Magnetic resonance imaging features of intra-articular tenosynovial giant cell tumor in children

• Published in: Pediatric radiology 2021-03, Vol.51 (3), p.441-449
• Main Authors: Nguyen, Jie C., Biko, David M., Nguyen, Michael K., Othman, Sammy, Weber, Kristy L., Ganley, Theodore J., Arkader, Alexandre
• Format: Article
• Language: English
• Subjects: Ankle; Children; Chondromalacia; Effusion; Imaging; Joint diseases; Joints (anatomy); Knee; Magnetic resonance imaging; Medical imaging; Medicine; Medicine & Public Health; Neuroradiology; Nuclear Medicine; Oncology; Original Article; Pediatrics; Radiology; Tumors; Ultrasound
• ISSN: 0301-0449; 1432-1998
• DOI: 10.1007/s00247-020-04861-4

Background Magnetic resonance imaging (MRI) plays a critical role in disease characterization of intra-articular tenosynovial giant cell tumor. Objective To characterize the MRI features of intra-articular tenosynovial giant cell tumor in children with respect to disease subtype and anatomical location. Materials and methods This retrospective study included children with tenosynovial giant cell tumor who underwent preoperative MRI between January 2006 and May 2020. Two radiologists reviewed each examination to determine disease subtype, signal intensities and the presence of an effusion, osseous changes, chondromalacia, juxtacapsular disease and concomitant joint involvement. Fisher exact, Mann-Whitney U , and Kruskal-Wallis H tests were used to compare findings between subtypes and locations. Results Twenty-four children (16 girls, 8 boys; mean age: 13.1±3.8 years) with 19 knee and 5 ankle-hindfoot tenosynovial giant cell tumor had either diffuse ( n =15) or localized ( n =9) disease. An effusion ( P =0.004) was significantly more common with diffuse than localized disease. There was no significant difference in MRI signal ( P -range: 0.09–1) or other imaging findings ( P -range: 0.12–0.67) between subtypes. Children with knee involvement were significantly more likely to present with diffuse disease while those with ankle-hindfoot involvement all presented with focal disease ( P =0.004). Juxtacapsular ( n =4) and concomitant proximal tibiofibular joint involvement ( n =5) were observed with diffuse disease in the knee. Erosions ( P =0.01) were significantly more common in the ankle than in the knee. Conclusion In our study, diffuse tenosynovial giant cell tumor was more common than localized disease, particularly in the knee where juxtacapsular and concomitant proximal tibiofibular joint disease can occur.