OS4.4 Long term follow up of radionecrotic brain metastases assessed by serial F-DOPA PET/CT scans

Abstract

BACKGROUND AND AIM: Differential diagnosis between radionecrotic (RN) and progressive (PD) brain metastases after radiosurgery (SRS) is a diagnostic challenge. In a former exploratory study, we showed that 6-[18F]-fluoro-L-3,4-dihydroxyphenylalanine (F-DOPA) PET/CT is able to discriminate between RN and PD metastases with excellent diagnostic accuracy, outperforming MRI including perfusion-weighted imaging. Borderline uptake values often represent the coexistence of necrotic and viable neoplastic cells. The aim of the present report was to analyze the long term follow-up of predominantly radionecrotic metastases by means of serial F-DOPA PET/CT scans. MATERIALS AND METHODS: Eighty-one patients with suspected radiological progression at the site of previously irradiated brain metastases referred for F-DOPA PET at our institution between May 2010 and March 2016 were screened. All metastases had previously been treated with single fraction or multifraction SRS. To be included in the final analysis, patients were required to have i) a mildly positive F-DOPA defined by visual analysis and borderline lesion-to-background uptake ratio (rSUV) values defined according to a previously published method ii) a diagnosis of radionecrosis after initial PET/CT evaluation, defined histologically as the presence of <20% neoplastic features within the surgical specimen or radiologically as a stable or shrinking lesion over an observational period of 6 months and ii) a minimum of 12 month follow up. Standard patient monitoring in this setting includes repeated magnetic resonance imaging (MRI) every 3-6 months and repeated F-DOPA every 6-12 months. RESULTS: A total of 68 F-DOPA scans from 19 patients (11 F, 8 M) were evaluated. Median follow-up was 30.5 months (range:14-51). At the time of writing, 4 (21%) patients have died because of intracranial (n=3) or extracranial (n=1) PD. Ten (52%) metastases remained in histologically confirmed (n=4) or unconfirmed RN, while 9 (48%) showed a slow PD (n=3 histological confirmations). All progressive metastases showed a steady rSUV increase over time; however, PET/CT was false positive in one case of histologically confirmed RN. In two cases of confirmed RN (one histologically, one radiologically), there was an increase of rSUV followed by a subsequent slight decrease. PET/CT and MRI were concordant in 12 (63%) and discordant in 7 (37%) cases. MRI was false positive in 4/10 (40%) cases of RN and false negative in 2/9 (22%) cases of PD. CONCLUSION: Amino acid PET with F-DOPA is a reliable tool to assess the evolution over time of metastatic brain lesions after SRS, and performs better than MRI. Different patterns of F-DOPA uptake over time can discriminate with excellent diagnostic accuracy between PD and RN in the context of mixed viable and radionecrotic neoplastic cells.