Value of F-DOPA PET in the long term follow up of radionecrotic brain metastases after radiosurgery: comparison with MRI

Aim. To analyse the value of sequential F-DOPA PET/CT scans in the long term follow-up of predominantly radionecrotic brain metastases after stereotactic radiosurgery. Comparison with MRI-derived parameters is given. Materials and Methods. Eighty-one patients with previously irradiated brain metastases who underwent F-DOPA PET at our institution for differential diagnosis between radionecrosis (RN) and progression (PD) were screened. Inclusion criteria were i) initial negative or borderline uptake values (rSUV) defined according to a previously published method, ii) availability of repeated F-DOPA scans over an observational period >12 months. Standard patient monitoring in this setting includes repeated MRI and F-DOPA scans every 3-6 and 6-12 months, respectively. F-DOPA PET/CT consisted of a 20 min acquisition starting 15 minutes after tracer injection. rSUV was calculated as the ratio between the lesion SUVmax and the SUVmax of the controlateral background. Tumor maximal transverse diameter (Dmax) was calculated on contrast-enhanced, T1-weighted sequences; tumour volumes were calculated by assuming a spherical geometry with Dmax as diameter. Cerebral blood flow (rCBV) was derived from perfusion-weighted images and quantified relative to the normal controlateral white matter. ∆rSUV, ∆Dmax and ∆Vol were calculated between first and last time point imaging; PD vs RN group comparisons were made with two-tailed Student’s t test. Results. Twenty patients (12 F, 8 M) with a total of 70 F-DOPA scans were included. Median follow-up was 30.5 months (range:13-51). So far, 4 (20%) patients have died because of intracranial (n=3) or extracranial (n=1) PD. Eleven (55%) metastases remained in histologically confirmed (n=5) or unconfirmed RN, while 9 (45%) showed a slow PD (n=3 histological confirmations) over time. All PD showed a steady rSUV increase over time; however, F-DOPA was false positive in one case of histologically confirmed RN. In two RN, there was an increase of rSUV followed by a subsequent slight decrease. An increase in Dmax was observed in all patients but two cases of RN. rCBV was falsely negative in 2 (22%) PD and falsely positive in 5 (45%) RN. Initial rSUV did not differ between groups (p=0.75), while final rSUV and ∆rSUV were significantly different between PD and RN (p<0.001 and p=0.006, respectively). All MRI derived parameters were not significantly different between PD and RN (all p>0.25). Conclusion. F-DOPA PET is a reliable tool to assess the evolution over time of predominantly radionecrotic brain metastases after stereotactic radiosurgery. In this setting, performances of MRI including perfusion-weighted imaging are lower.