| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1
Department of Medical Imaging, Mount Sinai Hospital and the University Health
Network, University of Toronto, 600 University Ave., Toronto, Ontario, Canada
M5G 1X5.
2
Present address: Department of Radiology, St. James University Hospital,
Beckett St., Leeds LS9 7TF, United Kingdom.
3
Department of Radiology, St. Louis University School of Medicine, Health
Sciences Center, St. Louis, MO 63110-0250.
4
Present address: Department of Radiology, Mayo Clinic, 200 First St., S.W.,
Rochester, MN 55905.
5
Department of Pathology, Mount Sinai Hospital, University of Toronto, Toronto,
Ontario, Canada M5G 1X5.
6
Department of Orthopedic Oncology, Mount Sinai Hospital, University of
Toronto, Toronto, Ontario, Canada M5G 1X5.
7
Department of Orthopedic Surgery, St. Louis University School of Medicine,
Health Sciences Center, St. Louis, MO 63110-0250.
8
Department of Pathology, St. Louis University School of Medicine, Health
Sciences Center, St. Louis, MO 63110-0250.
Address correspondence to L. M. White.
Abstract
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
MATERIALS AND METHODS. Twenty-two patients (nine women and 13 men; mean age, 33 years) with pathologically proven periosteal chondroid lesions were retrospectively reviewed. The imaging modalities included conventional radiography (n = 17), CT (n = 10), and MR imaging (n = 14). The images were reviewed by two osteoradiologists, with agreement by consensus. Evaluation criteria included lesion location, mineralization, and size; periosteal reaction; and cortical response. Intramedullary extension, adjacent intramedullary edema, soft-tissue edema, and intrinsic characteristics were also evaluated on MR imaging. After the evaluation, a radiologic diagnosis of chondroma or chondrosarcoma was obtained. An experienced osteopathologist who was unaware of the patient's medical history and radiologic findings reviewed all histopathology. Agreement between the radiologic and the histopathologic diagnosis was tested using the kappa analysis. Imaging features were correlated with the pathologic findings, and a statistical analysis was performed.
RESULTS. Using strict pathologic criteria, we diagnosed 11
chondromas and 11 chondrosarcomas (nine, grade I; two, grade II). Moderate
agreement was reached between the radiologic and the pathologic diagnosis
(
= 0.55). The size of periosteal chondrosarcomas (range, 3-14 cm;
median, 4 cm) was considerably larger than the size of the chondromas (range,
1-6.5 cm; median, 2.5 cm; p < 0.05). Other imaging features did
not significantly correlate with benign versus malignant disease at pathology
(all p > 0.05).
CONCLUSION. A variable overlap existed in the imaging appearances of benign and malignant periosteal chondroid lesions, with size being the most reliable indicator in distinguishing the two lesions. This and the fact that histologic differentiation of the entities can be difficult, suggests that surgical wide excision may be the most appropriate procedure in treating patients with lesions greater than 3 cm.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Clinical onset, even with malignant tumors, is relatively insidious, and pain and swelling are usually the presenting features. Conventional radiography has been the primary imaging modality assessed in previous studies, and descriptions of cross-sectional imaging were limited to a small number of cases [8,9,10,11]. The typical radiographic appearance of a benign chondroma is described as saucerization of the adjacent cortex, with a well-formed sclerotic periosteal reaction. A soft-tissue mass is not frequently shown, and matrix calcification occurs in approximately 50% of patients [5, 6]. Treatment of these lesions is surgical, with a wide excision performed for chondrosarcoma and a local excision for chondroma [5,6,7,8, 10].
Pathologic interpretation of such lesions is difficult. Histologic features can overlap because chondromas can exhibit cellular atypia, and most periosteal chondrosarcomas are of low histologic grade [5, 8, 12,13,14]. Most reviews state that when pathology is indeterminate, conventional radiographic correlation is advisable because malignant lesions are typically larger, more ill-defined, and may show medullary extension [5, 8, 11, 13,14,15]. However, some studies have described discrepancies using this approach because periosteal chondromas may have aggressive radiographic appearances, and small chondrosarcomas may appear radiologically benign [5, 6].
The purpose of this study was to determine the features of periosteal chondroid tumors on conventional radiography and cross-sectional imaging and to correlate these appearances with histopathology.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Thirteen men and nine women, with a mean age of 33 years (age range, 22-64 years), were involved in this study. The frequency of bone involvement included the following: humerus (n = 10), tibia (n = 4), femur (n = 3), phalanges (n = 2), iliac bone (n = 2), and radius (n = 1). Long-bone involvement was metaphyseal in 16 patients and diaphyseal in four patients. In all patients, a chart review was performed to determine the extent of resection, length of follow-up, and presence of recurrent disease.
