Computed Tomography and Magnetic Resonance Imaging Findings in Early Glioblastoma (e-GBM): Spotting the Wolf in Sheep's Clothing.

INTRODUCTION: Radiological differences exist between early-stage and late-stage glioblastoma (GBM). Most clinicians are aware of the latter appearances, but in our experience, early GBM (e-GBM) is often missed due to the absence of these classical findings. In this retrospective cohort study, we aim to highlight the radiological findings of e-GBM in order to improve recognition. This is important, as earlier recognition and treatment of this serious condition may improve prognosis. METHODS: A retrospective study of all patients presented at our institutional neuro-oncology multidisciplinary team meeting between 2012 and 2023 was undertaken. All patients had histologically confirmed GBM and early imaging that did not demonstrate the typical characteristics of malignant tumour. Imaging performed prior to the development of classical magnetic resonance imaging (MRI) features was reviewed by a single experienced neuroradiologist to investigate common imaging characteristics of e-GBM. RESULTS: Thirty-four patients (21 male, 13 female) were included. All patients presented with neurological symptoms and underwent MRI, with 30 of 34 patients having gadolinium-enhanced MRI. On T2-weighted and fluid-attenuated inversion recovery (FLAIR) sequences, all 34 patients had cortical signal change, while 33 of 34 patients had signal change in the subcortical region. All regions of FLAIR and T2 signal change demonstrated associated restricted diffusion. Thirty-two of 34 patients underwent computed tomography (CT) prior to MRI. Twenty-three of 32 patients had hyperdensity on CT correlating with areas of signal change on MRI. CONCLUSION: e-GBM should be considered in patients presenting with neurological symptoms, in particular seizures, when T2WI or FLAIR signal abnormalities in a cortical or subcortical location are seen with associated restricted diffusion. Hyperdensity on non-contrast CT imaging congruent with areas of signal change on MRI should also prompt clinicians to consider e-GBM as a differential diagnosis.