Development of a Natural Language Processing (NLP) model to automatically extract clinical data from electronic health records: results from an Italian comprehensive stroke center.

INTRODUCTION: Data collection often relies on time-consuming manual inputs, with a vast amount of information embedded in unstructured texts such as patients' medical records and clinical notes. Our study aims to develop a pipeline that combines active learning (AL) and NLP techniques to enhance data extraction in an acute ischemic stroke cohort. MATERIALS AND METHODS: Consecutive acute ischemic stroke patients who received reperfusion therapies at IRCCS Humanitas Research Hospital were included. The Italian NLP Bidirectional Encoder Representations from Transformers (BERT) model was trained with AL to automatically extract clinical variables from electronic health text. Simulated active learning performances were evaluated on a set of labels representing patients' comorbidities, comparing Bayesian Uncertainty Sampling by Disagreement (BALD) and random text selection. Prognostic models predicting patients' functional outcomes using Gradient Boosting were trained on manually labelled and semi-automatically extracted data and their performance was compared. RESULTS: The active learning process initially showed null performance until around 20% of texts were labelled, possibly due to root layers freezing in the BERT model, yet overall, active learning improves model learning efficiency across most comorbidities. Prognostic modelling showed no significant difference in performance between models trained on manually labelled versus semi-automatically extracted data, indicating effective prediction capabilities in both settings. CONCLUSIONS: We developed an efficient language model to automate the extraction of clinical data from Italian unstructured health texts in a cohort of ischemic stroke patients. In a preliminary analysis, we demonstrated its potential applicability for enhancing prediction model accuracy.

Gliomagenesis, Epileptogenesis, and Remodeling of Neural Circuits: Relevance for Novel Treatment Strategies in Low- and High-Grade Gliomas.

Gliomas present a complex challenge in neuro-oncology, often accompanied by the debilitating complication of epilepsy. Understanding the biological interaction and common pathways between gliomagenesis and epileptogenesis is crucial for improving the current understanding of tumorigenesis and also for developing effective management strategies. Shared genetic and molecular mechanisms, such as IDH mutations and dysregulated glutamate signaling, contribute to both tumor progression and seizure development. Targeting these pathways, such as through direct inhibition of mutant IDH enzymes or modulation of glutamate receptors, holds promise for improving patient outcomes. Additionally, advancements in surgical techniques, like supratotal resection guided by connectomics, offer opportunities for maximally safe tumor resection and enhanced seizure control. Advanced imaging modalities further aid in identifying epileptogenic foci and tailoring treatment approaches based on the tumor's metabolic characteristics. This review aims to explore the complex interplay between gliomagenesis, epileptogenesis, and neural circuit remodeling, offering insights into shared molecular pathways and innovative treatment strategies to improve outcomes for patients with gliomas and associated epilepsy.

Late-onset relapsing neurotoxicity after Brexucabtagene autoleucel associated with high chimeric antigen receptor T cells in cerebrospinal fluid.

BACKGROUND AIMS: Mounting evidence suggests that persistent cell expansion is the main driver for both efficacy and toxicity of chimeric antigen receptor (CAR) T-cell therapy. Hereby, we describe a case of delayed recurrent neurotoxicity associated with late CAR T-cells re-expansion. CASE DESCRIPTION: A 44-year-old man suffering from mantle cell lymphoma received brexu-cel. After infusion, he developed grade 2 cytokine release syndrome. On day +11, grade 3 neurotoxicity was reported and high-dose methylprednisolone was started with a complete resolution of neurological manifestations. On day +30, he experienced a late-onset CAR T-cell toxicity associated with CAR T-cell re-expansion. The patient was treated with tocilizumab and dexamethasone, with resolution of symptoms. On day +58, he was readmitted for new onset of neurotoxicity. Notably, a new CAR T-cell expansion was observed, with an unexpectedly elevated cerebrospinal fluid/blood ratio. The patient was promptly treated with dexamethasone and then escalated to high-dose methylprednisolone and anakinra, with resolution of his neurologic condition noted. CONCLUSIONS: CAR T-cell-related neurotoxicity usually has an early monophasic course. To our knowledge, this is the first case of late-onset, recurrent neurotoxicity. Moreover, an elevated level of cerebrospinal fluid CAR T cells was observed, which may suggest that the delayed neurotoxicity was primarily caused by the brain infiltration of CAR T cells rather than driven by cytokine-mediated neuroinflammation.

