Prediction of immediate and long-term benefit after kyphoplasty of painful osteoporotic vertebral fractures by preoperative MRI

Background It is unclear if an MR-detectable bone marrow edema is a prerequisite for pain reduction and morphological correction by kyphoplasty. This comparative trial evaluates clinical and radiomorphological outcomes after kyphoplasty of painful osteoporotic vertebral fractures with and without preoperative MR-detectable bone marrow edema for 1 year of follow-up. Methods Preoperative MR-images of 45 patients who received kyphoplasty for treatment of painful osteoporotic vertebral fractures were evaluated with regard to presence (n = 27) or absence (n = 18) of vertebral bone marrow edema. Pain scores (VAS 0-100) and radiomorphological measures (midline vertebral height, kyphosis angle) were analysed at baseline, postoperatively and after 12 months. Results In the "bone edema" group, pain scores improved from 72.7 to 46.8 (postoperative) and 48.0 (12 months, P < 0.001, both). In the group without preoperative bone edema, pain score improved from 70.7 to 60.3 (postoperative, P = 0.013) and to 50.1 (12 months, P = 0.001). Pain scores of both groups were significantly different directly postoperative (P = 0.026), but not after 12 months (P = 0.714). Vertebral height restoration was slightly greater in the "bone edema" group (10.2% vs. 7.8%, P = 0.289). Correction of the kyphosis angle was greater in the "bone edema" group (P = 0.014) compared to the "no bone edema" group (P = 0.838). Conclusion A preoperative MR-detectable vertebral bone marrow edema predicts a better short-term outcome after kyphoplasty, but is not a prerequisite for long-term pain reduction in patients with old, chronically painful osteoporotic vertebral fractures.