Written language preservation in glioma patients undergoing awake surgery: The value of tailored intra-operative assessment.

Written language is increasingly important, as contemporary society strongly relies on text-based communication. Nonetheless, in neurosurgical practice, language preservation has classically focused on spoken language. The current study aimed to evaluate the potential role of intra-operative assessments in the preservation of written language skills in glioma patients undergoing awake surgery. It is the first feasibility study to use a standardized and detailed Written language battery in glioma patients undergoing awakening surgery. Reading and spelling were assessed pre- and post-operatively in eleven patients. Intra-operatively, 7 cases underwent written language assessment in addition to spoken object naming. Results show that reading and spelling deficits may arise before and after glioma surgery and that written language may be differently affected than spoken language. In our case series, task-specific preservation of function was obtained in all cases when a specific written language skill was monitored intra-operatively. However, the benefits of intra-operative testing did not always generalize, and non-monitored written language tasks may not be preserved. Hence, when a specific written language skill needs to be preserved, to facilitate return to work and maintain quality of life, results indicate that intra-operative assessment of that skill is advised. An illustrative case report demonstrates how profile analyses can be used pre-operatively to identify cognitive components at risk and intra-operatively to preserve written language abilities in clinical practice.

Understanding language and cognition after brain surgery - Tumour grade, fine-grained assessment tools and, most of all, individualized approach.

Cognitive performance influences the quality of life and survival of people with glioma. Thus, a detailed neuropsychological and language evaluation is essential. In this work, we tested if an analysis of errors in naming can indicate semantic and/or phonological impairments in 87 awake brain surgery patients. Secondly, we explored how language and cognition change after brain tumour resection. Finally, we checked if low-tumour grade had a protective effect on cognition. Our results indicated that naming errors can be useful to monitor semantic and phonological processing, as their number correlated with scores on tasks developed by our team for testing these domains. Secondly, we showed that - although an analysis at a whole group level indicates a decline in language functions - significantly more individual patients improve or remain stable when compared to the ones who declined. Finally, we observed that having LGG, when compared with HGG, favours patients' outcome after surgery, most probably due to brain plasticity mechanisms. We provide new evidence of the importance of applying a broader neuropsychological assessment and an analysis of naming errors in patients with glioma. Our approach may potentially ensure better detection of cognitive deficits and contribute to better postoperative outcomes. Our study also shows that an individualized approach in post-surgical follow-ups can reveal reassuring results showing that significantly more patients remain stable or improve and can be a promising avenue for similar reports. Finally, the study captures that plasticity mechanisms may act as protective in LGG versus HGG after surgery.

Preoperative assessment of eloquence in neurosurgery: a systematic review.

BACKGROUND AND OBJECTIVES: Tumor location and eloquence are two crucial preoperative factors when deciding on the optimal treatment choice in glioma management. Consensus is currently lacking regarding the preoperative assessment and definition of eloquent areas. This systematic review aims to evaluate the existing definitions and assessment methods of eloquent areas that are used in current clinical practice. METHODS: A computer-aided search of Embase, Medline (OvidSP), and Google Scholar was performed to identify relevant studies. This review includes articles describing preoperative definitions of eloquence in the study's Methods section. These definitions were compared and categorized by anatomical structure. Additionally, various techniques to preoperatively assess tumor eloquence were extracted, along with their benefits, drawbacks and ease of use. RESULTS: This review covers 98 articles including 12,714 participants. Evaluation of these studies indicated considerable variability in defining eloquence. Categorization of these definitions yielded a list of 32 brain regions that were considered eloquent. The most commonly used methods to preoperatively determine tumor eloquence were anatomical classification systems and structural MRI, followed by DTI-FT, functional MRI and nTMS. CONCLUSIONS: There were major differences in the definitions and assessment methods of eloquence, and none of them proved to be satisfactory to express eloquence as an objective, quantifiable, preoperative factor to use in glioma decision making. Therefore, we propose the development of a novel, objective, reliable, preoperative classification system to assess eloquence. This should in the future aid neurosurgeons in their preoperative decision making to facilitate personalized treatment paradigms and to improve surgical outcomes.

Research Progress on Amide Proton Transfer Imaging in Preoperative and Postoperative Glioma Assessment.

