View metadata, citation and similar papers at core.ac.uk brought to you by CORE provided by RERO DOC Digital Library Skeletal Radiol (2002) 31:88–95 DOI 10.1007/s00256-001-0450-3 ARTICLE Achim H. Kaim Chondroblastoma and clear cell Rolf Hügli Harald M. Bonél chondrosarcoma: radiological Gernot Jundt and MRI characteristics with histopathological correlation Received: 12 June 2001 Abstract Objective: To analyze and droblastomas were more confined to Accepted: 15 October 2001 compare the radiological and magnetic the epiphysis (10/12) than the clear Published online: 14 December 2001 resonance imaging (MRI) appearances cell chondrosarcomas. All chondro- © ISS 2001 of chondroblastoma and clear cell blastomas and clear cell chondrosar- chondrosarcoma with histopathologi- comas except one were classified as cal correlation. Design and patients: grade 1A or 1B according to the Twelve patients with histologically LRGS; one clear cell chondrosarcoma proven chondroblastoma and of an- was judged as grade 2. Signal intensi- other four patients with clear cell ty of the tumors on MRI was very chondrosarcoma were investigated by heterogeneous in both groups. High radiographs and MRI (T1-, T2- signal intensity on T2-weighted MR weighted sequences, intravenous ga- images in chondroblastoma mostly dolinium application). Additionally, corresponded to areas with aneurys- the clinical and radiologic data of mal bone cyst components and in seven cases of clear cell chondrosar- clear cell chondrosarcoma to islands coma without available MRI were of hyaline cartilage. Contrast en- considered. The localization, calcifi- hancement occurred in all tumors and cation of tumor matrix, periosteal re- tended to be more intense with clear action, cortical bone and patterns of cell chondrosarcoma. Chondroblasto- bone destruction were analyzed ac- ma was more frequently associated cording to the Lodwick radiological with bone marrow edema (11/12), pe- grading system (LRGS). The signal riosteal reaction (10/12), soft tissue intensity on T1- and T2-weighted se- reaction (7/12) and synovitis (3/12). quences, characteristics of contrast Conclusion: Chondroblastoma occurs enhancement, associated bone mar- in younger patients, is smaller than row edema, soft tissue reaction and clear cell chondrosarcoma and is A.H. Kaim (✉) · R. Hügli joint involvement were evaluated. more confined to the epiphysis. The Institute of Diagnostic Radiology, University Hospital Basle, Petersgraben 4, Histopathological specimens were overlap of signal intensity and con- 4031 Basle, Switzerland available in all cases. Results: The trast enhancement patterns does not e-mail: [email protected] age of patients with chondroblastoma allow a reliable differentiation of the Tel.: +41-61-2654344 (range 15–59 years, mean 22.3 years) two tumors by MRI. Chondroblasto- Fax: +41-61-2655383 was lower than that of those with mas are typically associated with H.M. Bonél clear cell chondrosarcoma (range bone marrow edema, periosteal reac- Institute of Clinical Radiology, 19–61 years, mean 36.6 years), and tion and soft tissue reaction. University Hospital Munich-Grosshadern, Germany the lesions were smaller in the chon- droblastoma group (range 1–4 cm, Keywords Chondroblastoma · Clear G. Jundt Institute of Pathology, mean 2.3 cm) than in patients with cell chondrosarcoma · Magnetic University Hospital Basle, clear cell chondrosarcoma (range resonance imaging · MRI · Basle, Switzerland 3–7.5 cm, mean 5.2 cm). The chon- Histopathology 89 Introduction and an MRI examination were obtained. Imaging was done prior to any diagnostic or therapeutic intervention. Group 2 comprised 11 patients (9 men, 2 women), median age 36.6 years (range Chondroblastoma is a rare, benign cartilaginous bone tu- 19–61 years) in whom the tumor was histologically diagnosed as mor that is treated by curettage, and overall has a good clear cell chondrosarcoma. Conventional radiographs were ob- prognosis with a low morbidity. The tumor appears at tained in all patients, but MR images in only four. Five lesions centers of enchondral ossification, most commonly in the were located in the proximal femur, two in the proximal humerus, and one each in the distal femur, proximal tibia, upper pubic rim epiphysis of the long bones. Very rarely the epicenter and maxilla. may be in the metaphysis or apophysis. Radiographically it is a sharply delineated osteolytic defect with a sclerotic border. Subtle cartilaginous calcifications are seen in Imaging over 50% of cases and a periosteal reaction may also be The MRI parameters varied depending on the institution perform- present [1, 2]. The differential diagnoses of a tumor lo- ing the imaging, and the field strength ranged from 0.5 to 1.5 T. calized in the epiphysis are limited, but include osteomy- T1- and T2-weighted spin echo or fast spin echo images were ac- elitis, giant cell tumor, and clear cell chondrosarcoma. quired in all patients with gadolinium-DTPA (standard intravenous Giant cell tumors can mostly be differentiated by the age dose of 0.1 mmol/kg body weight) enhanced T1-weighted se- quences, four of which had spectral fat