Dynamic Brain [F18]-Fluoroestradiol (FES) PET Protocol Optimization for Improved Detection of Breast Cancer Brain...

Brain [¹⁸F]Fluoroestradiol (FES) PET enables noninvasive assessment of estrogen receptor (ER) expression and may improve detection of ER-positive (ER⁺) breast cancer brain metastases (BC-BM), with pilot studies demonstrating clinical utility. Our aim was to leverage our dynamic brain [¹⁸F]FES PET acquisition protocol to characterize BC-BM uptake kinetics across different post-injection intervals, with the goal of improving lesion detectability.

Patients with ER⁺ primary disease and BC-BM were enrolled prospectively and underwent dynamic [¹⁸F]FES PET/CT and contemporaneous standard-of-care contrast-enhanced brain MRI. PET was acquired over 90 minutes using a dynamic protocol (34 frames). Four 15-minute static time windows were reconstructed including 30-45, 45-60, 60-75, and 75-90 minutes post-injection, respectively. MRI was used as gold standard for clinical lesion characterization and segmentation, with subsequent [¹⁸F]FES PET co-registration. Time-window-specific PET-MRI lesion matching rate (LMR) and mean standardized uptake value ratio (SUVR, normalized to cerebellar cortex) were calculated.

Eleven patients with a total of 92 MRI-confirmed lesions were included (73 parenchymal, 8 leptomeningeal, 6 calvarial, 4 dural-based, and 1 pituitary stalk). Time window-3 yielded the highest LMR at 68% (63/92 lesions), followed by time window-2 (62%), time window-4 (60%), and time window-1 (57%). Mean SUVR increased across time, with means of 1.83 ± 1.46, 2.08 ± 1.88, 2.29 ± 2.37, and 2.38 ± 2.50 for time window-1, -2, -3, and -4, respectively. Despite higher SUVR at later time points, 8 lesions (6 parenchymal, 1 leptomeningeal, and 1 calvarial) that were visible at 60-75 min were undetectable at 75-90 min.

Time window-3 at 60-75 minutes post-injection provided optimal LMR for brain [¹⁸F]FES PET in ER⁺ BC and suspected BM. Imaging beyond this interval may not further improve diagnostic yield and may reduce visibility. Limitations of this study include small cohort size, population heterogeneity with regard to prior treatment, and lack of pathologic confirmation. Our findings suggest that 60-75 minutes post-injection may represent the optimal acquisition time window and highlight the importance of timing standardization for neuro-oncologic applications of brain [¹⁸F]FES PET.