RETURN TO ABSTRACT LISTING


Objective:
The purpose of this study was to describe the MR imaging features of malignant tenosynovial giant cell tumors (TSGCTs) and to assess imaging characteristics that may be helpful in differentiating these lesions from benign TSGCTs.

Materials and Methods:
We conducted an IRB-approved retrospective review of the MR imaging characteristics of pathologically proven malignant and benign TSGCTs. Our search of the institutional pathology and radiology databases yielded 10 pathologically proven cases of malignant TSGCT and 23 cases of benign localized TSGCT with MR imaging. The MRIs were reviewed by a radiology resident and board-certified fellowship trained musculoskeletal radiologist in consensus for multiple MRI features including size, T1 and T2 signal characteristics, margins, location, perilesional vessels and edema, and degree/pattern of enhancement.

Results:
There was no significant difference in age between the benign and malignant TSGCT groups (43.7 +/- 18.7 versus 51.6 +/- 17.2 years, respectively; p = 0.26) or gender distribution (13 women/10 men in the benign cohort and 4 women/6 men in the malignant cohort). There was no significant difference in MRI signal characteristics between benign and malignant lesions (T1 signal intensity, p = 1.0; T2 signal intensity, p = 0.39; heterogeneous T2 signal, p = 1.0; presence of low T2 signal, p = 0.52; presence of perilesional edema, p = 0.26; blooming, p = 1.0). There was no difference in whether gadolinium was administered (p = 0.40) or the enhancement pattern (heterogeneous vs. homogeneous, p = 0.82; degree of enhancement, p = 0.13; presence of internal non-enhancing areas, p = 0.11). Malignant TSGCTs had a larger average maximal dimension compared to benign TSGCTs (8.1 +/- 7.9 cm versus 2.8 +/- 2.2 cm, p = 0.0051). Malignant TSGCTs were more likely to have irregular margins (6/10 versus 1/23, p = 0.0012), prominent perilesional vessels (6/10 versus 1/23, p = 0.0012), and be extra-articular (8/10 versus 7/23, p = 0.02). Interestingly, 3/10 malignant TSGCTs were subcutaneous and not clearly associated with a synovial lined structure (joint, tendon sheath, or bursa).

Conclusion:
This study is the largest series to date examining MRI imaging characteristics of malignant TSGCTs. We found no difference in T1 signal, T2 signal, enhancement pattern, blooming, or perilesional edema between benign and malignant TSGCTs. Malignant TSGCTs were more likely to be larger, extra-articular, have irregular margins, and prominent perilesional vessels. Additionally, malignant TSGCTs may not be clearly associated with a synovial-lined structure.