Image Analysis
For each patient, all imaging was reviewed by two musculoskeletal
radiologists who were unaware of the diagnosis, with agreement reached by
consensus. The imaging modalities included conventional radiography
(n = 17), CT (n = 10), and MR imaging (n = 15);
cross-sectional imaging was available in all cases. MR imaging consisted of
T1- and T2-weighted fat-suppressed sequences performed in at least two
orthogonal planes. Gadolinium-enhanced sequences were obtained in only four
cases and therefore were not included in the analysis.
Conventional radiography and CT evaluation included an assessment of lesion size, lesion mineralization, periosteal reaction (graded as complete or incomplete if present), and intramedullary extension. The MR imaging evaluation consisted of the same features recorded on other imaging modalities but also included intrinsic signal characteristics, intramedullary invasion, intramedullary edema, soft-tissue edema, and definition of surrounding soft-tissue margins.
In each patient, a radiologic assessment was obtained that classified the disease as benign (chondroma) or malignant (chondrosarcoma) on the basis of features traditionally thought to indicate malignancy (larger size, irregular margins, medullary invasion, and soft-tissue invasion) [5, 6].
Histopathologic Analysis
An experienced osteopathologist, unaware of the clinical and imaging
findings, reviewed all histopathology. Tumors were considered malignant if
they fulfilled the criteria of increased cellularity, cytologic atypia, and
binucleate cells invading soft tissue, or if they were located beyond the
limiting lamellar bone plate in the medullary canal
[5,
9].
Statistical Analysis
The recorded imaging features were correlated with the histopathologic
diagnosis, and the statistical significance was tested by Fisher's exact test.
For lesion size, the Mann-Whitney test was used to evaluate statistical
significance. Agreement between the radiologic and pathologic diagnoses was
assessed using the kappa statistic
[16].
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Image Analysis
Lesion size ranged from 1 to 14 cm (mean, 3.8 cm) in maximal dimension. For
the 19 patients who underwent conventional radiography and CT, the findings
were as follows: periosteal-based lesions, in 19 of 19; cortical scalloping,
in 10 of 19; cortical sclerosis, in eight of 19; periosteal shell, in 11 of 19
(complete in 1/11); internal chondroid matrix, in 14 of 19; and intramedullary
invasion, in five of 19.
For the 14 patients who underwent MR imaging, the findings were as follows: isointense or low T1 signal and increased T2 signal, in 14 of 14; intramedullary extension, in three of 14; intramedullary edema, in four of 14; irregular soft-tissue margins, in five of 14; and peritumoral soft-tissue edema, in two of 14.
Comparison of Imaging Findings in Chondroma and Chondrosarcoma
The median size of a low-grade periosteal chondrosarcoma (range, 3-14 cm;
median, 4 cm; mean, 5.3 cm) was considerably larger than that of chondroma
(range, 1-6.5 cm; median, 2.5 cm; mean, 2.2 cm; p = 0.0005).
The distribution of other imaging findings was as follows: matrix calcification (8/11 chondromas, 6/8 chondrosarcomas), intramedullary extension on conventional radiography or CT (3/11 chondromas, 2/8 chondrosarcomas) (Fig. 1A,1B), intramedullary extension on MR imaging (1/5 chondromas, 2/9 chondrosarcomas) (Fig. 1A,1B), intramedullary edema on MR imaging (1/5 chondromas, 3/9 chondrosarcomas) (Figs. 2A,2B,2C,2D and 3), soft-tissue edema on MR imaging (0/5 chondromas, 2/9 chondrosarcomas) (Fig. 3), and irregular soft-tissue margins on MR imaging (2/5 chondromas, 3/9 chondrosarcomas) (Figs. 3 and 4A,4B,4C). None of these features were closely related to pathology (p > 0.05).
|
|
|
|
|
|
|
|
|
|
Agreement of Imaging and Pathologic Diagnosis
Radiologically, 14 lesions were evaluated as nonaggressive or benign, with
pathology classifying 10 of 14 chondromas and four of 14 chondrosarcomas (Fig.
5A,5B,5C).
Radiologically, eight lesions were evaluated as aggressive or malignant, with
pathology classifying one of eight chondromas (Fig.
2A,2B,2C,2D)
and seven of eight chondrosarcomas. Thus only moderate agreement was seen
between the radiologic and pathologic assessment of lesions, with a kappa
score of 0.55 [17].
|
|
|
Clinical Follow-Up
The average follow-up for all patients (n = 22) was 45 months
(range, 4-120 months). Marginal excision and curettage were performed in seven
of 22 patients (five chondromas, two chondrosarcomas) and a wide excision was
performed in 15 of 22 patients (six chondromas, nine chondrosarcomas).
