Prognostic factors for survival in adult patients with recurrent glioblastoma: a decision-tree-based model

Imaging treatment efficacy of repeated photodynamic therapy in glioblastoma using chemical exchange transfer saturation MRI

PURPOSE

To observe the tumor responses during photodynamic therapy in a murine glioblastoma model using chemical exchange saturation transfer (CEST) MRI and to compare the treatment effectiveness between single photodynamic therapy (sPDT) and repeated PDT (rePDT).

METHODS

After tumor cell implantation in NSG mouse brain (n = 27), mice were subjected to four PDT sessions (rePDT), sPDT after the administration of 5-aminolevulinic acid 6 h before each session, and a non-PDT session (control). A 630-nm LED light was used to effectuate PDT. After 24 h for each PDT session, T2-weighted and CEST MRI were performed over 7 days.

RESULTS

We observed that rePDT resulted in a continuous suppression of tumors according to T2-weighted images; thus, the tumor volume was the smallest among three groups on Day 7. Both CEST contrasts at 3.5 ppm (amide proton transfer, APT) and- $$ - $$ 3.5 ppm (relayed nuclear Overhauser enhancement, rNOE) in the rePDT group were significantly lower (p < 0.05) than those in the control group starting from Day 5, which corresponds to lower protein and cellularity in tumors in the rePDT group, respectively. CEST contrast decreased by 17.9% at 3.5 ppm and 11.3% at- $$ - $$ 3.5 ppm for rePDT group. This was validated by histology, where we observed moderate correlations between APT with cell proliferation (R = 0.730, p < 0.01) and cell apoptosis (R = 0.715, p < 0.05) and moderate correlation between rNOE with cellularity (R = 0.796, p < 0.01).

CONCLUSIONS

rePDT has a better effect in tumor growth suppression when compared with sPDT, and CEST could be a robust and noninvasive mean to assess the molecular changes related to treatment efficacy.

Machine Learning to Predict Mortality in Older Patients With Cancer: Development and External Validation of the Geriatric Cancer Scoring System Using Two Large French Cohorts

PURPOSE

Establishing an accurate prognosis remains challenging in older patients with cancer because of the population's heterogeneity and the current predictive models' reduced ability to capture the complex interactions between oncologic and geriatric predictors. We aim to develop and externally validate a new predictive score (the Geriatric Cancer Scoring System [GCSS]) to refine individualized prognosis for older patients with cancer during the first year after a geriatric assessment (GA).

MATERIALS AND METHODS

Data were collected from two French prospective multicenter cohorts of patients with cancer 70 years and older, referred for GA: ELCAPA (training set January 2007-March 2016) and ONCODAGE (validation set August 2008-March 2010). Candidate predictors included baseline oncologic and geriatric factors and routine biomarkers. We built predictive models using Cox regression, single decision tree (DT), and random survival forest (RSF) methods, comparing their predictive performance for 3-, 6-, and 12-month mortalities by computing time-dependent area under the receiver operator curve (tAUC).

RESULTS

A total of 2,012 and 1,397 patients were included in the training and validation set, respectively (mean age: 81 ± 6 years/78 ± 5 years; women: 47%/70%; metastatic cancer: 50%/34%; 12-month mortality: 43%/16%). Tumor site/metastatic status, cancer treatment, weight loss, ≥five prescription drugs, impaired functional status and mobility, abnormal G-8 score, low creatinine clearance, and elevated C-reactive protein (CRP)/albumin were identified as relevant predictors in the Cox model. DT and RSF identified more complex combinations of features, with G-8 score, tumor site/metastatic status, and CRP/albumin ratio contributing most to the predictions. The RSF approach gave the highest tAUC (12 months: 0.87 [RSF], 0.82 [Cox], 0.82 [DT]) and was retained as the final model.

CONCLUSION

The GCSS on the basis of a machine learning approach applied to two large French cohorts gave an accurate externally validated mortality prediction. The GCSS might improve decision making and counseling in older patients with cancer referred for pretherapeutic GA. GCSS's generalizability must now be confirmed in an international setting.

