Intravertebral vacuum phenomenon following fractures: CT study on frequency and etiology

PURPOSE: The purpose of this work was to determine the frequency and etiology of the intravertebral vacuum phenomenon (IVP). METHOD: CT examinations of 96 vertebral fractures were evaluated for IVP. Bone mineral density (BMD) was determined in nonfractured vertebrae. For calibration purposes, densities of a standard phantom measured in 30 patients who underwent quantitative CT examinations in the same time period were used and precision was calculated. RESULTS: Eleven of 96 fractures (11.5%) showed IVP. IVP was present in 4 of 20 fractures at T12 (20%), in 4 of 23 at L1 (17.4%), and 1 IVP was found at L2-4 each. Mean +/- SD age of patients with IVP was 68.3 +/- 10.5 years and without 47.8 +/- 19.4 years (p < 0.001). Mean +/- SD density of nonaffected vertebra was 45.9 +/- 17.0 mg of hydroxyapatite/ml for patients with IVP and 139.5 +/- 62.6 mg/ml for those without IVP (p < 0.0005). An average precision of 1.2% was calculated for the density measurements over the investigated time. CONCLUSION: Following vertebral fractures, IVP on CT scans is more common than presumed and increases with age. There exists a significant inverse correlation between the BMD and the frequency of IVP.