Differentiation of osteoblastic metastases and bone islands on dual-energy computed tomography in patients with untreated lung cancer

Abstract

Objective: To evaluate the diagnostic efficacy of quantitative dual-energy computed tomography (CT) parameters for distinguishing osteoblastic metastases (OBMs) from bone islands (BIs) in untreated lung cancer.

Material and methods: Dual-energy CT images of 24 patients with OBMs and 56 patients with BIs obtained between January 2019 and December 2021 were retrospectively analyzed. The CT_70keV value, calcium_(water) density [Dcalcium_(water)], and water_(calcium) density [Dwater_(calcium)] were analyzed. Diagnostic performance was assessed by measuring the area under the curve (AUC), and specificity, sensitivity, and accuracy were determined.

Results: A total of 70 OBMs and 67 BIs were included. The AUC values of CT_70keV, Dcalcium_(water), and Dwater_(calcium) showed no significant differences (0.950 vs. 0.947 vs. 0.929, respectively; P > 0.05). The optimal CT_70keV cutoff value was 885.1 HU, with specificity, sensitivity, and accuracy of 81.4 %, 92.5 %, and 86.9 %, respectively. When using Dcalcium(water) < 254.9 mg/cm³ and Dwater(calcium) < 1250.6 mg/cm³, respectively, 119 of 137 lesions showed consistent diagnostic results (true or false). Sub-analysis of these 119 lesions showed specificity of 92.1 %, which was higher than that of CT_70keV (P = 0.021). The AUC, sensitivity, and accuracy were 0.974, 92.9 %, and 92.4 %, respectively, which were not significantly different from those of CT_70keV (P = 0.230, 0.906, and 0.220, respectively). Among the 18 lesions showing inconsistent diagnoses, Dcalcium_(water) diagnosed 11 lesions correctly, and Dwater_(calcium) diagnosed the remaining seven lesions correctly.

Conclusion: The combination of Dcalcium_(water) and Dwater_(calcium) demonstrated a promising role in the differentiation of OBMs from BIs in lung cancer patients.

Keywords: Bone island; Computed tomography; Dual energy; Metastasis; Osteoblast.