Fereidan‐Esfahani, Mahboubeh

Abstract Clinical management of IDH wildtype glioblastoma (GBM) and tumefactive multiple sclerosis (tMS) is drastically different. GBM requires maximal safe resection followed by chemoradiation, while tMS outcome is worsened by surgery and radiotherapy. Misdiagnosis of brain lesions may expose patients to unnecessary anxiety, surgery, or radiotherapy. Noninvasive methods to help with accurate diagnosis of tumor and non-tumor etiologies are needed. tMS subjects were identified by the Mayo Clinic Center for Multiple Sclerosis and Autoimmune Neurology (CMSAN). GBM subjects were matched to tMS by age at diagnosis, index MRI date, and sex. Inclusion criteria included one cm minimal lesion size and pre-operative post-contrast T1 and T2 images available for analysis. To identify potential experimental design bias, MRI technology and clinical characteristics were compared across groups. A 3D-DenseNet121 was used to develop a classification model using prespecified parameters: 600 epochs, batch size 16, learning rate 10-3, cross entropy loss, and AdamW optimizer. The stopping rule was defined as three sequential epoch cross entropy loss < 0.02. Five-fold cross validation (CV) was utilized to obtain an unbiased estimate of predictive performance (overfitting) and sensitivity of the 3D-DenseNet121 algorithm was evaluated. 224 subjects (114 GBM, 110 tMS) were analyzed: 40% male, median age 46. After one round of CV, the average unbiased estimate of sensitivity and specificity for predicting GBM from the final classification model was 0.86 and 0.73, respectively. To assess the reproducibility of the deep learning algorithm (3D-DenseNet121), the exact same CV folds were rerun resulting in average sensitivity and specificity of 0.82 and 0.75, respectively. Obtaining unbiased estimates of model performance using internal validation is critical before evaluating a final model on an external dataset. Understanding sensitivity of a deep learning algorithm is important for model validity. Differentiation of GBM and tMS appears reasonable with MRI; external validation is in progress.

AbstractObjectiveTo describe demographic, clinical, and radiographic features of tumefactive demyelination (TD) and identify factors associated with severe attacks and poor outcomes.MethodsRetrospective review of TD cases seen at Mayo Clinic, 1990–2021.ResultsOf 257 patients with TD, 183/257 (71%) fulfilled the 2017 multiple sclerosis (MS) McDonald criteria at the last follow‐up, 12/257 (5%) had myelin oligodendrocyte glycoprotein antibody‐associated disease (MOGAD), 0 had aquaporin‐4‐IgG seropositive neuromyelitis optic spectrum disorders (AQP4+ NMOSD), and 62/257 (24%) were cryptogenic. Onset before age 18 was present in 18/257 (7%). Female to male ratio was 1.3:1. Cerebrospinal fluid oligoclonal (CSF) bands were present in 95/153 (62%). TD was the first demyelinating attack in 176/257 (69%). At presentation, 59/126 (47%) fulfilled Barkhof criteria for dissemination in space, 59/100 (59%) had apparent diffusion coefficient (ADC) restriction, and 57/126 (45%) had mass effect. Despite aggressive clinical presentation at onset, 181/257 (70%) of patients remained fully ambulatory (Expanded Disability Status Scale [EDSS] ≤4) after a 3.0‐year median follow‐up duration. Severe initial attack‐related disability (EDSS ≥4) was more common in patients with motor symptoms (81/143 vs. 35/106, p < 0.0001), encephalopathy (20/143 vs. 2/106, p < 0.0001) and ADC restriction on initial MRI (42/63 vs. 15/33, p = 0.04). Poor long‐term outcome (EDSS ≥4) was more common in patients with older onset age (41.9 ± 15 vs. 36.8 ± 15.6, p = 0.02) and motor symptoms at onset (49/76 vs. 66/171, p < 0.0001).InterpretationMost TD patients should be considered part of the MS spectrum after excluding MOGAD and NMOSD. Motor symptoms and older age at presentation portend a poor outcome.

