39:
523-7.39523
1988
[PubMed][CrossRef]
Harris WH, Dudley HR Jr, Barry
RJ. The natural history of fibrous dysplasia. An orthopaedic,
pathological, and roentgenographic study. Am J Orthop.1962;44:
207-33.44207
1962
Yalniz E, Er T, Ozyilmaz F.
Fibrous dysplasia of the spine with sarcomatous transformation: a case report
and review of the literature. Eur Spine J.1995; 4:
372-4.4372
1995
[PubMed][CrossRef]
Buckwalter JA, Bonfiglio MM,
el-Khoury G, Platz CC. Case report 490: monostotic fibrous dysplasia of
the pubis. Skeletal Radiol.1988;17:
361-3.17361
1988
[PubMed][CrossRef]