Background: The response of bone-dominant (BD) breast cancer to therapy is difficult to assess by conventional imaging. Our preliminary studies have shown that quantitative serial 2-[(18)F] fluoro-2-deoxy-D: -glucose positron emission tomography (FDG PET) correlates with therapeutic response of BD breast cancer, but the relationship to long-term outcome measures is unknown. Our goal was to evaluate the prognostic power of serial FDG PET in BD breast cancer patients undergoing treatment.
Methods: We reviewed medical records of 405 consecutive breast cancer patients referred for FDG PET. Of these, 28 demonstrated metastatic BD breast cancer, were undergoing treatment, had at least 2 serial PET scans, and had abnormal FDG uptake on the first scan. Standardized uptake value (SUV) for the most conspicuous bone lesion at the initial scan, absolute change in SUV over an interval of 1-17 months, and percent change in SUV were considered as predictors of time-to-progression (TTP) and time to skeletal-related event (t-SRE).
Results: Using proportional hazards regression, smaller percentage decreases in SUV (or increases in SUV) were associated with a shorter TTP (P < 0.006). A patient with no change in SUV was twice as likely to progress compared to a patient with a 42% median decrease in SUV. A higher SUV on the initial FDG PET predicted a shorter t-SRE (hazard ratio = 1.30, P < 0.02).
Conclusions: Changes in serial FDG PET may predict TTP in BD metastatic breast cancer patients. However, larger prospective trials are needed to validate changes in FDG PET as a surrogate endpoint for treatment response.