Top Line: When post-SRS MRI leaves you guessing between necrosis and recurrence, could 18F-fluciclovine PET tip the scales?
The Study:
The phase 2 PURSUE trial tackled a perennial challenge in radiation oncology: distinguishing recurrent brain metastasis from treatment-related change after SRS. The trial enrolled 23 patients with previously irradiated, MRI-equivocal lesions who were planned to undergo resection. All underwent 18F-fluciclovine PET prior to surgery allowing histopathology to serve as ground truth. Among the 23 lesions, 10 (43%) represented true recurrence. Several uptake metrics were evaluated, but one simple qualitative measure stood out at predicting recurrence: lesion uptake greater than the parotid gland. On static PET at 10–20 minutes post-injection, this threshold delivered specificity of 92–100%. However, sensitivity ranged from 40–80% across readers. Semi-quantitative analysis also performed well: SUVmax ≥4.8 achieved 80% sensitivity and 85% specificity (AUC 0.87). Further analysis suggested an optimal visual interpretation window 15–25 minutes post-injection, where visual reads reached 60–90% sensitivity and 77–100% specificity. Another semi-quantitative alternative, the SUVpeak lesion-to-pituitary ratio ≥1.1, produced 70–80% sensitivity and ~85% specificity across timepoints.
TBL: Results from the PURSUE study established visual criteria for differentiating tumor recurrence and necrosis after brain SRS. The investigators ultimately proposed straightforward visual criteria suggestive of recurrence: uptake ≥ parotid (or the pituitary if the parotids aren’t in field). | Kotecha, Int J Radiat Oncol Biol Phys 2026
The Study:
The phase 2 PURSUE trial tackled a perennial challenge in radiation oncology: distinguishing recurrent brain metastasis from treatment-related change after SRS. The trial enrolled 23 patients with previously irradiated, MRI-equivocal lesions who were planned to undergo resection. All underwent 18F-fluciclovine PET prior to surgery allowing histopathology to serve as ground truth. Among the 23 lesions, 10 (43%) represented true recurrence. Several uptake metrics were evaluated, but one simple qualitative measure stood out at predicting recurrence: lesion uptake greater than the parotid gland. On static PET at 10–20 minutes post-injection, this threshold delivered specificity of 92–100%. However, sensitivity ranged from 40–80% across readers. Semi-quantitative analysis also performed well: SUVmax ≥4.8 achieved 80% sensitivity and 85% specificity (AUC 0.87). Further analysis suggested an optimal visual interpretation window 15–25 minutes post-injection, where visual reads reached 60–90% sensitivity and 77–100% specificity. Another semi-quantitative alternative, the SUVpeak lesion-to-pituitary ratio ≥1.1, produced 70–80% sensitivity and ~85% specificity across timepoints.
TBL: Results from the PURSUE study established visual criteria for differentiating tumor recurrence and necrosis after brain SRS. The investigators ultimately proposed straightforward visual criteria suggestive of recurrence: uptake ≥ parotid (or the pituitary if the parotids aren’t in field). | Kotecha, Int J Radiat Oncol Biol Phys 2026