In our patient group, there have been two recurrences. Both cases of recurrent disease were seen in patients initially diagnosed with chondrosarcoma (both, grade I). The first patient underwent marginal excision of an iliac lesion that recurred 12 months after surgery. The second patient underwent marginal excision of a femoral lesion that recurred 24 months after surgery.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
To our knowledge, no large series has evaluated the use of cross-sectional imaging in these entities. The largest series were performed before the introduction of MR imaging, with conventional radiographs available in 31 of a possible 60 cases [5]. Most reports state that chondrosarcomas are typically larger, more irregular, and permeative on conventional radiography than chondromas [5, 6, 8, 13, 14]. In fact, the reports advise that in most patients these radiographic features should be used to determine the diagnosis when pathology is indeterminate. Case reports and short series have reported findings on CT and MR imaging predominantly for chondromas [8,9,10]. One case report described intramedullary extension and edema on MR imaging in patients with periosteal chondrosarcoma [11].
Our results show that periosteal chondrosarcomas are larger (range, 3-14 cm; median, 4 cm) than chondromas (range, 1-6.5 cm; median, 2.5 cm), and this finding was statistically significant (p < 0.05). No chondrosarcomas were smaller than 3 cm. However, an overlap was observed in the ranges obtained. Although other findings thought to be more typical of chondrosarcoma occurred more commonly with this diagnosis, the features were not sensitive (soft-tissue edema [2/9 chondrosarcomas]) or specific (intramedullary edema on MR imaging [1/5 chondromas, 3/9 chondrosarcomas], intramedullary extension on conventional radiography or CT [3/11 chondromas, 2/8 chondrosarcomas], intramedullary extension on MR imaging [1/5 chondromas, 2/9 chondrosarcomas], and irregular soft-tissue margins on MR imaging [2/5 chondromas, 3/9 chondrosarcomas]). The difficulty in clearly defining distinct imaging features differentiating benign and low-grade malignant periosteal chondroid lesions is similar to the problem of differentiating benign and low-grade malignant intramedullary chondroid lesions on imaging [17, 18].
In our series, clinical follow-up showed recurrence in two cases of chondrosarcoma treated by marginal excision. No recurrences were reported in lesions treated by wide (segmental) excision. Although several studies agree that local excision is sufficient for chondroma, some have advocated wide excision for treatment of periosteal chondrosarcoma [5, 12,13,14].
Limitations of our study are its retrospective design and surgical bias, which may explain the larger proportion of chondrosarcomas in our series compared with previous studies [5]. However, given the rarity of this condition, a true prospective study would be difficult. Pathologic confirmation (at surgery) is necessary to make a comparison meaningful. Another limitation of MR imaging was the lack of gadolinium-enhanced sequences for analysis. A recent study of dynamic gadolinium enhancement showed some differentiation of benign versus malignant intraosseous chondrois lesions, although overlap of enhancement patterns was seen in this study [19]. To our knowledge, this differentiation has not been assessed in periosteal-based lesions.
We found only moderate agreement between radiologic assessment of these
tumors and the pathologic diagnosis ( = 0.55). This finding suggests that
traditionally assessed imaging features of periosteal chondroid lesions are
insensitive and may not be sufficient in differentiating some cases of benign
versus malignant disease. Indeed, our results show an overlap in the imaging
appearances of benign and malignant periosteal chondroid lesions.
The most reliable predictor in our series was lesion size. This finding suggests that surgical wide (segmental) excision may be the most appropriate treatment for these patients to prevent locally recurrent disease, especially when lesion size exceeds 3 cm.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  S. Chaabane, M. C. Bouaziz, C. Drissi, L. Abid, and M. F. Ladeb Periosteal Chondrosarcoma Am. J. Roentgenol., January 1, 2009; 192(1): W1 - W6. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. J. Walden, M. D. Murphey, and J. A. Vidal Incidental Enchondromas of the Knee Am. J. Roentgenol., June 1, 2008; 190(6): 1611 - 1615. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Y.-Y. An, J.-Y. Kim, M.-I. Ahn, Y.-K. Kang, and H.-J. Choi Enchondroma Protuberans of the Hand Am. J. Roentgenol., January 1, 2008; 190(1): 40 - 44. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. D. Murphey, J. S. Jelinek, H. T. Temple, D. J. Flemming, and F. H. Gannon Imaging of Periosteal Osteosarcoma: Radiologic-Pathologic Comparison Radiology, October 1, 2004; 233(1): 129 - 138. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. D. Murphey, E. A. Walker, A. J. Wilson, M. J. Kransdorf, H. T. Temple, and F. H. Gannon From the Archives of the AFIP: Imaging of Primary Chondrosarcoma: Radiologic-Pathologic Correlation RadioGraphics, September 1, 2003; 23(5): 1245 - 1278. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