Statin use and long-term risk of recurrent intracerebral haemorrhage: the MUCH-Italy.

BACKGROUND: Whether statin use after spontaneous intracerebral haemorrhage (ICH) increases the risk of recurrent ICH is uncertain. METHODS: In the setting of the Multicentric Study on Cerebral Haemorrhage in Italy we followed up a cohort of 30-day ICH survivors, consecutively admitted from January 2002 to July 2014, to assess whether the use of statins after the acute event is associated with recurrent cerebral bleeding. RESULTS: 1623 patients (mean age, 73.9±10.3 years; males, 55.9%) qualified for the analysis. After a median follow-up of 40.5 months (25th to 75th percentile, 67.7) statin use was not associated with increased risk of recurrent ICH either in the whole study group (adjusted HR, 0.99; 95% CI 0.64 to 1.53) or in the subgroups defined by haematoma location (deep ICH, adjusted HR, 0.74; 95% CI 0.35 to 1.57; lobar ICH, adjusted HR, 1.09; 95% CI 0.62 to 1.90), intensity of statins (low-moderate intensity statins, adjusted HR, 0.93; 95% CI 0.58 to 1.49; high-intensity statins, adjusted HR, 1.48; 95% CI 0.66 to 3.31) and use of statins before the index event (adjusted HR, 0.66; 95% CI 0.38 to 1.17). CONCLUSIONS: Statin use appears to be unrelated to the risk of ICH recurrence.

Reperfusion therapies in patients with acute ischaemic stroke and atrial fibrillation: data on safety and effectiveness from a multi-centre cohort study.

BACKGROUND: Intravenous thrombolysis (IVT) and/or endovascular therapy (EVT) are currently considered best practices in acute stroke patients. Data regarding the efficacy and safety of reperfusion therapies in patients with atrial fibrillation (AF) are conflicting as regards haemorrhagic transformation, mortality, and functional outcome. This study sought to investigate for any differences, in terms of safety and effectiveness, between AF patients with acute ischaemic stroke (AIS) treated and untreated with reperfusion therapies. METHODS: Data from two multicenter cohort studies (RAF and RAF-NOACs) on consecutive patients with AF and AIS were analyzed to compare patients treated and not treated with reperfusion therapies (IVT and/or EVT). Multivariable logistic regression analysis was performed to identify independent predictors for outcome events: 90-day good functional outcome and mortality. A propensity score matching (PSM) analysis compared treated and untreated patients. RESULTS: Overall, 441 (25.4%) were included in the reperfusion-treated group and 1,295 (74.6%) in the untreated group. The multivariable model suggested that reperfusion therapies were significantly associated with good functional outcome. Rates of mortality and disability were higher in patients not treated, especially in the case of higher NIHSS scores. In the PSM comparison, 173/250 patients (69.2%) who had received reperfusion therapies had good functional outcome at 90 days, compared to 146/250 (58.4%) untreated patients (p = 0.009, OR: 1.60, 95% CI:1.11-2.31). CONCLUSIONS: Patients with AF and AIS treated with reperfusion therapies had a significantly higher rate of good functional outcome and lower rates of mortality compared to those patients with AF and AIS who had undergone conservative treatment.

Early seizures and risk of epilepsy and death after intracerebral haemorrhage: The MUCH Italy.

INTRODUCTION: It is unclear which patients with non-traumatic (spontaneous) intracerebral haemorrhage (ICH) are at risk of developing acute symptomatic seizures (provoked seizures occurring within the first week after stroke onset; early seizures, ES) and whether ES predispose to the occurrence of remote symptomatic seizures (unprovoked seizures occurring more than 1 week after stroke; post-stroke epilepsy, PSE) and long-term mortality. PATIENTS AND METHODS: In the setting of the Multicenter Study on Cerebral Haemorrhage in Italy (MUCH-Italy) we examined the risk of ES and whether they predict the occurrence of PSE and all-cause mortality in a cohort of patients with first-ever spontaneous ICH and no previous history of epilepsy, consecutively hospitalized in 12 Italian neurological centers from 2002 to 2014. RESULTS: Among 2570 patients (mean age, 73.4 ± 12.5 years; males, 55.4%) 228 (8.9%) had acute ES (183 (7.1%) short seizures and 45 (1.8%) status epilepticus (SE)). Lobar location of the hematoma (OR, 1.49; 95% CI, 1.06-2.08) was independently associated with the occurrence of ES. Of the 2,037 patients who were followed-up (median follow-up time, 68.0 months (25th-75th percentile, 77.0)), 155 (7.6%) developed PSE. ES (aHR, 2.34; 95% CI, 1.42-3.85), especially when presenting as short seizures (aHR, 2.35; 95% CI, 1.38-4.00) were associated to PSE occurrence. Unlike short seizures, SE was an independent predictor of all-cause mortality (aHR, 1.50; 95% CI, 1.005-2.26). DISCUSSION AND CONCLUSION: The long-term risk of PSE and death after an ICH vary according to ES subtype. This might have implications for the design of future clinical trials targeting post-ICH epileptic seizures.