saturation. The lesions of the patient, the size of the lesion, its extension into the were documented in at least two orthogonal planes; the slice thick- metaphysis and its poorer delineation. Clear cell chon- ness varied between 3 and 6 mm according to the size of the le- drosarcoma and chondroblastoma, however, may have a sion. The in-plane resolution was 1 mm2 or less for all MRI exam- very similar radiographic appearance, which prevents a inations. All examinations of this retrospective evaluation were reliable differentiation of the two tumors. The mean age performed for clinical indications. of tumor occurrence in clear cell chondrosarcoma is one decade later, but the overlap of the age range compli- Analysis criteria cates the differential diagnosis. A precise diagnosis is important as the therapy of clear cell chondrosarcoma is Radiographs and MRI were analyzed by two radiologists (A.H.K., more aggressive than that of chondroblastoma and may R.H.) in consensus. They were masked to the clinical and histo- logical findings. involve en-bloc resection [3, 4, 5]. The radiographs were analyzed with respect to the anatomical Previous studies of the MRI appearance of chondro- location of the tumor, size (long axis size on radiographs), matrix blastoma have demonstrated that it is frequently associ- calcification, periosteal reaction and tumor extent applying the ated with bone marrow edema, periosteal and soft tissue Lodwick classification [10, 11]. The MRI features evaluated were signal intensity on T1- and reactions and varying signal intensity on T2-weighted T2- weighted images, morphology of the lesion itself (homogene- sequences [6, 7, 8]. It seems doubtful whether the MRI ity, heterogeneity), contrast medium uptake (intense/ medium/ criteria as described in previous case reports [9] allow a low, heterogeneous/ homogeneous), periosteal reaction, bone mar- reliable differentiation between chondroblastoma and the row edema, soft tissue reaction and synovitis with joint effusion. Lesions with a higher signal intensity on the T1- and T2-weighted histologically malignant clear cell chondrosarcoma, sequences compared with fatty bone marrow were considered as which requires a more extended surgical approach and hyperintense, and lesions with lower signal intensity were defined has an overall poorer prognosis. as hypointense. The aim of this study was to compare the radiological and MRI patterns of the two tumors and to correlate spe- cific radiological findings with histopathology. Histology All histological samples were analyzed by the same experienced musculoskeletal pathologist (G.J.). As all the chondroblastomas Patients and methods were treated by curettage and therefore no en-bloc resection was available, a general correlation with the radiological result was Patients performed. Three of four chondrosarcomas that had been imaged with MRI were treated by en-bloc-resection and their histomor- Between 1971 and 1999 a total of 93 cases of chondroblastoma phology could be correlated with the MR study, whereas in one and 11 cases of clear cell chondrosarcoma were referred to the case only a diagnostic biopsy of the tumor was available that al- Bone Tumor Reference Center for analysis. A retrospective analy- lowed a rather general correlation. sis was performed by two experienced musculoskeletal radiolo- gists (A.H.K., R.H.) and divided into one group (group 1) consist- ing of 12 cases of histologically proven chondroblastoma in which a diagnostic MRI examination was available, and a second group Results (group 2) comprising 11 patients with clear cell chondrosarcoma only four of whom had an MRI examination. Seven patients with chondroblastoma had a fully fused Group 1 consisted of 11 men and one woman, mean age growth plate and in the remaining five patients it was 22.3 years (range 15–59 years). Seven lesions were located in the distal femur, two in the proximal tibia, and one each in the meta- partially fused. All 11 patients with clear cell chondro- carpal bone of the first digit, the first metatarsal bone and the sarcoma had completely fused growth plates and apo- proximal humerus. In all the patients conventional radiographs physes. 90 Fig. 1A–D Chondroblastoma of the first metatarsal bone in a 24-year-old male patient. A Radiography depicts an ex- pansile, sharply delineated le- sion (Lodwick grade 1B; black arrowhead). B Sagittal, fat- suppressed, T2-weighted MR image depicts a lesion with a solid and a macrocystic compo- nent. C Sagittal T1-weighted MR image before (above) and after (below)intravenous gado- linium application shows mod- erate and homogeneous en- hancement of the solid portion. D Pseudoepithelial lining of the aneurysmal bone cyst compo- nent (black arrows). Typical matrix calcification and chon- droblastic tumor cells (hema- toxylin-eosin (HE), ×10) Radiography seal, two were metadiaphyseal (Fig. 4A), two were epiphyseal with extension into the growth plate and two Chondroblastoma lesions were in the axial skeleton (pubic and maxillary Two chondroblastomas were strictly confined to the bones). The size varied from 3 to 7.5 cm (mean 5.2 cm).