Amide proton transfer (APT) imaging is a technique that reflects the level of amide proton transport of hydrogen ions in water to peptides and proteins within the tumour by using magnetic resonance imaging without contrast administration. This technique can more accurately reflect the protein level, pH value, and other relevant information within the tumour. Currently, APT is widely used in glioma research. In this paper, we briefly describe the research progress on this technique.

Subpial technique in supratentorial glioma resection: state of the art and analysis of costs and effectiveness in a single institute experience.

BACKGROUND: Many neurosurgeons advocate subpial technique as the best technique to remove supratentorial gliomas. However, few authors clearly defined advantages and features of this technique. The aim of our study is to describe microsurgical subpial technique related to glioma surgery, with regard to its safety and cost effectiveness. METHODS: We analyzed retrospectively all consecutive patients surgically treated for supratentorial glioma from January 2017 to April 2018 at Neurosurgery Department of Neuromed Institute. All patients underwent to surgical glioma resection performing microsurgical subpial technique. Extent of resection and neurological complications were evaluated as primary outcomes; Karnofsky Performance Status and postoperative edema extent were secondary outcomes. Statistical analysis was obtained. RESULTS: The study included 70 patients. Gross Total Removal was obtained in 91.3% of patients with low grade glioma (LGG) and in 81% of patients with high grade glioma. Neurological complications amounted to 34% at early assessment in LGG patients, which were permanent at 3 months in 17% of patients. In high grade glioma patients, neurological complications amounted to 51% at early assessment, which were permanent at 3 months in 25% of them. CONCLUSIONS: We obtained good postoperative results with regard to the extent of tumor resection using this technique. Subpial resection is an effective surgical technique to get a safer and more complete tumor resection. It should be combined with other modern neurosurgical tools such as neuronavigation, ultrasound and cortical mapping to obtain the best tumor resection and functional neurological preservation.

DeepEOR: automated perioperative volumetric assessment of variable grade gliomas using deep learning.

PURPOSE: Volumetric assessments, such as extent of resection (EOR) or residual tumor volume, are essential criterions in glioma resection surgery. Our goal is to develop and validate segmentation machine learning models for pre- and postoperative magnetic resonance imaging scans, allowing us to assess the percentagewise tumor reduction after intracranial surgery for gliomas. METHODS: For the development of the preoperative segmentation model (U-Net), MRI scans of 1053 patients from the Multimodal Brain Tumor Segmentation Challenge (BraTS) 2021 as well as from patients who underwent surgery at the University Hospital in Zurich were used. Subsequently, the model was evaluated on a holdout set containing 285 images from the same sources. The postoperative model was developed using 72 scans and validated on 45 scans obtained from the BraTS 2015 and Zurich dataset. Performance is evaluated using Dice Similarity score, Jaccard coefficient and Hausdorff 95%. RESULTS: We were able to achieve an overall mean Dice Similarity Score of 0.59 and 0.29 on the pre- and postoperative holdout sets, respectively. Our algorithm managed to determine correct EOR in 44.1%. CONCLUSION: Although our models are not suitable for clinical use at this point, the possible applications are vast, going from automated lesion detection to disease progression evaluation. Precise determination of EOR is a challenging task, but we managed to show that deep learning can provide fast and objective estimates.

A new study protocol for in-vivo assessment of tumor diagnosis and microscopic tumor infiltration at the resection cavity in central nervous system tumors by a new miniature confocal endomicroscope (CONVIVO system).

BACKGROUND: Confocal laser technology has been recently suggested as a promising method to obtain near real-time intraoperative histological data. We recently demonstrated the accuracy of a newly designed confocal endomiscroscope (CONVIVO) in offering an intraoperative diagnosis during high-grade gliomas (HGGs) removal in an ex vivo study. With this work we aim to perform a standardized, prospective and blinded-to-histological section study for evaluating the potentiality of CONVIVO in offering in-vivo data regarding histological diagnosis and presence of tumor at margins during resection of central nervous system (CNS) tumors. METHODS: This prospective, observational, standardized, blinded-to-histological section, clinical trial was approved by the institutional review board in Carlo Besta Neurologic Institute IRCCS Foundation in Milan and is expected to last 24 months. 75 patients will be included, with at least 53 of them being HGGs based on the statistical sample size calculation. Main objectives will be the assessing of the concordance of tumor diagnoses between CONVIVO images and frozen section at the center of all tumor subtypes and the evaluation of the accuracy of CONVIVO in the identification of tumor tissue at the margins, compared to standard histology. For this purpose, "virtual biopsies" and physical biopsies will be performed directly on patient tumor tissue and surrounding brain parenchima during tumor resection, comparing the results of CONVIVO analysis and frozen and histological sections. RESULTS: Despite promising preliminary data on ex vivo usefulness of CLE machines are emerging in literature, still few studies are available when looking at in vivo potentiality of CONVIVO. At this regard, this study will be the first work where a standardized, prospective, and blinded-to-histological section CONVIVO analysis will be performed in an in-vivo setting in neuro-oncological surgery. CONCLUSIONS: We hypothesize that this new technique may have a role in offering data regarding presence of tumor tissue, eventually giving an intraoperative diagnosis in neuro-oncological surgery, rendering more fluid the decision-making process in the operating room. Furthermore, the result of this study will provide a solid base for further expanding the clinical applications of confocal machines in neurosurgery.

