Diagnosis of Vertebral Fractures by Vertebral Fracture Assessment

Abstract

Vertebral fractures are independent risk factors for both vertebral and peripheral fractures and only one-third of these fractures come to clinical attention. Vertebral fracture assessment (VFA) is a radiographic method using dual X-ray absorptiometry (DXA) to assess vertebral deformities during bone density measurement. We performed VFA of the spine from T4 to L5 on a Delphi W device (Hologic, Bedford, MA) in 136 postmenopausal patients (69 ± 10 yr). These patients also had X-rays of the thoracic and lumbar spine. VFA was independently compared with X-rays by two rheumatologists, for the diagnosis of vertebral fractures at both the patient and vertebral levels. Using X-rays, 61 patients (45%) had at least one vertebral fracture. The percentage of unreadable vertebrae was 1% and 12.4% on X-rays and VFA, respectively (p < 0.0001). At the patient level, VFA allowed to diagnose if the patient had no fracture or had at least one fracture in 74% of patients. In 11.2% of cases, VFA misclassified the patients. At the vertebral level, diagnostic efficacy of VFA as compared with X-rays was 97%. Concordance between both observers was good (κ-score = 0.69). We designed an algorithm for decision of performing X-rays in postmenopausal women: Using results of VFA would avoid X-rays in 32% of our patients. VFA is a reliable technique with low radiation, and is easily and rapidly applicable during bone density measurement by DXA, which could improve management of osteoporotic patients.

Introduction

Vertebral fractures are the hallmark manifestation of osteoporosis, associated with back pain, increased morbidity, and functional limitations (1). Postmenopausal women with vertebral fractures are at higher risk of both vertebral and peripheral fractures, including hip fractures (2). Risk of vertebral fractures in women with one prevalent fracture is twice that of women without a prevalent fracture (1). Prevalent vertebral fractures are associated with increased mortality (3), probably related to comorbidity; the association of clinical vertebral fractures with the increase in mortality is similar to the one observed after hip fracture among older women (4). Efficient treatments decreasing by 50%, on average, the risk of incident fracture in patients having prevalent vertebral fractures are available. Postmenopausal women with prevalent vertebral fractures should be recognized in order to prevent the debilitating consequences of subsequent fractures 5, 6. However, only one-third of vertebral fractures come to clinical attention, even if they are present on X-rays (7). There is no single clinical sign, or combination of clinical risk factors, able to predict existing vertebral fracture (8).

Performing systematic spine X-rays in asymptomatic osteoporotic patients exposes these patients to radiation and would increase the cost of diagnosis of vertebral fractures. A possible way to check for these asymptomatic vertebral fractures is available through dual energy X-ray absorptiometry (DXA) devices. This technique has several advantages: it delivers a lower dose of radiation to the patient (entrance dose = 70 μSv, constructor data) than spine X-rays (2100 μSv effective dose) 9, 10, and the X-ray source is in theory always orthogonal to the vertebral bodies, avoiding geometric distortion. The main limitation of the technique is lack of visualization of the upper thoracic spine (above T7) 11, 12, 13, 14, 15, 16, 17, 18.

Quantitative measurements of the vertebral bodies have been used. They are time-consuming and not applicable in clinical settings. They are, therefore, used only in clinical research studies. Qualitative or semi-quantitative evaluation of the vertebrae could easily be used in clinical practice. DXA devices of recent generations have better image quality, allowing both a semi-quantitative and a quantitative evaluation of the vertebrae using visual assessment (vertebral fracture assessment [VFA]). This technique seems promising and its indications remain to be standardized.

The aim of our study was to evaluate the sensitivity and specificity of this technique (VFA) for diagnosis of vertebral fractures in postmenopausal women and to design an algorithm using this technique to avoid unnecessary X-rays in this population.