Radiographic methods for evaluating osteoporotic vertebral fractures.

UNLABELLED: Reproducible methods for the radiological assessment of osteoporotic vertebral fractures, defined based on accurate criteria, are needed in everyday practice and in therapeutic trials and epidemiological studies.
OBJECTIVES: To describe and to evaluate methods for osteoporotic vertebral fracture assessment based on standard radiographs or dual-energy X-ray absorptiometry (DXA) and to determine the role for each method in clinical practice, therapeutic trials, and epidemiological studies.
METHODS: A review written by a rheumatologist based on his clinical experience and on a literature review was submitted to four experts. Studies in English or French published between 1975 and February 2008 were retrieved from Medline using the keywords vertebral fracture, osteoporosis, vertebral deformity, and vertebral fracture assessment.
RESULTS: One hundred forty-nine articles were selected and read in their full-text version. There was no consensus regarding the definition of osteoporotic vertebral fractures. The following methods were evaluated: visual assessment, Genant's semi-quantitative assessment, Jiang's algorithm-based qualitative method, morphometric radiography, and DXA of the spine. In everyday practice, Genant's semi-quantitative assessment on standard radiographs may provide useful information on the severity and prognosis of osteoporosis. DXA done for bone mineral density measurement may detect vertebral fractures in asymptomatic patients. Assessment of standard radiographs remains the reference standard for diagnosing vertebral fractures in patients with suggestive symptoms (e.g., pain in the thoracic or lumbar spine, height loss, or thoracic kyphosis). For therapeutic trials and epidemiological studies, Genant's semi-quantitative assessment used by a trained and experienced observer is the preferred method, based on its good reproducibility and ability to differentiate fractures from other deformities. However, thousands of radiographs may be needed, making routine interpretation by an expert impractical. A visual semi-quantitative method may be used to separate normal radiographs from radiographs showing possible or obvious fractures, which can then be read by an expert. Alternatively, radiomorphometric indices can be determined on digitized radiographs in combination with a semi-quantitative assessment, with discordant cases being reviewed by an expert. We do not recommend Jiang's method at present, as it is still undergoing validation.