Launching Awake Craniotomy Technique in a Resource-Limited Center: New Insights into the Patient Experience, Costs, and Long-Term Outcomes and a Narrative Review of the Literature.

BACKGROUND: Awake craniotomy (AC) with brain mapping is a standard surgical technique for the excision of lesions located in eloquent areas. We aimed to assess the clinical challenges, patient experience, costs, and long-term outcomes of AC in a resource-limited setting. METHODS: In this cross-sectional study, electronic documents of 12 patients who underwent AC with functional brain mapping were prospectively collected from August 2017 to October 2020. Patient characteristics, surgical specifications, hospitalization period, intraoperative and postoperative events, functional outcome, patients' satisfaction, costs, and survivals were collected and analyzed. RESULTS: Twelve patients with a median age of 42.5 (interquartile range, 13.5) were enrolled, of whom 8 were male (66.7%), and 9 (75%) were harboring grade 2 glioma. Of the patients, 8.34%, 33.34%, and 58.33% had partial, subtotal, and gross total excision of the tumors, respectively. The intraoperative seizure was the only complication and occurred in 2 cases (16.67%). At 1 year follow-up, none of the patients experienced any neurologic deficit. Eleven patients (91.6%) had a satisfactory opinion about reappearing in the AC. At 38 months follow-up, mortality was 8% for AC group and 25% among the historically matched controls who had surgery under general anesthesia (P = 0.27). Most costs belonged to the neurosurgery team (43%), and the overall expenses were reduced by 13% compared with a putatively well-equipped setting in our country. CONCLUSIONS: In carefully selected individuals, AC with brain mapping for excision of gliomas could be a safe, effective, and affordable strategy in a resource-limited setting and can be successfully performed with satisfactory outcomes.

Targeted metabolomics analyses for brain tumor margin assessment during surgery.

MOTIVATION: Identification and removal of micro-scale residual tumor tissue during brain tumor surgery are key for survival in glioma patients. For this goal, High-Resolution Magic Angle Spinning Nuclear Magnetic Resonance (HRMAS NMR) spectroscopy-based assessment of tumor margins during surgery has been an effective method. However, the time required for metabolite quantification and the need for human experts such as a pathologist to be present during surgery are major bottlenecks of this technique. While machine learning techniques that analyze the NMR spectrum in an untargeted manner (i.e. using the full raw signal) have been shown to effectively automate this feedback mechanism, high dimensional and noisy structure of the NMR signal limits the attained performance. RESULTS: In this study, we show that identifying informative regions in the HRMAS NMR spectrum and using them for tumor margin assessment improves the prediction power. We use the spectra normalized with the ERETIC (electronic reference to access in vivo concentrations) method which uses an external reference signal to calibrate the HRMAS NMR spectrum. We train models to predict quantities of metabolites from annotated regions of this spectrum. Using these predictions for tumor margin assessment provides performance improvements up to 4.6% the Area Under the ROC Curve (AUC-ROC) and 2.8% the Area Under the Precision-Recall Curve (AUC-PR). We validate the importance of various tumor biomarkers and identify a novel region between 7.97 ppm and 8.09 ppm as a new candidate for a glioma biomarker. AVAILABILITY AND IMPLEMENTATION: The code is released at https://github.com/ciceklab/targeted_brain_tumor_margin_assessment. The data underlying this article are available in Zenodo, at https://doi.org/10.5281/zenodo.5781769. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

Assessment of neuropsychological function in brain tumor treatment: a comparison of traditional neuropsychological assessment with app-based cognitive screening.

