Background Vertebral fractures (VFs) are the hallmark of osteoporosis and are more predictive of future fracture than areal Bone Mineral Density (BMD). Number and severity of VFs are related to bone microarchitecture deterioration. Trabecular Bone Score (TBS), derived from the texture of the DXA image, has been shown to be related to bone microarchitecture and fracture risk. Our hypothesis is that TBS measurement could be related to the severity of VFs.
Objectives to evaluate performance of TBS, alone or added to aBMD, in the prediction of the presence of VFs and of the severity of these fractures.
Methods Patients were selected from the Fracture Liaison Service (FLS) of our department, aiming at providing assessment of osteoporosis to patients over the age of 50 years who had sustained low trauma fractures and who are hospitalized in the Orthopaedic surgery department. aBMD and Vertebral Fracture Assessment (VFA) were performed one week to 3 months after the fracture using DXA. VFs were classified using the Genant’s semiquantitative evaluation and severity was assessed using the spinal deformity index (SDI), i.e. the sum of number and grades of VFs. TBS was obtained after re-analysis of DXA lumbar spine (L2-L4) scans. Performance of TBS, BMD and their combination was assessed using Receiver operator characteristic (ROC) and areas under receiver operating characteristics curves (AUCs).
Results Data from 528 patients over 50 years with a non vertebral fragility fracture were examined between February 2009 and October 2012. VFA and TBS were not performed in 166 of them due to technical reasons. 362 patients (77.3% women; mean age 74.3±11.7 years) were analysed; 182 (50.3%) had hip fractures and 49 (13.5%) received an anti-osteoporotic treatment. Prevalence of VFs by VFA was 36.7%; 189 (52.2%) patients were osteoporotic (T score≤-2.5 at at least one site). TBS was lower in the patients with VFs than in patients without VFs in the whole population (1.16 ±0.11 vs 1.23±0.11, p<0.0001) and in non osteoporotic patients (1.19 ±0.12 vs 1.25±0.10, p=0.001). In the whole population, performance of TBS (AUC=0.677) was similar to lumbar spine (LS) BMD (AUC= 0.669) and hip BMD (AUC= 0.692) for the identification of VFs. However combination of TBS and LS BMD improves the discrimination as compared to LS BMD alone (AUC=0.707, p=0.043). In the non osteoporotic population (n=173), AUC of TBS for the discrimination of VFs was higher than AUC of LS BMD (0.67 0vs 0.541, p=0.035). There was a negative correlation between TBS and SDI: r= -0.31 (p<0.0001).
Conclusions Our study suggests that about 40% of patients with a non vertebral fracture have vertebral fractures that were not previously diagnosed. TBS is able to discriminate patients with vertebral fractures, and adds information on LS aBMD. TBS is correlated to the number and severity of vertebral fractures evaluated by SDI. TBS measurement may be useful to identify subjects with non vertebral fracture requiring spine imaging.
Disclosure of Interest None Declared
Statistics from Altmetric.com
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.