FDG PET-CT of bone metastases in breast cancer

Abstract

97

Objectives: To retrospectively correlate the metabolic activity and the CT appearance of metabolically active metastatic osseous lesions in patients with breast cancer.

Methods: The FDG PET-CT images of 272 consecutive patients with breast cancer who were scanned during 07/02 to 11/05 were retrospectively reviewed. Forty-eight patients (age 35-75 years) with metabolically active bony lesions consistent with either highly probable or definite bony metastases were selected for further analysis. All metabolically active bony lesions greater than 1.5 cm were classified into three different morphologic groups using the CT bone window level (center 300 HU, width 1500 HU) as either osteoblastic, osteolytic, or no discernable structural lesions. The means of peak SUV for the three types of CT lesions were compared two at a time using the Student t-test with probability less than 0.05 considered as significant.

Results: Two-hundred seventy-seven lesions in 48 patients were considered to be either highly probable or definite bone metastases on PET. CT revealed osteoblastic changes in 63 (22%), osteolytic changes in 65 (24%), and no structural changes in 149 (54%) of lesions. The range and mean peak SUV were 1.6-9.0 and 4.1 for osteoblastic, 2.0-11.5 and 4.7 for osteolytic, and 2.1-13.0 and 4.7 for no structural change lesions, respectively. The mean peak SUV for osteobalstic lesions was significantly lower than that for no structural change bony lesions (p=0.02). However, there was no statistically significant difference in the metabolic activities of the osteoblastic to the osteolytic, and the osteolytic to no structural change lesions (p=0.06, and p=0.1, respectively).

Conclusions: FDG uptake in metabolic bone deposits from breast cancer does not appear to be highly correlated with structural morphology on CT. Approximately half of the lesions noted on PET had no morphological correlate on CT, although pathological confirmation was not obtained in this study. The broad range of SUV in each subgroup also suggests potential overlap with benign bone lesions.