BACKGROUND: Gliomas are typically considered to cause relatively few neurological impairments. However, cognitive difficulties can arise, for example during treatment, with potential detrimental effects on quality of life. Accurate, reproducible, and accessible cognitive assessment is therefore vital in understanding the effects of both tumor and treatments. Our aim is to compare traditional neuropsychological assessment with an app-based cognitive screening tool in patients with glioma before and after surgical resection. Our hypotheses were that cognitive impairments would be apparent, even in a young and high functioning cohort, and that app-based cognitive screening would complement traditional neuropsychological assessment. METHODS: Seventeen patients with diffuse gliomas completed a traditional neuropsychological assessment and an app-based touchscreen tablet assessment pre- and post-operatively. The app assessment was also conducted at 3- and 12-month follow-up. Impairment rates, mean performance, and pre- and post-operative changes were compared using standardized Z-scores. RESULTS: Approximately 2-3 h of traditional assessment indicated an average of 2.88 cognitive impairments per patient, while the 30-min screen indicated 1.18. As might be expected, traditional assessment using multiple items across the difficulty range proved more sensitive than brief screening measures in areas such as memory and attention. However, the capacity of the screening app to capture reaction times enhanced its sensitivity, relative to traditional assessment, in the area of non-verbal function. Where there was overlap between the two assessments, for example digit span tasks, the results were broadly equivalent. CONCLUSIONS: Cognitive impairments were common in this sample and app-based screening complemented traditional neuropsychological assessment. Implications for clinical assessment and follow-up are discussed.

Deep learning-based automatic tumor burden assessment of pediatric high-grade gliomas, medulloblastomas, and other leptomeningeal seeding tumors.

BACKGROUND: Longitudinal measurement of tumor burden with magnetic resonance imaging (MRI) is an essential component of response assessment in pediatric brain tumors. We developed a fully automated pipeline for the segmentation of tumors in pediatric high-grade gliomas, medulloblastomas, and leptomeningeal seeding tumors. We further developed an algorithm for automatic 2D and volumetric size measurement of tumors. METHODS: The preoperative and postoperative cohorts were randomly split into training and testing sets in a 4:1 ratio. A 3D U-Net neural network was trained to automatically segment the tumor on T1 contrast-enhanced and T2/FLAIR images. The product of the maximum bidimensional diameters according to the RAPNO (Response Assessment in Pediatric Neuro-Oncology) criteria (AutoRAPNO) was determined. Performance was compared to that of 2 expert human raters who performed assessments independently. Volumetric measurements of predicted and expert segmentations were computationally derived and compared. RESULTS: A total of 794 preoperative MRIs from 794 patients and 1003 postoperative MRIs from 122 patients were included. There was excellent agreement of volumes between preoperative and postoperative predicted and manual segmentations, with intraclass correlation coefficients (ICCs) of 0.912 and 0.960 for the 2 preoperative and 0.947 and 0.896 for the 2 postoperative models. There was high agreement between AutoRAPNO scores on predicted segmentations and manually calculated scores based on manual segmentations (Rater 2 ICC = 0.909; Rater 3 ICC = 0.851). Lastly, the performance of AutoRAPNO was superior in repeatability to that of human raters for MRIs with multiple lesions. CONCLUSIONS: Our automated deep learning pipeline demonstrates potential utility for response assessment in pediatric brain tumors. The tool should be further validated in prospective studies.

Avaliação da linguagem em cirurgia de craniotomia com paciente acordado: relato de caso / Language assessment in awake craniotomy: case report