Primary vs. pre-emptive anti-seizure medication prophylaxis in anti-CD19 CAR T-cell therapy.

INTRODUCTION: Seizures may occur in up to 30% of non-Hodgkin lymphoma patients who received anti-CD19 CAR T-cell therapy, yet the optimal anti-seizure medication (ASM) prevention strategy has not been thoroughly investigated. METHODS: Consecutive patients affected by refractory non-Hodgkin lymphoma who received anti-CD19 CAR T-cells were included. Patients were selected and assessed using similar internal protocols. ASM was started either as a primary prophylaxis (PP-group) before CAR T-cells infusion or as a pre-emptive therapy (PET-group) only upon the onset of neurotoxicity development. RESULTS: One hundred fifty-six patients were included (PP-group = 88, PET-group = 66). Overall, neurotoxicity and severe neurotoxicity occurred in 45 (29%) and 20 (13%) patients, respectively, equally distributed between the two groups. Five patients experienced epileptic events (PET-group = 3 [4%]; PP-group = 2 [2%]). For all the PET-group patients, seizure/status epilepticus occurred in the absence of overt CAR-T-related neurotoxicity, whereas patients in the PP-group experienced brief seizures only in the context of critical neurotoxicity with progressive severe encephalopathy. ASMs were well-tolerated by all patients, even without titration. No patients developed epilepsy or required long-term ASMs. CONCLUSION: Our data suggest that both primary and pre-emptive anti-seizure prophylaxis are safe and effective in anti-CD19 CAR T-cell recipients. Clinical rationale suggests a possible more favourable profile of primary prophylaxis, yet no definitive conclusion of superiority between the two ASM strategies can be drawn from our study.

Long-Term Risk of Arterial Thrombosis After Intracerebral Hemorrhage: MUCH-Italy.

BACKGROUND: The identification of patients surviving an acute intracerebral hemorrhage who are at a long-term risk of arterial thrombosis is a poorly defined, crucial issue for clinicians. METHODS: In the setting of the MUCH-Italy (Multicenter Study on Cerebral Haemorrhage in Italy) prospective observational cohort, we enrolled and followed up consecutive 30-day intracerebral hemorrhage survivors to assess the long-term incidence of arterial thrombotic events, to assess the impact of clinical and radiological variables on the risk of these events, and to develop a tool for estimating such a risk at the individual level. Primary end point was a composite of ischemic stroke, myocardial infarction, or other arterial thrombotic events. A point-scoring system was generated by the ß-coefficients of the variables independently associated with the long-term risk of arterial thrombosis, and the predictive MUCH score was calculated as the sum of the weighted scores. RESULTS: Overall, 1729 patients (median follow-up time, 43 months [25th to 75th percentile, 69.0]) qualified for inclusion. Arterial thrombotic events occurred in 169 (9.7%) patients. Male sex, diabetes, hypercholesterolemia, atrial fibrillation, and personal history of coronary artery disease were associated with increased long-term risk of arterial thrombosis, whereas the use of statins and antithrombotic medications after the acute intracerebral hemorrhage was associated with a reduced risk. The area under the receiver operating characteristic curve of the MUCH score predictive validity was 0.716 (95% CI, 0.56-0.81) for the 0- to 1-year score, 0.672 (95% CI, 0.58-0.73) for the 0- to 5-year score, and 0.744 (95% CI, 0.65-0.81) for the 0- to 10-year score. C statistic for the prediction of events that occur from 0 to 10 years was 0.69 (95% CI, 0.64-0.74). CONCLUSIONS: Intracerebral hemorrhage survivors are at high long-term risk of arterial thrombosis. The MUCH score may serve as a simple tool for risk estimation.

