Analysis of 18F-FET PET for grading of recurrent gliomas: is evaluation of uptake kinetics superior to standard methods?

Abstract

The aim of the present study was to evaluate whether extended analyses of O-(2-18F-fluoroethyl)-L-tyrosine (FET) uptake kinetics provide results superior to those of standard tumor-to-background ratios in predicting tumor grade in patients with pretreated gliomas.

Methods: Dynamic 18F-FET PET studies (0-40 min after injection of 180 MBq of 18F-FET) were performed on 45 glioma patients with suspected tumor recurrence after multimodal treatment. For the standard method, tumoral maximal standardized uptake value (SUVmax) and the ratio to the background were derived from a summed image 20-40 min after injection. Dynamic data evaluation comprised several approaches: first, SUV within a 90% isocontour threshold (SUV90) and the respective ratio to the background calculated for each time frame between 5 and 40 min after injection; second, the time to peak analysis; and third, various parameters accounting for the individual time course of 18F-FET uptake. Results were correlated with the histopathologic findings of MRI/PET-guided stereotactic biopsies and were evaluated with respect to their discriminatory power to separate low- from high-grade tumors using receiver-operating characteristic (ROC) analyses.

Results: The parameters taking into account the individual time course of 18F-FET uptake were able to differentiate low-grade from high-grade recurrent astrocytomas with high diagnostic accuracy, reaching the best differentiation with a sensitivity and specificity of 92% and an area under the ROC curve (AUC) of 0.94. For the other parameters, the respective values were considerably lower (time to peak: 85% sensitivity and 88% specificity; SUV90-to-background ratio for single-frame evaluation of the early-uptake phase: 100% sensitivity, 62% specificity, and 0.81 AUC). The lowest performance was provided by the standard method (SUVmax: 73% sensitivity, 54% specificity, and 0.60 AUC; SUVmax-to-background ratio: 62% sensitivity, 62% specificity, and 0.59 AUC). Time-activity curves (5-40 min after injection) slightly and steadily increased in tumor-free patients and in low-grade tumors, whereas high-grade tumors showed an early peak around 10-15 min after injection followed by a decrease.

Conclusion: This study has shown differences in the dynamics of 18F-FET uptake between recurrent low- and high-grade gliomas. Therefore, parameters addressing the different kinetic behaviors allow discrimination with high diagnostic power between these 2 prognostically different groups. Thus, the techniques introduced here are clearly superior to the yet most widely used standard method.