RESUMO O objetivo deste relato foi descrever o caso de um paciente submetido à craniotomia, acordado, para a ressecção neurocirúrgica de um glioma e a avaliação linguística pré-operatória, intraoperatória e pós-operatória. Paciente do gênero masculino, 27 anos, escolaridade nível superior incompleto, apresentando vômitos, confusão mental e crise convulsiva tônico-clônica. Após a avaliação do paciente pela equipe e devidas orientações pré-operatórias, a proposta de excisão da lesão em estado de vigília foi esclarecida e aceita. Ao iniciar o procedimento, os campos foram ajustados para manter as vias aéreas e os olhos acessíveis para mapeamento com estimulação elétrica e avaliação da linguagem no período intraoperatório. Devido à localização do tumor próximo à área motora da fala, foram propostas tarefas para a avaliação da linguagem em quatro momentos: pré-operatório, intraoperatório, pós-operatório imediato e pós-operatório mediato. As habilidades linguísticas testadas nas quatro avaliações foram: compreensão e expressão da linguagem oral, transposição linguística, linguagem associativa, nomeação, discriminação visual, fluência e organização da sintaxe. Com o objetivo de controlar e eliminar o efeito de aprendizagem da testagem, foram solicitadas as mesmas tarefas, porém, com diferentes conteúdos para a testagem das habilidades nas quatro fases. A cirurgia com o paciente acordado permitiu a ressecção completa e segura do tumor, sem prejuízo motor ou linguístico. O engajamento da equipe, a interação interdisciplinar e o planejamento cirúrgico detalhado constituem um pilar para o bom resultado de um procedimento tão complexo e delicado.

ABSTRACT The purpose of this report is to describe the case of a patient who underwent awake craniotomy for neurosurgical resection of a glioma and pre, intra and postoperative linguistic assessment. Male patient, 27 years old, incomplete higher education presenting vomiting, mental confusion and tonic-clonic seizures. After the evaluation of the patient by the team and due preoperative guidance, the proposal of excision of the lesion while awake was clarified and accepted. At the start of the procedure, the fields were adjusted to keep the airway and eyes accessible for mapping with electrical stimulation and intraoperative language assessment. Due to the location of the tumor close to the speech motor area, tasks were proposed for the assessment of language in four moments: preoperative, intraoperative, immediate postoperative and mediate postoperative. The language skills tested in the four assessments were: comprehension and expression of oral language, linguistic transposition, associative language, naming, visual discrimination, fluency and syntax organization. In order to control and eliminate the learning effect of testing, the same tasks were requested, but with different contents for testing skills in the four phases. Surgery with the patient awake allowed the complete and safe resection of the tumor, without motor or linguistic damage to the patient. Team engagement, interdisciplinary interaction and detailed surgical planning constitute the pillar for the good result of such a complex and delicate procedure

[Magnetic resonance relaxometry in high-grade glioma subregion assessment - neuroimaging and morphological correlates]. / Magnitno-rezonansnaya relaksometriya v otsenke subregionov gliom golovnogo mozga vysokoi stepeni zlokachestvennosti ­ neirovizualizatsionnye i morfologicheskie korrelyaty.

OBJECTIVE: To analyze the differences of high-grade glioma subregions using magnetic resonance relaxometry with compilation of images (MAGiC) and arterial spin labeling (ASL), as well as to compare quantitative measurements of these techniques with morphological data. MATERIAL AND METHODS: The study enrolled 35 patients with newly diagnosed supratentorial gliomas (23 - grade IV, 12 - grade III). We measured relaxometric values (T1, T2, proton density), tumor blood flow (TBF) in glioma subregions and normal-appearing brain matter. Neuronavigation was intraoperatively used to obtain tissue samples from active tumor growth zone, perifocal infiltrative edema zone and adjacent brain matter along surgical approach. RESULTS: ASL perfusion revealed higher tumor blood flow (TBF) in active tumor growth region compared to perifocal infiltrative edema zone (p<0.01). Relaxometric values (T1, T2, proton density) in perifocal zone were higher (p<0.01) compared to adjacent intact white matter along surgical approach. However, there were no differences in TBF between these zones. Proton density in tumor-adjacent intact white matter was higher (p<0.01) compared to normal-appearing white matter in ipsilateral hemisphere. There was inverse correlation between T2 and TBF in active tumor growth zone (Spearman rank R= -0.58; p=0.0016). We found inverse correlation between T2 and Ki67 proliferative index and direct correlation between TBF and Ki67 in this zone. Nevertheless, these relationships were insignificant after multiple test adjustment. CONCLUSION: Our study advocates for complementary power of ASL perfusion and MR relaxometry in assessment of high-grade brain glioma subregions. More malignant tumor zones tend to have higher TBF and shorter T2. Further investigation is needed to prove the capability of MAGiC to reveal foci of increased relaxometric values in tumor-adjacent normal-appearing white matter.

Label-free imaging of human brain tissue at subcellular resolution for potential rapid intra-operative assessment of glioma surgery.

