부산대학교 도서관

Objective: To determine the value of CT and dynamic contrast-enhanced (DCE-)MRI for monitoring denosumab therapy of giant cell tumors of bone (GCTB) by correlating it to histopathology. Materials and methods: Patients with GCTB under denosumab treatment and monitored with CT and (DCE-)MRI (2012-2021) were retrospectively included. Imaging and (semi-)quantitative measurements were used to assess response/relapse. Tissue samples were analyzed using computerized segmentation for vascularization and number of neoplastic and giant cells. Pearson's correlation/Spearman's rank coefficient and Kruskal-Wallis tests were used to assess correlations between histopathology and radiology. Results: Six patients (28 ± 8years; five men) were evaluated. On CT, good responders showed progressive re-ossification (+7.8HU/month) and cortical remodeling (woven bone). MRI showed an SI decrease relative to muscle on T1-weighted (−0.01 A.U./month) and on fat-saturated T2-weighted sequences (−0.03 A.U./month). Time-intensity-curves evolved from a type IV with high first pass, high amplitude, and steep wash-out to a slow type II. An increase in time-to-peak (+100%) and a decrease in Ktrans (−71%) were observed. This is consistent with microscopic examination, showing a decrease of giant cells (−76%), neoplastic cells (-63%), and blood vessels (−28%). There was a strong statistical significant inverse correlation between time-to-peak and microvessel density (ρ = −0.9, p = 0.01). Significantly less neoplastic (p = 0.03) and giant cells (p = 0.04) were found with a time-intensity curve type II, compared to a type IV. Two patients showed relapse after initial good response when stopping denosumab. Inverse imaging and pathological findings were observed. Conclusion: CT and (DCE-)MRI show a good correlation with pathology and allow adequate evaluation of response to denosumab and detection of therapy failure. [ABSTRACT FROM AUTHOR]