Differentiation of osteolytic metastases and Schmorl's nodes in cancer patients using dual-energy CT: Advantage of spectral CT imaging

Purpose: To assess the reliability of dual-energy CT (DECT) spectral imaging for the differentiation of bone metastases (BMs) from Schmorl's nodes (SNs) in the vertebrae of cancer patients. Materials and methods: In this retrospective study, 102 cancer patients who underwent DECT (GE spectral CT Discovery CT750 HD scanner) had 110 low density vertebral lesions. Each lesion was characterized as a BM or SN, based on the typical MRI or SPECT/PET-CT findings as well as size and number change in the 6 months follow-up. The means of 140 kVp polychromatic CT values, 40-140 keV monochromatic CT values, slopes (k) of the spectral curves, bone(water) and water(bone) densities of BMs and SNs were measured and compared with independent-samples t-test. The difference values of the two lesions and their respective normal bone tissue were calculated (normal density - lesion density) and compared using independent-samples t-test. ROC curves were used to compare the diagnostic efficacies of these measures in the identification of SNs and BMs. Results: 110 lesions consisting of 69 BMs and 41 SNs were identified. The spectral curve patterns and slopes for BMs and SNs were different (p < 0.05). The water(bone) density of BMs (1009.02 +/- 59.25 mg/cm3) was higher than that of SNs (892.00 +/- 83.65 mg/cm(3)) (p < 0.01) while the bone(water) density (43.57 +/- 50.87 mg/cm3) was lower than that of SNs (174.60 +/- 94.61 mg/cm3) (p < 0.01). The 40 keV CT value, k, bone(water) density and water(bone) density had a higher diagnostic efficacy for differentiating the two lesions than polychromatic CT value (p <0.05). Conclusions: Dual-energy CT imaging is accurate enough for identification of osteolytic metastases and Schmorl's nodes. (C) 2014 Published by Elsevier Ireland Ltd.