Background: Frozen section and smear preparation are the current standard for intraoperative histopathology during cancer surgery. However, these methods are time-consuming and subject to limited sampling. Multiphoton microscopy (MPM) is a high-resolution non-destructive imaging technique capable of optical sectioning in real time with subcellular resolution. In this report, we systematically investigated the feasibility and translation potential of MPM for rapid histopathological assessment of label- and processing-free surgical specimens. Methods: We employed a customized MPM platform to capture architectural and cytological features of biological tissues based on two-photon excited NADH and FAD autofluorescence and second harmonic generation from collagen. Infiltrating glioma, an aggressive disease that requires subcellular resolution for definitive characterization during surgery, was chosen as an example for this validation study. MPM images were collected from resected brain specimens of 19 patients and correlated with histopathology. Deep learning was introduced to assist with image feature recognition. Results: MPM robustly captures diagnostic features of glioma including increased cellularity, cellular and nuclear pleomorphism, microvascular proliferation, necrosis, and collagen deposition. Preliminary application of deep learning to MPM images achieves high accuracy in distinguishing gray from white matter and cancer from non-cancer. We also demonstrate the ability to obtain such images from intact brain tissue with a multiphoton endomicroscope for intraoperative application. Conclusion: Multiphoton imaging correlates well with histopathology and is a promising tool for characterization of cancer and delineation of infiltration within seconds during brain surgery.

Intraoperative BOLD-fMRI Cerebrovascular Reactivity Assessment.

Blood oxygenation-level dependent cerebrovascular reactivity (BOLD-CVR) has gained attention in recent years as an effective way to investigate CVR, a measure of the hemodynamic state of the brain, with high spatial and temporal resolution. An association between impaired CVR and diverse pathologies has been observed, especially in ischemic cerebrovascular diseases and brain gliomas. The ability to obtain this information intraoperatively is novel and has not been widely tested. We report our first experience with this intraoperative technique in vascular and oncologic neurosurgical patients, discuss the results of its feasibility, and the possible developments of the intraoperative employment of BOLD-CVR.

Updated perspectives on awake neurosurgery with cognitive and emotional assessment for patients with low-grade gliomas.

Introduction: Thanks to early extensive surgical resection combined with medical oncological therapies, life expectancy dramatically increased in low-grade glioma (LGG), with an overall survival currently over 15 years. Therefore, patients should be able to maintain valuable family and socio-professional activities.Areas covered: For many decades, cognitive and emotional aspects were neglected by surgical and medical neurooncologists. The goal of surgery was to avoid hemiplegia and/or aphasia, with no considerations regarding behavior. However, because LGG patients live longer, they must be cognitively and affectively able to make long-term projects. Preservation of higher-order functions should be considered systematically in LGG surgery by means of awake cognitive/emotional mapping and monitoring.Expert opinion: The aim is to incorporate recent advances in neurosciences, which proposed revisited models of cerebral processing relying on a meta-network perspective, into the pre-, intra- and postoperative procedure. In this connectomal approach, brain functions result from complex interactions within and between neural networks. This improved understanding of a constant instability of the neural system allows a better cognitive/emotional assessment before and after each treatment over years, in order to preserve personality and adaptive behavior for each LGG patient, based on his/her own definition of quality of life. It is time to create oncological neurosciences.

Predictive Model and Online Calculator for Discharge Disposition in Brain Tumor Patients.

BACKGROUND: In the era of value-based payment models, it is imperative for neurosurgeons to eliminate inefficiencies and provide high-quality care. Discharge disposition is a relevant consideration with clinical and economic ramifications in brain tumor patients. We developed a predictive model and online calculator for postoperative non-home discharge disposition in brain tumor patients that can be incorporated into preoperative workflows. METHODS: We reviewed all brain tumor patients at our institution from 2017 to 2019. A predictive model of discharge disposition containing preoperatively available variables was developed using stepwise multivariable logistic regression. Model performance was assessed using receiver operating characteristic curves and calibration curves. Internal validation was performed using bootstrapping with 2000 samples. RESULTS: Our cohort included 2335 patients who underwent 2586 surgeries with a 16% non-home discharge rate. Significant predictors of non-home discharge were age >60 years (odds ratio [OR], 2.02), African American (OR, 1.73) or Asian (OR, 2.05) race, unmarried status (OR, 1.48), Medicaid insurance (OR, 1.90), admission from another health care facility (OR, 2.30), higher 5-factor modified frailty index (OR, 1.61 for 5-factor modified frailty index ≥2), and lower Karnofsky Performance Status (increasing OR with each 10-point decrease in Karnofsky Performance Status). The model was well calibrated and had excellent discrimination (optimism-corrected C-statistic, 0.82). An open-access calculator was deployed (https://neurooncsurgery.shinyapps.io/discharge_calc/). CONCLUSIONS: A strongly performing predictive model and online calculator for non-home discharge disposition in brain tumor patients was developed. With further validation, this tool may facilitate more efficient discharge planning, with consequent improvements in quality and value of care for brain tumor patients.

