Background: Glioblastoma (GBM) is the most common primary brain tumor in adults, with a poor prognosis and survival. Although typically presenting with focal neurological deficits, seizures, or cognitive decline, GBM can occasionally mimic autoimmune encephalitis (AE), leading to significant diagnostic delay. The overlap in clinical, radiological, and serological findings between GBM and AE underscores the need for thorough evaluation. Methods: We retrospectively reviewed cases of patients diagnosed between 2016 and 2023 with pathology-confirmed GBM, critically rethinking those cases initially diagnosed with AE at symptom onset. The diagnostic workup included magnetic resonance imaging (MRI), cerebrospinal fluid (CSF) analysis, autoantibody testing, and whole-body nuclear scanning to exclude extracranial malignancies. Results: We found five female patients diagnosed with GBM who initially presented with signs and symptoms suggestive for AE. Initial MRI showed non-specific brain tissue alterations, without definitive tumor features. CSF analysis was largely unremarkable, though some cases exhibited positive autoantibodies. Despite therapy, clinical deterioration and follow-up MRI revealed infiltrative intra-axial lesions with contrast enhancement, leading to pathology-confirmed GBM diagnoses. All patients had poor prognoses, with a mean survival of 10 ± 4 months. Conclusions: GBM can mimic AE, delaying appropriate treatment. In patients with atypical MRI findings and suboptimal response to therapy, early follow-up imaging and biopsy should be considered to exclude malignancy. A multidisciplinary approach is critical for timely diagnosis and improved management.

Different types of software to analyze CTP data in patients with ischemic stroke are available. Assessing their comparability and interchangeability in clinical practice represents an only partly addressed question. Here we present a comparison between 2 distinct commercially available CTP types of software, analyzing their performance in estimating ischemic core volumes and evaluating the possible impact on patient stratification strategies to endovascular treatment (EVT). In this single-center retrospective monocentric observational study, 109 patients with stroke (mean age = 72.4 ± 12.4 years, M/F = 41/68) were included from January 2023 to June 2024. To evaluate the possible clinical relevance of the use of different types of software, DAWN and DEFUSE-3 criteria were applied to stratify the population. The software was compared (Viz.ai and syngo.via), and for both programs, different relative CBF thresholds were used to define the ischemic core. The 2 software programs showed significant differences in core volume identification, independently from the used threshold (all comparisons with P < .001). When the DAWN criteria for EVT were applied, the use of one software compared with another led to a significant (P = .005) increase in subjects excluded from EVT. The use of a more conservative threshold significantly reduced (P = .68) this discrepancy. Within-subject analysis of CTP data with different software and thresholds might lead to significantly different core estimation and treatment stratification in patients with stroke. Though this effect can be mitigated by using specific thresholds, the physician should be aware of these differences when evaluating CTP data in clinical practice, given the possible direct implications in their decision-making process.