Survival prediction of glioblastoma patients using machine learning and deep learning: a systematic review

Glioblastoma Multiforme (GBM), classified as a grade IV glioma by the World Health Organization (WHO), is a prevalent and notably aggressive form of brain tumor derived from glial cells. It stands as one of the most severe forms of primary brain cancer in humans. The median survival time of GBM patients is only 12-15 months, making it the most lethal type of brain tumor. Every year, about 200,000 people worldwide succumb to this disease. GBM is also highly heterogeneous, meaning that its characteristics and behavior vary widely among different patients. This leads to different outcomes and survival times for each individual. Predicting the survival of GBM patients accurately can have multiple benefits. It can enable optimal and personalized treatment planning based on the patient's condition and prognosis. It can also support the patients and their families to cope with the possible outcomes and make informed decisions about their care and quality of life. Furthermore, it can assist the researchers and scientists to discover the most relevant biomarkers, features, and mechanisms of the disease and to design more effective and personalized therapies. Artificial intelligence methods, such as machine learning and deep learning, have been widely applied to survival prediction in various fields, such as breast cancer, lung cancer, gastric cancer, cervical cancer, liver cancer, prostate cancer, and covid 19. This systematic review summarizes the current state-of-the-art methods for predicting glioblastoma survival using different types of input data, such as clinical features, molecular markers, imaging features, radiomics features, omics data or a combination of them. Following PRISMA guidelines, we searched databases from 2015 to 2024, reviewing 107 articles meeting our criteria. We analyzed the data sources, methods, performance metrics and outcomes of the studies. We found that random forest was the most popular method, and a combination of radiomics and clinical data was the most common input data.

Do clinical outcomes in individuals with malignant gliomas differ between sexes?

Introduction

Sex-related differences in the epidemiology of malignant gliomas are acknowledged; however, information regarding their clinical characteristics and outcomes after surgery is limited.

Research question

To identify sex-specific differences of all patients with high-grade glioma at our institution and assessed clinical outcomes and prognostic factors.

Material and methods

This single-center study included those who underwent surgery for malignant gliomas between 2010 and 2020. Categorical, normally distributed, and skewed continuous variables were compared between men and women using the chi-square test, independent samples t-test, and Mann-Whitney U test, respectively. Survival was calculated using the log-rank and Kaplan-Meier methods.

Results

In total, 621 patients with WHO grade IV gliomas were identified (370 (59.58%) male). Men were significantly younger, underwent surgery faster after imaging diagnosis, and had a slightly higher surgical complications incidence than women. Women reported a worse preoperative performance status. Multivariate analysis showed that sex did not affect survival, surgical complications, nicotine or alcohol abuse, or preoperative tumor volume. Age, Karnofsky performance status, neurosurgical resection, and adjuvant radiotherapy with temozolomide showed a survival advantage.

Discussion and conclusions

Men are diagnosed with malignant glioma at a younger age than women; however, no advantage in clinical outcomes was observed. No sex-related differences were observed.

Glutathione peroxidase 4 overexpression induces anomalous subdiffusion and impairs glioblastoma cell growth

Glioblastoma tumors are the most common and aggressive adult central nervous system malignancy. Nearly all patients experience disease progression, which significantly contributes to disease mortality. Recently, it has been suggested that recurrent tumors may be characterized by a ferroptosis-prone phenotype with a significant decrease in glutathione peroxidase 4 (GPx4) expression. This led to the hypothesis that GPx4 expression negatively influences GBM cell growth. This study utilizes a doxycycline inducible GPx4 overexpression model to test this hypothesis. Consistently, the overexpression of GPx4 significantly impairs cell growth and colony formation while also causing an accumulation of cells in G1/G0 phase of the cell cycle. From a biophysical perspective, GPx4 overexpressing cells have significantly greater surface area, increased Young's modulus, and experience anomalous sub-diffusion as opposed to normal diffusion associated with Brownian motion. Moreover, analysis of patient derived GBM cells reveal that cell growth rates, plating efficiency, and Young's modulus are all inversely proportional to GPx4 expression. Therefore, GPx4 appears to be a biophysical regulator of GBM cell growth that warrants further mechanistic investigation in its role in GBM progression.

