[Vertebral morphometry: evaluation of osteoporosis-caused fractures]

To compare visual reading of spine radiographs and quantitative morphometric approach for assessing the prevalence of vertebral fractures in postmenopausal osteoporosis.

In 473 postmenopausal women afferent to our Centre of Osteoporosis under-went lateral thoracic and lumbar spine radiograph to identify vertebral fractures and dual energy X-ray absorptiometry (DXA) to measure bone mineral density (BMD) of the lumbar spine (L1-L4). Osteoporosis was defined according to the World Health Organization (WHO) guidelines. To identify vertebral fractures the radiographs were visually analyzed by two radiologists; a woman was judged as fractured only if both readers independently found at least one vertebral fracture on her films. Then the spine radiographs were digitized by means of a scanner to perform quantitative vertebral morphometry (QVM) using specific software. An expert operator manually located the calipers on the vertebral bodies from T4 to L5 and the computer automatically calculated the anterior, middle and posterior vertebral heights and their ratios. A vertebral fracture was defined by morphometry as a reduction by at least 20%, with an absolute decrease of at least 4 mm, in one of three height ratios of any vertebral body compared to the corresponding reference ratio for fertile women.

Visual reading by radiologists detected 9.5% (45/473) women with vertebral fractures and QVM detected 13.7% (65/473) with statistical significance (p < 0.001). In the 75-80 years age group the prevalence of vertebral fractures reaches the maximum value, 26.3% by visual reading and 36.8% by QVM. Among fractured women, 34 were osteoporotic by DXA; 11 women found fractured by visual reading and 21 by QVM were osteopenic women, with bone mineral densities between -1 and -2.5 SD of the T-score.

This study showed that quantitative assessment of spine radiographs by vertebral morphometry is an objective method that allows to identify a larger number of vertebral fractures compared to visual inspection. This is very important not only for epidemiological studies, but also for clinical use because a previous vertebral fracture increases the risk of subsequent fractures significantly. Therefore, to improve the risk assessment of vertebral fractures for osteoporotic patients it is necessary to combine the use of QVM and BMD.