Purpose Localized tenosynovial giant cell tumors (GCTTS), also known as giant cell tumors of tendon sheath, are benign lesions that represent the second most common tumor of the hand. The standard treatment for GCTTS is marginal excision; however, recurrence is common, with rates of 4% to 56% reported in the literature. This study aimed to identify clinical, imaging, and surgical factors associated with recurrence following surgical excision of localized GCTTS. Methods In this retrospective cohort study, following institutional review board approval, we analyzed 74 patients with histologically confirmed localized GCTTS treated at a single academic medical center between 2001 and 2024. The mean follow-up was 7.4 years. All surgeries were performed by fellowship-trained hand surgeons using marginal excision with surgical adjuncts including extensile incisions, loupe magnification, tourniquet control, and secondary survey of the surgical bed. Patients were divided into nonrecurrent (n = 63) and recurrent (n = 11) cohorts. Fifty-five clinical parameters were collected from medical records and patient interviews. Statistical analyses were performed using Fisher exact tests to compare clinical parameters between the nonrecurrent and recurrent cohorts. Results The overall recurrence rate was 15% with a mean follow-up of 7.4 years. Three statistically significant predictors of recurrence were identified: (1) Bone involvement was present in 100% of recurrent cases versus 47.6% of nonrecurrent cases, representing the strongest predictor. Although bone involvement demonstrated high sensitivity (100%), it had low specificity (52.4%). (2) Neurovascular bundle involvement (63.6% vs 25.4%) and (3) presence of satellite lesions (18.2% vs. 0%) were also identified; however, this last finding is based on only two cases and should be interpreted with caution as a preliminary observation rather than a confirmed predictor. Joint involvement (81.8% vs 58.7%) and flexor tendon involvement (45.5% vs 23.8%) demonstrated trends toward recurrence without reaching statistical significance. Tumor size, duration of symptoms, pseudocapsule integrity, surgeon's experience, and osteoarthritis were not associated with recurrence. The wide confidence intervals reflect the limited precision inherent in the small sample size. Conclusions Bone involvement and neurovascular bundle involvement are significant predictors of recurrence in localized GCTTS. The exclusive occurrence of satellite lesions in the recurrent group suggests greater vigilance for satellite lesions during surgery; however, the sample is insufficient to establish a definitive predictive relationship. The relatively low recurrence rate (15%) achieved in this series likely reflects the standardized surgical approach with meticulous technique. Notably, the high nonrecurrence rate (73.2%) among patients with bone involvement suggests that aggressive curettage as part of a standardized protocol can substantially mitigate recurrence risk even in these technically challenging cases. Preoperative magnetic resonance imaging is valuable for identifying these high-risk features and guiding surgical planning. Until additional prognostic markers become available, meticulous surgical technique with attention to identified risk factors remains the cornerstone of preventing recurrence in localized tenosynovial giant cell tumors. Patients with these high-risk factors of recurrence should also be counseled about recurrence and followed at least 5 years. Type of study/level of evidence Prognostic III.
Hurst et al. (Wed,) studied this question.