Abstract
This study aimed to investigate the diagnostic performance of chest computed tomography (CT) for opportunistic screening and longitudinal follow-up of osteoporosis in breast cancer patients, compared to dual-energy X-ray absorptiometry (DXA). The association between L1 vertebral attenuation on chest CT and incidental fracture was also evaluated. We retrospectively reviewed 414 consecutive breast cancer patients who underwent both non-enhanced chest CT and DXA within a 3-month interval and had at least two DXA and two chest CT examinations over more than 1 year. The attenuation value of the L1 trabecular bone was measured on an axial CT image and compared to the corresponding DXA T-score. The diagnostic performance of L1 vertebral attenuation on CT for osteoporosis was calculated at different thresholds (90 HU, 100 HU, 110 HU), and the correlation between L1 vertebral attenuation values and DXA T-scores was statistically analyzed. Overall fracture-free survival was estimated and compared with the threshold of 90 HU on CT and -2.5 T-score on DXA. Of 414 patients (median age, 53.0 years), 88 (21.3%) had either vertebral or non-vertebral fractures. The median follow-up duration between initial and final DXA was 902.9 days. There was a moderate correlation between L1 vertebral attenuation value and DXA T-score (ρ = 0.684, CI 0.653-0.712). Fracture-free survival was significantly lower in patients with attenuation values ≤90 HU on CT and T-scores ≤-2.5 on DXA (P < .001). Multivariate analysis revealed that attenuation values ≤90 HU on CT (P < .001), T-scores ≤-2.5 on DXA (P = .003), and age ≥65 years (P = .03) were independent significant prognostic factors associated with overall fracture-free survival. The sensitivities and specificities of L1 attenuation value were 54.9% and 85.8% at 90-HU threshold, 74.0% and78.4% at 100-HU threshold, and 83.9% and 70.1% at 110-HU threshold, respectively. In conclusion, CT can be used for predicting osteoporosis and discriminating incidental fracture risk in breast cancer patients.