Point-Of-Care low-field MRI in acute Stroke (POCS): protocol for a multicentric prospective open-label study evaluating diagnostic accuracy.

INTRODUCTION: Fast and accurate diagnosis of acute stroke is crucial to timely initiate reperfusion therapies. Conventional high-field (HF) MRI yields the highest accuracy in discriminating early ischaemia from haemorrhages and mimics. Rapid access to HF-MRI is often limited by contraindications or unavailability. Low-field (LF) MRI (<0.5T) can detect several types of brain injury, including ischaemic and haemorrhagic stroke. Implementing LF-MRI in acute stroke care may offer several advantages, including extended applicability, increased safety, faster administration, reduced staffing and costs. This multicentric prospective open-label trial aims to evaluate the diagnostic accuracy of LF-MRI, as a tool to guide treatment decision in acute stroke. METHODS AND ANALYSIS: Consecutive patients accessing the emergency department with suspected stroke dispatch will be recruited at three Italian study units: Azienda Sanitaria Locale (ASL) Abruzzo 1 and 2, Istituto di Ricerca e Cura a Carattere Scientifico (IRCCS) Humanitas Research Hospital. The estimated sample size is 300 patients. Anonymised clinical and LF-MRI data, along with conventional neuroimaging data, will be independently assessed by two external units: Marche Polytechnic University and 'G. Martino' Polyclinic University Hospital. Both units will independently adjudicate the best treatment option, while the latter will provide historical HF-MRI data to develop artificial intelligence algorithms for LF-MRI images interpretation (Free University of Bozen-Bolzano). Agreement with conventional neuroimaging will be evaluated at different time points: hyperacute, acute (24 hours), subacute (72 hours), at discharge and chronic (4 weeks). Further investigations will include feasibility study to develop a mobile stroke unit equipped with LF-MRI and cost-effectiveness analysis. This trial will provide necessary data to validate the use of LF-MRI in acute stroke care. ETHICS AND DISSEMINATION: The study was approved by the Research Ethics Committee of the Abruzzo Region (CEtRA) on 11 May 2023 (approval code: richyvgrg). Results will be disseminated in peer-reviewed journals and presented in academic conferences. TRIAL REGISTRATION NUMBER: NCT05816213; Pre-Results.

Carotid free-floating thrombus stemming from carotid web: co-occurrence of two rare causes of ischemic stroke.

BACKGROUND: Carotid web (CaW) and carotid free-floating thrombus (CFFT) are rare yet critical causes of ischemic stroke in young adults. CASE PRESENTATION: A 54-year-old woman presented with a fluctuating right sensory-motor faciobrachial syndrome. A brain MRI scan revealed multiple small recent asynchronous cortico-subcortical ischemic foci in the vascular territory of the left internal carotid artery. A CT angiography identified a CFFT in the left internal carotid artery arising from an underlying CaW. The patient was treated with excellent clinical outcomes with carotid artery stenting and dual antiplatelet therapy. CONCLUSIONS: We provide a structured pathophysiological rationale connecting CaW and CFFT and highlight pivotal therapeutic implications. Further studies are needed to investigate this relationship and guide assessment and treatment.

Sex differences in ischemic stroke during COVID-19 first outbreak in northern Italy.

INTRODUCTION: COVID-19 pandemic had a great impact on outcome in SARS-CoV-2 positive patients with ischemic stroke during the first wave in Italy. Few data are available on outcome stratified by sex. METHODS: The Italian Society of Hospital Neuroscience conducted a multi-center, retrospective, observational study on neurological complications in COVID-19 patients with ischemic stroke. All the patients admitted from March 1st to April 30th, 2020 in 20 Neurology Units in Northern Italy were recruited. Demographical and clinical features, treatment and outcome data were compared focusing on sex differences. RESULTS: 812 patients with ischemic stroke were enrolled, of whom 129 with COVID-19; males were 53.8%. In-hospital mortality in COVID-19 patients was 35.3% in males and 27.9% in females while 8.5% in male and 5.8% in female patients without COVID-19. SARS-CoV-2 positive patients had a higher frequency of stroke of undetermined etiology, than negative ones (32.8% vs 22.5%; p = 0.02), especially in females compared to males (36.1% vs 27.9%), albeit without statistical significance. Male patients with SARS-CoV-2 were more likely to require cPAP (30.9% vs 14.8%; p = 0.03), endotracheal tube (14.9% vs 3.3%; p = 0.02) and reperfusion strategies (29.4% vs 11.5%; p = 0.01) than females, as well as to have a higher CRP and D-dimer. These elements together with older age, a total anterior circulation stroke and lymphopenia were predictors of a worse outcome. DISCUSSION: Our study detected some differences due to sex in ischemic stroke with and without COVID-19, supporting the possibility to perform sex analyses for SARS-CoV-2 positive patients for a better clinical management.

