Combining magnetic resonance imaging (MRI) with 18F-fluorodeoxyglucose positron emission tomography/computed tomography (FDG PET/CT) more effectively diagnoses and/or defines glioma grade than does MRI alone, according to an Australian study in the October 2019 Journal of Medical Imaging and Radiation Oncology. However, the researchers at the Royal Brisbane and Women’s Hospital in Brisbane, Queensland, caution that due to limited sensitivity and negative predictive values, negative scans should not guide decisions for observation in cases where surgery would be recommended.
Defining glioma grade is vital to performing targeted treatment approaches. High-grade glioma is difficult to treat, with rapid progression and poor prognosis.
MRI is the primary imaging modality recommended for glioma, but it can be limited in defining tumor grade and boundaries, the study authors explain, because there is substantial overlap of apparent diffusion coefficients (ADCs) between grades and some of the tumor may not enhance. FDG PET/CT identifies increased glucose uptake, which is greater in high-grade gliomas than in low-grade tumors.
The multi-specialty team sought to quantify the utility FDG PET/CT in identifying high-grade gliomas and to compare its diagnostic performance with MRI alone and combined with MRI. Their retrospective study included 33 patients who underwent both examinations within 12 months. Thirty-six scans were performed, 17 for initial diagnosis and 19 for follow-up.
When all tumor grades were included, MRI and PET/CT were concordant in predicting high-grade disease in only 23 cases, with a combined specificity of 100% and a positive predictive value (PPV) of 100%. Sensitivity was 79% and negative predictive value was 75%. By comparison, MRI alone had a sensitivity of 77%, a specificity of 86%, a PPV of 89%, and an NPV of 71%. FDG PET/CT was lower in all categories, at 59%, 79%, 81%, and 55% respectively.
The researchers reported that MRI results were inaccurate in 4 scans, and FDG PET/CT in 9. These occurred in 12 patients who had World Health Organization (WHO) grade II or III gliomas, as well as one patient with melanoma.
Most study subjects had grade II (11) or grade III (17) tumors. When both modalities scans were concordant, specificity and PPV were both 100%, sensitivity was 70%, and NPV was 67%. MRI outperformed FDG PET/CT, with a sensitivity of 76%, specificity of 82%, PPV of 87%, and NPV of 69% compared to 47%, 73%, 73%, and 47% respectively for FDG PET/CT.
“Diffusely or focally increased FDG uptake on PET/CT in our series was not a clinically reliable imaging modality alone for the diagnosis or surveillance of suspected high-grade histology,” wrote the authors. “We have found that FDG-PET/CT is an accurate modality for identification of grade IV glioma... The difficulty appears to be differentiating between grade II and III disease.”
“Unfortunately, the negative predictive value of the combined imaging modalities in our study is still poor at 75%, and does not allow safe exclusion of patients from biopsy or surgical management. We do not support the use of FDG-PET/CT in the evaluation of glioma grade. Contrast-enhanced MRI with histological analysis of lesions suspicious for high-grade disease remains our standard for diagnosis,” they concluded.
MRI plus 18F-FDG PET/CT recommended, with caveats, to identify gliomas. Appl Radiol.