The purpose was to investigate the interrater agreement of FDG-PET/CT and bone scintigraphy for diagnosing bone recurrence in breast cancer patients. A total of 100 women with suspected recurrence of breast cancer underwent planar whole-body bone scintigraphy with [99mTc]DPD and FDG-PET/CT. Scans were evaluated independently by experienced nuclear medicine physicians and the results for one modality were blinded to the other. Images were visually interpreted using a 4-point assessment scale (0 = no metastases, 1 = probably no metastases, 2 = probably metastases, 3 = definite metastases). Out of 100 women, 22 (22%) were verified with distant recurrence, 18 of these had bone involvement. The proportions of agreement between readers were 93% (86.3-96.6) for bone recurrence with FDG-PET/CT and 47% (37.5-56.7) for bone recurrence with planar bone scintigraphy. The strengths of agreement between readers for diagnosing bone recurrence was 'almost perfect' with FDG-PET/CT and was 'fair' with planar bone scintigraphy according to Cohen's kappa value of 0.82 (0.70-0.95) and 0.28 (0.18-0.39), respectively. Interrater agreement yielded improved reproducibility with FDG-PET/CT versus with bone scintigraphy when diagnosing recurrence with bone metastasis in this patient cohort.
Keywords: agreement; bone scintigraphy; breast cancer; interrater; positron emission tomography; recurrence; repeatability; reproducibility.