IV thrombolysis plus thrombectomy versus IV thrombolysis alone for minor stroke with anterior circulation large vessel occlusion from the IRETAS and Italian SITS-ISTR cohorts.

INTRODUCTION: The aim of this study was to compare the outcomes of patients treated with intravenous thrombolysis (IVT) <4.5 h after symptom onset plus mechanical thrombectomy (MT) <6 h with those treated with IVT alone <4.5 h for minor stroke (NIHSS ≤5) with large vessel occlusion (LVO) in the anterior circulation. PATIENTS AND METHODS: Patients enrolled in the Italian Registry of Endovascular Treatment in Acute Stroke (IRETAS) and in the Italian centers included in the SITS-ISTR were analyzed. RESULTS: Among the patients with complete data on 24-h ICH type, 236 received IVT plus MT and 382 received IVT alone. IVT plus MT was significantly associated with unfavorable shift on 24-h ICH types (from no ICH to PH-2) (OR, 2.130; 95% CI, 1.173-3.868; p=0.013) and higher rate of PH (OR, 4.363; 95% CI, 1.579-12.055; p=0.005), sICH per ECASS II definition (OR, 5.527; 95% CI, 1.378-22.167; p=0.016), and sICH per NINDS definition (OR, 3.805; 95% CI, 1.310-11.046; p=0.014). Among the patients with complete data on 3-month mRS score, 226 received IVT plus MT and 262 received IVT alone. No significant difference was reported between IVT plus MT and IVT alone on mRS score 0-1 (72.1% versus 69.1%), mRS score 0-2 (79.6% versus 79%), and death (6.2% versus 6.1%). CONCLUSIONS: Compared with IVT alone, IVT plus MT was associated with unfavorable shift on 24-h ICH types and higher rate of 24-h PH and sICH in patients with minor stroke and LVO in the anterior circulation. However, no difference was reported between the groups on 3-month functional outcome measures.

Risk Profile of Patients with Spontaneous Cervical Artery Dissection.

OBJECTIVE: Epidemiological data to characterize the individual risk profile of patients with spontaneous cervical artery dissection (sCeAD) are rather inconsistent. METHODS AND RESULTS: In the setting of the Italian Project on Stroke in Young Adults Cervical Artery Dissection (IPSYS CeAD), we compared the characteristics of 1,468 patients with sCeAD (mean age = 47.3 ± 11.3 years, men = 56.7%) prospectively recruited at 39 Italian centers with those of 2 control groups, composed of (1) patients whose ischemic stroke was caused by mechanisms other than dissection (non-CeAD IS) selected from the prospective IPSYS registry and Brescia Stroke Registry and (2) stroke-free individuals selected from the staff members of participating hospitals, matched 1:1:1 by sex, age, and race. Compared to stroke-free subjects, patients with sCeAD were more likely to be hypertensive (odds ratio [OR] = 1.65, 95% confidence interval [CI] = 1.37-1.98), to have personal history of migraine with aura (OR = 2.45, 95% CI = 1.74-3.34), without aura (OR = 2.67, 95% CI = 2.15-3.32), and family history of vascular disease in first-degree relatives (OR = 1.69, 95% CI = 1.39-2.05), and less likely to be diabetic (OR = 0.65, 95% CI = 0.47-0.91), hypercholesterolemic (OR = 0.75, 95% CI = 0.62-0.91), and obese (OR = 0.41, 95% CI = 0.31-0.54). Migraine without aura was also associated with sCeAD (OR = 1.81, 95% CI = 1.47-2.22) in comparison with patients with non-CeAD IS. In the subgroup of patients with migraine, patients with sCeAD had higher frequency of migraine attacks and were less likely to take anti-migraine preventive medications, especially beta-blockers, compared with the other groups. INTERPRETATION: The risk of sCeAD is influenced by migraine, especially migraine without aura, more than by other factors, increases with increasing frequency of attacks, and seems to be reduced by migraine preventive medications, namely beta-blockers. ANN NEUROL 2023;94:585-595.

