Purpose: We assessed the performance of 6-[(18)F]-fluoro-L-3,4-dihydroxyphenylalanine (F-DOPA) PET for differentiating radionecrosis (RN) from tumour progression (PD) in a population of patients with brain metastases, treated with stereotactic radiosurgery. The accuracy of F-DOPA PET was compared with that of perfusion-weighted magnetic resonance (perfusion-MR).
Methods: In 42 patients with a total of 50 brain metastases from various primaries F-DOPA PET/CT was performed because of suspected radiological progression at the site of previously irradiated brain metastasis. Several semiquantitative PET parameters were recorded, and their diagnostic accuracy was compared by receiver operating characteristic curve analyses. The diagnosis was established by either surgery or follow-up. A comparison was made between F-DOPA PET and perfusion-MR sequences acquired no more than 3 weeks apart.
Results: Definitive outcome was available in 46 of the 50 lesions (20 PD, 26 RN). Of the 46 lesions, 11 were surgically excised while in the remaining 35 lesions the diagnosis was established by radiological and clinical criteria. The best diagnostic performance was obtained using the semiquantitative PET parameter maximum lesion to maximum background uptake ratio (SUVLmax/Bkgrmax). With a cut-off value of 1.59, a sensitivity of 90 % and a specificity of 92.3 % were achieved in differentiating RN from PD lesions (accuracy 91.3 %). Relative cerebral blood volume (rCBV) derived from perfusion-MR was available for comparison in 37 of the 46 metastases. Overall accuracy of rCBV was lower than that of all semiquantitative PET parameters under study. The best differentiating rCBV cut-off value was 2.14; this yielded a sensitivity of 86.7 % and a specificity of 68.2 % (accuracy 75.6 %).
Conclusion: F-DOPA PET is a highly accurate tool for differentiating RN from PD brain metastases after stereotactic radiosurgery. In this specific setting, F-DOPA PET seems to perform better than perfusion-MR.