Objective: To evaluate the accuracy of thallium-201 (201TI) scintigraphy in distinguishing a benign from a malignant recent non-traumatic vertebral fracture.
Methods: STUDY DESIGN--Single center, prospective study. PARTICIPANTS--Patients hospitalized for a recent non-traumatic vertebral fracture. EVALUATION--Usual clinical, laboratory and radiological assessment; 201TI vertebral scintigraphy: patients were injected with iv 3 mCi 201TI. Early and delayed images of the fractured vertebra were obtained. DATA ANALYSIS--(1) Two examinators, unaware of the other findings, rated the images as hyperfixation or not of the fractured vertebra; (2) the ratio (average count per pixel of the fractured vertebra/normal adjacent vertebrae) were calculated. The FINAL DIAGNOSIS was established on the result of vertebral biopsy or on follow-up.
Results: Twenty-one patients were included. The final diagnosis was a benign vertebral fracture in 14 patients and a malignant vertebral fracture in 7. The sensitivity, specificity, positive and negative predictive values for a malignant fracture on early 201TI vertebral scintigraphy images were 28.6, 92.9, 66.6, and 72.2%, respectively, and on delayed images were 28.6, 100, 100, and 73.7%, respectively. The ratio of lesioned over normal tissue was not increased in malignant, compared with benign fractures.
Conclusion: The weak sensitivity does not support the wide use of 201TI bone scintigraphy to distinguish a benign from a malignant recent non traumatic vertebral fracture. However, the high specificity suggests that such evaluation might be proposed prior to vertebral biopsy in some difficult cases.