Assessment of Amide proton transfer weighted (APTw) MRI for pre-surgical prediction of final diagnosis in gliomas.

PURPOSE: Radiological assessment of primary brain neoplasms, both high (HGG) and low grade tumors (LGG), based on contrast-enhancement alone can be inaccurate. We evaluated the radiological value of amide proton transfer weighted (APTw) MRI as an imaging complement for pre-surgical radiological diagnosis of brain tumors. METHODS: Twenty-six patients were evaluated prospectively; (22 males, 4 females, mean age 55 years, range 26-76 years) underwent MRI at 3T using T1-MPRAGE pre- and post-contrast administration, conventional T2w, FLAIR, and APTw imaging pre-surgically for suspected primary/secondary brain tumor. Assessment of the additional value of APTw imaging compared to conventional MRI for correct pre-surgical brain tumor diagnosis. The initial radiological pre-operative diagnosis was based on the conventional contrast-enhanced MR images. The range, minimum, maximum, and mean APTw signals were evaluated. Conventional normality testing was performed; with boxplots/outliers/skewness/kurtosis and a Shapiro-Wilk's test. Mann-Whitney U for analysis of significance for mean/max/min and range APTw signal. A logistic regression model was constructed for mean, max, range and Receiver Operating Characteristic (ROC) curves calculated for individual and combined APTw signals. RESULTS: Conventional radiological diagnosis prior to surgery/biopsy was HGG (8 patients), LGG (12 patients), and metastasis (6 patients). Using the mean and maximum: APTw signal would have changed the pre-operative evaluation the diagnosis in 8 of 22 patients (two LGGs excluded, two METs excluded). Using a cut off value of >2.0% for mean APTw signal integral, 4 of the 12 radiologically suspected LGG would have been diagnosed as high grade glioma, which was confirmed by histopathological diagnosis. APTw mean of >2.0% and max >2.48% outperformed four separate clinical radiological assessments of tumor type, P-values = .004 and = .002, respectively. CONCLUSIONS: Using APTw-images as part of the daily clinical pre-operative radiological evaluation may improve diagnostic precision in differentiating LGGs from HGGs, with potential improvement of patient management and treatment.

Machine learning assisted intraoperative assessment of brain tumor margins using HRMAS NMR spectroscopy.

Complete resection of the tumor is important for survival in glioma patients. Even if the gross total resection was achieved, left-over micro-scale tissue in the excision cavity risks recurrence. High Resolution Magic Angle Spinning Nuclear Magnetic Resonance (HRMAS NMR) technique can distinguish healthy and malign tissue efficiently using peak intensities of biomarker metabolites. The method is fast, sensitive and can work with small and unprocessed samples, which makes it a good fit for real-time analysis during surgery. However, only a targeted analysis for the existence of known tumor biomarkers can be made and this requires a technician with chemistry background, and a pathologist with knowledge on tumor metabolism to be present during surgery. Here, we show that we can accurately perform this analysis in real-time and can analyze the full spectrum in an untargeted fashion using machine learning. We work on a new and large HRMAS NMR dataset of glioma and control samples (n = 565), which are also labeled with a quantitative pathology analysis. Our results show that a random forest based approach can distinguish samples with tumor cells and controls accurately and effectively with a median AUC of 85.6% and AUPR of 93.4%. We also show that we can further distinguish benign and malignant samples with a median AUC of 87.1% and AUPR of 96.1%. We analyze the feature (peak) importance for classification to interpret the results of the classifier. We validate that known malignancy biomarkers such as creatine and 2-hydroxyglutarate play an important role in distinguishing tumor and normal cells and suggest new biomarker regions. The code is released at http://github.com/ciceklab/HRMAS_NC.