Role of 18F FDG-PET/CT in Recurrent Breast Cancer: Correlation with CA15.3 and Survival Outcomes.

Background: Breast cancer (BC) is the most frequently diagnosed cancer globally. Following curative treatment of BC, monitoring locoregional and distant recurrence is a major clinical challenge. Patients and Methods: Seventy-two female patients with histo pathologically proven BC who were referred for 18F-FDG PET/CT due to a suspicion of recurrence. PET/CT images were analyzed both qualitatively and quantitatively. PET/CT findings and quantitative parameters (SUVmax, SUVpeak, SUVmean, MTV, and TLG) were correlated with the CA15.3 levels and survival outcomes. The gold standard was histopathological examination, radiologic, or clinical follow-up. Results: Fifty-eight patients had positive PET/CT findings, while the remainder 14 had negative ones. PET/CT demonstrated an excellent diagnostic performance in detecting recurrent BC (96.6% sensitivity, 92.3%specificity, 98.3% PPV, 85.7% NPV, and 95.8% accuracy). CA15.3 was significantly higher in the patients with positive PET/CT results compared to negative ones (p =0.044). There was a significant positive correlation (p<0.001) between MTV and TLG of the visceral and the most active recurrent lesions, and CA15.3. TLG of recurrent osseous lesions was highly correlated with CA15.3 (p = 0.025). SUVmax, peak, mean, and MTV were all significant predictors of overall survival (OS) (p = 0.005, 0.005, 0.014, 0.013,respectively).Conclusions: 18F-FDG-PET/CT is a useful addition to existing monitoring methods for verifying recurrence in BC patients suspected of having relapsed disease and mapping out recurrent sites. It should be considered in patients with suspected BC recurrence but normal CA15.3 levels. OS is significantly shorter in patients with elevated CA15.3 and positive PET/CT than in patients with normal CA15.3 and negative PET/CT recurrence. FDG PET/CT-based volumetric metrics (MTV and TLG) are significantly correlated with CA15.3. In patients with recurrent BC, FDG PET/CT-derived SUVmax, SUVpeak, SUVmean and MTV are significant predictors of survival.