Use of FDG-PET in differentiating benign from malignant compression fractures

Abstract

Objective: The objective was to evaluate the use of fluorodeoxyglucose positron emission tomography (FDG-PET) in differentiating benign from malignant compression fractures.

Patients and methods: In a retrospective analysis, we identified 33 patients with 43 compression fractures who underwent FDG-PET. On FDG-PET the uptake pattern was recorded qualitatively and semiquantitatively and fractures were categorized as benign or malignant. Standardized uptake values (SUV) were obtained. MRI, CT, and biopsy results as well as clinical follow-up for 1-3 years served as standards of reference. The Student's t test was used to determine whether there was a statistically significant difference between the SUV for benign and malignant compression fractures.

Results: There were 14 malignant and 29 benign compression fractures, including 5 acute benign fractures. On FDG-PET, 5 benign fractures were falsely classified as malignant (false-positive). Three of these patients underwent prior treatment with bone marrow-stimulating agents. There were two false-negative results. Sensitivity, specificity, positive predictive value, negative predictive value, and accuracy of FDG-PET in differentiating benign from malignant compression fractures were 86%, 83%, 84%, 71%, and 92% respectively. The difference between SUV values of benign and malignant fractures was statistically significant (1.9 +/- 0.97 for benign and 3.9 +/- 1.52 for malignant fractures, p < 0.001). SUV of benign acute and chronic fractures were not statistically significant.

Conclusion: Fluorodeoxyglucose positron emission tomography is useful in differentiating benign from malignant compression fractures. Therapy with bone marrow-stimulating agents can mimic malignant involvement.

Fig. 1

a Malignant compression fracture in a patient with lymphoma. Sagittal FDG-PET image demonstrates intense radiotracer uptake in the L1 vertebral body (SUV = 7.1; arrow). b Axial FDG-PET image demonstrates tumor infiltration of the L1 vertebral body with associated extra-osseous component (arrows). c Sagittal T1-weighted MRI demonstrates diffuse low signal intensity of the bone marrow of the lumbar spine with compression fracture of L1 (arrow). d Sagittal fat-suppressed T1-weighted MRI following the administration of gadolinium demonstrates diffuse enhancement of the L1 compression fracture (arrow). e Axial image from a CT-guided biopsy demonstrates large extra-osseous soft tissue component (white arrow) and infiltration of the L1 vertebral body (black arrows)

Fig. 2

a Malignant compression fracture from metastatic colon cancer. Sagittal fused FDG-PET/CT image demonstrates intense radiotracer uptake in the T7 compression fracture (SUV = 6.3; arrow). b Axial fused FDG-PET/CT image demonstrates focal increased radiotracer uptake in the T7 vertebral body (arrow). c Sagittal T1-weighted MRI demonstrates compression fracture of T7 with low signal marrow infiltration (arrow). d Sagittal STIR image demonstrates mild hyperintensity of the T7 vertebral compression fracture (arrow). e Sagittal fat-suppressed T1-weighted MRI following the administration of gadolinium demonstrates diffuse enhancement of the T7 compression fracture (arrow). Malignancy was indeterminate on MRI

Fig. 3

Benign compression fracture in a patient with lymphoma. Sagittal fused FDG-PET/CT image demonstrates compression fracture of T9 (arrow) without significant radiotracer uptake (SUV = 0.7)

Fig. 4

a Benign subacute compression fractures in a patient without history of malignancy. Sagittal FDG-PET image demonstrates mildly increased radiotracer uptake in the upper thoracic spine (SUV = 1.8; arrow). b Axial FDG-PET image at the level of T4 demonstrates mildly increased radiotracer uptake (arrow), consistent with subacute compression fracture. c Sagittal CT image demonstrates compression fracture of T4 (arrow) and multiple chronic osteoporotic compression fractures throughout the thoracic spine

Fig. 5

a False-negative compression fracture on FDG-PET in a patient with metastatic esophageal cancer. Sagittal fused FDG-PET/CT image demonstrates mild to moderately increased radiotracer uptake in the L2 compression fracture (arrow; SUV = 2.5) that was thought to represent a benign compression fracture but was found to be malignant on subsequent biopsy. Note the mildly increased radiotracer uptake in the T11 compression fracture (arrowhead), which was thought to represent a benign fracture. b Axial fused FDG-PET/CT image demonstrates increased radiotracer uptake in the L2 vertebral body (arrow). c Sagittal CT image demonstrates L2 and T11 compression fractures (arrows), for which malignancy was indeterminate

Fig. 6

a False-positive compression fracture in a patient on bone marrow-stimulating agents. Sagittal FDG-PET image demonstrates diffuse increased radiotracer activity throughout the spine. Note the compression fractures with focal kyphotic angulation at L1–L2 (arrow). The patient was thought to have diffuse metastatic disease with malignant compression fracture. b Sagittal FDG-PET image performed 2 months later demonstrates resolution of diffuse radiotracer uptake. Kyphotic angulation from compression fractures remains (arrow). The patient was off bone marrow-stimulating agents for 6 weeks at the time of the study

Fig. 7

a False-positive compression fracture in a patient with laryngeal cancer. Sagittal fused FDG-PET/CT image demonstrates increased radiotracer uptake in the T12 compression fracture (SUV = 4.9; arrow). b Sagittal T1-weighted MRI demonstrates compression fracture of T12 with linear low signal of the superior endplate and preservation of the normal marrow signal (arrow). Note compression fracture of L1 with relative preservation of normal marrow signal (arrowhead). c Sagittal STIR image demonstrates hyperintensity of the superior endplates of the T12 and L1 compression fractures, suggesting acute fractures (arrows). d Sagittal fat-suppressed T1-weighted MRI following the administration of gadolinium demonstrates enhancement of the T12 compression fracture (arrow). Mild enhancement is noted, involving the L1 compression fracture (arrowhead). Findings were thought to be benign on MRI and on follow-up imaging, there was resolution of bone marrow edema and enhancement

Fig. 8

a Acute and chronic benign compression fractures in a patient without history of malignancy. Sagittal fused FDG-PET/CT image demonstrates multiple compression fractures throughout the thoracic and lumbar spine with mild increased radiotracer uptake at T11 (SUV = 2.1; arrow). b Sagittal T1-weighted MRI demonstrates multiple chronic compression fractures (arrowheads). Mild decreased signal intensity of the T11 compression fracture suggests edema (arrow). c Sagittal STIR image demonstrates mild edema in the T11 compression fracture suggestive of an acute fracture (arrow). There is no evidence of marrow edema in the remaining compression fractures

Fig. 9

Mean standardized uptake value (SUV) of benign and malignant compression fractures. The top of the boxes represent the mean and error bars represent the range of SUV. There is a statistically significant difference between the SUV values of benign and those of malignant compression fractures (p < 0.001, Student’s t test)