Prediction of low grade glioma recurrence using dynamic [18F]-FET-PET

Objectives [18F]-fluoroethyl-L-tyrosine-PET (FET-PET) is increasingly used in the diagnostic work-up of suspected and recurrent brain tumors. Dynamic imaging has been shown to be superior to standard static analysis in differentiating high and low grade glioma (LGG). We now hypothesize that dynamic FET PET imaging may also predict tumor recurrence in LGG patients.

Methods 14 patients (7 m; 7 f) with histologically proven LGG were included. All patients had received a 40 min dynamic [18F]-FET-PET scan prior to surgery. Images were evaluated using an in-house multiparametric software (Anima). Briefly, 90 %-isocontour tumor and background ROIs were drawn and time-activity curves were created. Different kinetic parameters were extracted (time-to-peak, slope, relative slope, sumSUVmax). LGG recurrence was determined based on clinical follow-up including MR imaging (mean FU-period 29.1, SD 13.1 months) which served as a gold-standard.

Results 6 out of 14 patients (42.9 %) had recurrent disease with a mean recurrence-free interval of 16.3 (SD 9.4) months. The ‘relative slope’ kinetic parameter showed good performance in identifying patients with low/high risk for recurrence; according to ROC analysis an optimal threshold of 0.149 was determined (p=0.014; sensitivity 83.3 %, specifity 87.5 %, accuracy 85.7 %). Survival curve analysis showed significant longer recurrence-free survival in the low risk group (logrank test: p=0.007).

Conclusions Relative slope of dynamic FET uptake is a promising kinetic parameter in the prediction of recurrent LGGs. However, further studies in larger patient cohorts are necessary to confirm these results.