Background and objectives: Literature on surveillance for lung metastasis from giant cell tumor of bone (GCTB) is scarce. We aimed to develop one by determining: (1) the optimal surveillance schedule by analyzing time-to-event data, taking into account the predictive factors, and (2) the effective diagnostic modality.
Methods: A total of 333 patients who underwent surgery for GCTB were followed for at least 2 years. All had chest radiography, and 169 had additional CT for surveillance. Time to lung metastasis and cumulative incidence were calculated, and diagnostic performance between chest radiography and CT was compared.
Results: Twenty-five (7.5%) of 333 patients developed lung metastasis, and local recurrence (LR) was the only predictive factor (RR = 6.54). Median interval from LR to metastasis was 15 months, and 17 (85%) of the 20 metastases with LR occurred within 3 years of LR. Cumulative post-LR incidences at 1, 3, and 5 years were 15.4%, 21.5%, and 21.5%, respectively. CT was more sensitive (100% vs 32%), and had higher positive predictive value (81% vs 57%) and accuracy (96% vs 93%).
Conclusions: Intensified lung surveillance is warranted for GCTB patients with LR, especially for 3 years from diagnosis of LR. CT is effective for detecting lung metastasis from GCTB.
Keywords: chest CT; chest radiography; giant cell tumor; local recurrence; lung metastasis; surveillance guideline.
© 2017 Wiley Periodicals, Inc.