Quantitative measures of FDG PET after neoadjuvant chemotherapy to predict breast cancer patient survival.

Abstract

1088 Background: In patients with locally advanced breast cancer (LABC), qualitative FDG positivity following neo-adjuvant chemotherapy (NC) has been shown to be inversely associated with survival (Emmering, Annals Oncol, 2008). We investigated quantitative measures of post-therapy FDG uptake, namely standardized uptake value (SUV) and glycolytic flux (Ki), as predictors of breast cancer survival. Methods: Forty-seven patients with LABC underwent dynamic FDG PET scans close to or at the end of NC and prior to surgical resection. Post-therapy FDG uptake at the primary tumor site was measured by mean SUV from 45-60 minutes after FDG injection, maximum SUV (SUVmax) from 50-55 minutes, and FDG glycolytic flux (Ki). Pathologic response (PR) was assessed for at the time of surgical resection. Cox proportional hazards models were used to estimate associations between log-transformed measures of post-therapy FDG uptake, PR and outcome. Results: Median SUVmax was 1.9 (0.9 – 9.2) and median Ki was 2.2 (0.02 – 47.7) mL/min/g. Median follow-up for relapse was 5.7 years with 11 events and 6.3 years for survival with 10 deaths. PR was not significant for DFS (p = .39) or OS (p= .48). Post-therapy FDG uptake measures showed a statistically significant ability to predict survival. SUVmax predicted DFS (p=0.02) and OS (p=0.01). Ki was associated with DFS (p <0.01) and OS (p <0.01). PET measured hazard ratios were not attenuated in multivariate analysis controlling for known prognostic markers such as primary tumor PR and nodal status. However, multivariate survival models appeared highly influenced by one patient with the shortest survival time (1.3 years) and highest SUVmax and Ki. Without this patient, Ki remained a borderline independent predictor of DFS (p= .08) and OS (p= .07), but SUVmax was no longer significant for DFS (p= .32) or OS (p=0.26). Conclusions: Our analysis suggests that quantitative measures of post-therapy FDG PET provide information beyond PR for predicting which LABC patients are at highest risk for relapse and death. This information may be useful in directing post-surgery treatment. Supported by NIH grants CA42045, CA138293, and CA148131.