The radiological assessment of vertebral osteoporosis

The traditional skeletal X-ray is of little help in quantifying osteopenia in the spine, and indirect criteria, such as increased lucency, loss of horizontal trabeculae and reduction in end-plate thickness with relatively increased opaqueness, can be misleading. However, the clinical expression of osteoporosis is fracture, which can be identified by radiography. Any reduction in the anterior, middle, or total height of the vertebral body should be classified as vertebral fracture. In mild cases, such fractures are not easily detectable, particularly when previous radiographs are not available for comparison. Thus several objective methods for fracture identification have been developed. Most of these methods are based on the ratio between posterior and anterior or middle height (wedge or biconcave deformity, respectively) and on the ratio between posterior height of adjacent vertebrae or with a single vertebral reference (T4). Some of these indices are referred to normal ranges in order to take into account intervertebral and interindividual variability. In 36 women with postmenopausal osteoporosis we have compared the lateral radiographs of lumbar and thoracic spine to similar X-ray pictures taken by chance at least five years before menopause. By defining a fracture as any decrease in vertical height above 1 mm, we found 77 deforming events in 29 out of 36 patients. We then applied some of the methods to identify objectively fractures in our postmenopausal radiographs: the sensitivity of the various systems ranged from 50 to 100%; however there was a large overlap between false positives and false negatives and the methods with the highest sensitivity lack specificity and vice versa.