Impact of Preemptive Use of Tocilizumab on Chimeric Antigen Receptor T Cell Outcomes in Non-Hodgkin Lymphoma.

Despite the impressive results of chimeric antigen receptor (CAR) T cell treatment for lymphomas, adverse events such as cytokine release syndrome (CRS), immune effector cell-associated neurotoxicity syndrome (ICANS), and infections are major issues that can lead to intensive care unit (ICU) admission and death. Current guidelines recommend tocilizumab for treating patients with CRS grade (G) ≥2; however, the optimal timing of intervention has yet to be determined. Our institution adopted the preemptive use of tocilizumab in cases of persistent G1 CRS, defined as fever (≥38 °C) for >24 hours. This preemptive tocilizumab treatment aimed to reduce evolution to severe (G≥3) CRS, ICU admission, or death. We report on 48 prospectively collected consecutive patients with non-Hodgkin lymphoma treated with autologous CD19-targeted CAR T cells. In total, 39 patients (81%) developed CRS. CRS started as G1 in 28 patients, as G2 in patients, and as G3 in 1 patient. Tocilizumab was administered in 34 patients, including 23 patients who received "preemptive" tocilizumab and 11 patients who received tocilizumab for G2 or G3 CRS from the onset of symptoms. CRS resolved without worsening severity in 19 patients out of 23 (83%) who received preemptive tocilizumab; 4 patients (17%) progressed from G1 to G2 for the development of hypotension and quickly responded to the introduction of steroids. No patients treated with a preemptive approach developed G3 or G4 CRS. Ten out of 48 patients (21%) were diagnosed with ICANS, including 5 patients with G3 or G4. Six infectious events occurred. The overall ICU admission rate was 19%. ICANS management was the most relevant reason for ICU admission (7 patients), and no patient required ICU to manage CRS. No deaths from CAR-T toxicity were observed. Our data indicate that preemptive tocilizumab use is feasible and useful in reducing severe CRS and CRS-related ICU admission, with no impact on neurotoxicity or infection rate. Therefore, early use of tocilizumab can be considered, especially for patients at high risk of CRS.

Clinical Reasoning: Acute Monocular Vision Loss in a Patient With Ipsilateral Extracranial Chronic Internal Carotid Artery Occlusion.

We report the case of a middle-aged man who presented with acute painless monocular vision loss. His medical history was remarkable for chronic total occlusion of the ipsilateral internal carotid artery (ICA) and a recent carotid endarterectomy (CEA) on the contralateral ICA. In a stepwise multidisciplinary approach assessment, we review the differential diagnosis of acute vision loss and investigate how the patient's intracranial and extracranial hemodynamic reorganization after chronic ICA occlusion may affect the clinical reasoning. Early complications of CEA and the differential diagnosis of new-onset anisocoria are also discussed.

Case report: Reversible punctate inflammatory foci in the corpus callosum: A novel radiological finding of CAR T-cell therapy-related neurotoxicity.

Introduction: Chimeric antigen receptor T-cell therapy-related neurotoxicity is a novel cytokine-mediated neurological syndrome that may present with a broad spectrum of manifestations. Descriptions of novel distinctive features are pivotal to untangling this condition's clinical and instrumental signature in order to inform diagnosis and pathophysiology. Case: A 27-year-old female patient received anti-CD19 CAR T cells for a refractory primary mediastinal B-cell lymphoma. At 6 days after the infusion, she developed mild ideo-motor slowing, dysgraphia, and drowsiness. Despite specific treatment with dexamethasone, her neurological status progressively worsened to a comatose state within 24 h. EEG and CSF analyses were non-specific, showing background slowing and inflammatory findings. Brain MRI revealed multiple focal punctate areas of T2-weighted hyperintensity localized in the body and isthmus of the corpus callosum. Following the administration of high-dose intravenous methylprednisolone, her neurological status resolved within 48 h. Notably, the follow-up brain MRI did not reveal any abnormalities in the corpus callosum, except for a reduction of fractional anisotropy. Conclusion: Reversible punctate inflammatory foci of the body and isthmus of the corpus callosum may represent a novel radiological finding of CAR T-cell therapy-related neurotoxicity.