Early Recurrence Detection of Glioma Using 18F-Fluorocholine PET/CT

Abstract

ObjectiveThe aim of this study was to analyze the usefulness of 18F-fluorocholine PET/CT in the early diagnosis of tumor recurrence, increasing the diagnosis confidence of MRI.MethodsPatients with a previous gross total resection of glioma and the first suspicious or doubtful for recurrence MRI were prospectively included and subjected to 18F-fluorocholine PET/CT. An independent and combined assessment of 18F-fluorocholine PET/CT and multimodal MRI was performed classifying the studies as positive or negative for tumor recurrence. Final diagnosis (recurrence or not) was obtained by histological confirmation or clinical and imaging follow-up. The relation of SUVmax and tumor-to-background ratio with progression, the diagnostic performance of imaging techniques, and their concordance (κ Cohen) were analyzed.ResultsTwenty-four studies on 21 patients were assessed. Recurrence was diagnosed in 20 cases. PET/CT was positive in 23 cases (3 false positive), whereas MRI was positive in 15 cases (1 false positive). MRI was false negative in 6 cases. There was no false negative on 18F-fluorocholine PET/CT. Accuracy of PET/CT versus MRI was 87.5% and 70.8%, respectively. The combined evaluation of both techniques did not show any advantage with respect to PET/CT results alone. The concordance between both imaging techniques was low (κ = 0.135; P = 0.375). SUVmax and tumor-to-background ratio were related to recurrence (areas under the curve of 0.844 [P = 0.033] and 0.869 [P = 0.022], respectively).Conclusions18F-fluorocholine PET/CT was helpful for increasing the diagnostic confidence in the cases of MRI doubtful for recurrence in order to avoid a delayed diagnosis.

Prognostic Value of Quantitative Indexes in Glioblastoma Subregions

“…MRI is the best method for assessing GBM recurrence, but current studies have limitations. Some only focus on pre or post-chemoradiotherapy (CRT) MRI 3,4 , neglecting the impact of surgical factors on prognosis. Previous research used binary categories, making personalized risk assessments challenging.…”