gms | German Medical Science

Objective: After radiosurgery of brain metastases, the MRI-based differentiation between tumor recurrence and radiation-induced lesions is challenging. In this context, the place of [¹⁸F] Fluoroethyltyrosine positron emission tomography ([¹⁸F]-FET PET) is poorly defined. In the current series, static and dynamic PET findings were described in correlation with both histological findings and follow-up evaluation.

Method: Patients with MRI-suspected tumor recurrences after previously performed radiosurgical treatment were included (03/2006-11/2013). Radiosurgery alone (median dose 20 Gy) or in combination with whole brain radiotherapy (median dose 30 Gy) was applied in 6 and 5 patients, respectively. MRI criteria for a recurrence were i) the appearance of a new contrast enhancing area after previously observed complete response or ii) an increase of the enhancing area after previously observed partial response/ stable disease. [¹⁸F]-FET PET analysis was done in all patients and included the assessment of both maximum standardized uptake value to background activity (SUV_max/BG) and dynamic analysis of intratumoral ¹⁸F-FET uptake over time (increasing vs. decreasing time-activity curves). PET-guided stereotactic biopsy was performed in all patients. Follow-up evaluation was done in three months intervals. Informed consent was obtained from all patients.

Results: Biopsy revealed a tumor recurrence in 7 patients and a reactive lesion in 4 patients (false positive rate of MRI: 37%). All patients with a biopsy-proven tumor recurrence exhibited a SUV_max/BG > 2 in combination with decreasing time activity curves (TAC). Two out of 4 patients with proven radiogenic lesions showed a SUV_max/BG < 2 in combination with increasing TAC, whereas the other two displayed PET findings not distinguishable from that of the recurrence group. In the former two patients, regressive lesions were seen over time (median follow-up 12 months), whereas in the latter two patients tumor recurrence occurred 4 and 7 months after biopsy.

Conclusions: Conventional MRI is notoriously inaccurate in the differential diagnosis of radiosurgically treated brain metastases. SUV_max/BG values > 2 in combination with decreasing TAC are highly suspicious for a tumor recurrence. Whether increased SUV_max/BG values can predict early recurrence even in patients with biopsy-proven radiogenic lesions deserves further evaluation. At the moment, close follow-up evaluation is recommended for these patients.