METHODS: A total of 34 patients (11 with ARONJ, 23 without ARONJ) who were administered anti-resorptive-agents for bone metastasis and had incidentally consulted a dental surgeon within 3 months after regular whole-body bone scintigraphy were retrospectively included in the study. The bone scintigraphy data were subjected to semiquantitative analysis of uptake in the jaw using BONENAVI (FUJIFILM Toyama Chemical, Co. Ltd., Tokyo, Japan; EXINI Diagnostics AB, Lund, Sweden) and BUV software (Technical Society for Quantitative Bone Scintigraphy and Fujifilm Toyama Chemical Co., Ltd. Tokyo, Japan). The ROI was set semi-automatically on mandibular hotspots, and the regional BSI was termed BSIJ. Planar anterior and posterior images were then sent to BUV software, with the ROI set manually as for BSI, and the regional BUV was termed BUVJ.
RESULTS: Mean BSIJ values for the ARONJ positive and ARONJ negative groups were 0.17 ± 0.83 and 0.03 ± 0.50%, respectively. Mean BUVJ values for the ARONJ positive and ARONJ negative groups were 0.47 ± 0.17 and 0.19 ± 0.11, respectively. BSIJ × BUVJ values for the ARONJ positive versus ARONJ negative groups were 0.088 ± 0.067 vs. 0.007 ± 0.010. The AUC for BSIJ, BUVJ and BSIJ × BUVJ was 0.949, 0.951 and 0.988, respectively.
CONCLUSIONS: The BSI metric of a CAD system for bone scintigraphy was useful for the early detection of ARONJ. Accuracy was improved with the additional use of BUVJ data. We recommend that SPECT imaging be performed when bone scintigraphy reveals focal or diffuse uptake in the mandible with high BSIJ and BUVJ.