AbstractBackground and purposeTumefactive demyelination (TD) presents with large inflammatory lesions mimicking tumors or other space‐occupying lesions. Limited epidemiological, clinical and radiological data exist for TD. We aimed to report the incidence rate, and clinical and radiological features of TD in Olmsted County, Minnesota.MethodsWe retrospectively reviewed patients with central nervous system inflammatory demyelination‐related diagnostic codes (January 1, 1998 to December 31, 2018) in the Rochester Epidemiology Project database, and adjusted incidence rates by age and sex to the 2010 US total population. We used the Expanded Disability Status Scale (EDSS) to assess outcomes (index attack and last follow‐up).ResultsOf 792 multiple sclerosis (MS) patients, 15 (eight males, seven females) had tumefactive MS, representing 1.9% of the MS population. The median (range) age at attack onset was 34.2 (2–61) years. Tumefactive lesion was the first clinical MS attack in 8/16 patients. Cerebrospinal fluid oligoclonal bands (OCBs) were present in 8/12 patients and 11/16 patients met the Barkhof criteria for dissemination in space. Most patients remained fully ambulatory (EDSS score ≤4 in 13/16 patients [81%]) after a median (range) follow‐up duration of 10.5 (1–20.5) years. Age‐adjusted annual incidence rates were 0.46/100,000 (95% confidence interval [CI] 0.12–0.81) for female patients, 0.66/100,000 (95% CI 0.23–1.02) for male patients, and 0.56/100,000 [95% CI 0.28–0.83] overall. When age‐ and sex‐adjusted to the 2010 US total population, the overall annual incidence rate was 0.57 (95% CI 0.28–0.84). Despite aggressive clinical presentation at disease onset, most patients remained fully ambulatory (EDSS score ≤4 in 13/16 patients) with a relapsing‐remitting course.ConclusionsAlthough incidence is rare, TD should be suspected in patients presenting with subacutely progressive neurological deficits associated with magnetic resonance imaging findings of ring enhancement, apparent diffusion coefficient restriction, and OCB on spinal fluid analysis.

We previously applied Systematic Evolution of Ligands by EXponential enrichment (SELEX) technology to identify myelin-specific DNA aptamers, using crude mouse central nervous system myelin as bait. This selection identified a 40-nucleotide aptamer (LJM-3064). Multiple biotinylated LJM-3064 molecules were conjugated to a streptavidin core to mimic a multimeric immunoglobulin M (IgM) antibody, generating 3064-BS-streptavidin (Myaptavin-3064). We previously showed that Myaptavin-3064 induces remyelination in the Theiler’s murine encephalomyelitis virus (TMEV) model of chronic spinal cord demyelination. While details of target binding and the mechanism of action remain unclear, we hypothesized that Myaptavin-3064 induces remyelination by binding to oligodendrocytes (OLs). We now report the results of binding assays using the human oligodendroglioma (HOG) cell line, applying both flow cytometry and immunocytochemistry (IC) to assay aptamer conjugate binding to cells. IC assays were applied to compare aptamer conjugate binding to primary embryonic mouse mixed cortical cultures and primary adult rat mixed glial cultures. We show that Myaptavin-3064 binds to HOG cells, with increased binding upon differentiation. In contrast, a negative control aptamer conjugate, 3060-BS, which did not promote central nervous system (CNS) remyelination, does not bind to HOG cells. Myaptavin-3064 did not bind to lung (L2) or kidney (BHK) cell lines. Total internal reflection fluorescence (TIRF) imaging indicates that Myaptavin-3064 binds at the cell membrane of live cells. In addition to HOG cells, Myaptavin-3064 binds to adult rat OLs, but not to embryonic mouse mixed cortical cultures. These data support the hypothesis that Myaptavin-3064 binds to a surface molecule on both rodent and human OLs in a manner that triggers a remyelination signal pathway.