Predicting Postoperative Motor Function After Brain Tumor Resection With Motor Evoked Potential Monitoring Using Decision Tree Analysis

Background Motor evoked potential (MEP) monitoring is a commonly employed method in neurosurgery to prevent postoperative motor dysfunction. However, it has low prediction accuracy for postoperative paralysis. This study aimed to develop a decision tree (DT) model for predicting postoperative motor function using MEP monitoring data. Methodology In this retrospective cohort study, we used datasets, comprising 14 variables including MEP amplitudes, obtained from 125 patients who underwent brain tumor resection with intraoperative MEP monitoring at our hospital. Prediction models were developed using DT and receiver operating characteristic (ROC) curve analyses. Model performance was assessed for accuracy, sensitivity, specificity, kappa (κ) coefficient, and area under the ROC curve (AUC) for internal and external validation. For the external validation of the classification model, we retrospectively collected data from an additional 28 patients who underwent brain tumor surgery with MEP monitoring. Results The amplitude of the last measured MEP and amplitude ratio were independent predictors of outcomes. The DT model achieved an accuracy of 0.921, sensitivity of 0.917, specificity of 0.923, and AUC of 0.931 using the internal test. In comparison, the ROC curve based on the amplitude of the last measured MEP achieved a sensitivity of 0.875, specificity of 0.906, and AUC of 0.941. External validation was performed and the DT model was superior to prediction by cutoff values from ROC curves in terms of accuracy, sensitivity, specificity, and κ coefficient. Conclusions Our study suggested the usefulness of DT modeling for predicting postoperative paralysis. However, this study has several limitations, such as the retrospective design and small sample size of the validation dataset. Nonetheless, the DT modeling presented in this study might be applicable to surgeries using MEP monitoring and is expected to contribute to devising treatment strategies by predicting postoperative motor function in various patients.

Glioblastoma and Immune Checkpoint Inhibitors: A Glance at Available Treatment Options and Future Directions

Glioblastoma is known to be one of the most aggressive and fatal human cancers, with a poor prognosis and resistance to standard treatments. In the last few years, many solid tumor treatments have been revolutionized with the help of immunotherapy. However, this type of treatment has failed to improve the results in glioblastoma patients. Effective immunotherapeutic strategies may be developed after understanding how glioblastoma achieves tumor-mediated immune suppression in both local and systemic landscapes. Biomarkers may help identify patients most likely to benefit from this type of treatment. In this review, we discuss the use of immunotherapy in glioblastoma, with an emphasis on immune checkpoint inhibitors and the factors that influence clinical response. A Pubmed data search was performed for all existing information regarding immune checkpoint inhibitors used for the treatment of glioblastoma. All data evaluating the ongoing clinical trials involving the use of ICIs either as monotherapy or in combination with other drugs was compiled and analyzed.

Do presenting symptoms predict treatment decisions and survival in glioblastoma? Real-world data from 1458 patients in the Swedish brain tumor registry

Background

Glioblastoma is the most common malignant brain tumor in adults. Non-invasive clinical parameters could play a crucial role in treatment planning and serve as predictors of patient survival. Our register-based real-life study aimed to investigate the prognostic value of presenting symptoms.

Methods

Data on presenting symptoms and survival, as well as known prognostic factors, were retrieved for all glioblastoma patients in Sweden registered in the Swedish Brain Tumor Registry between 2018 and 2021. The prognostic impact of different presenting symptoms was calculated using the Cox proportional hazard model.

Results

Data from 1458 adults with pathologically verified IDH wild-type glioblastoma were analyzed. Median survival time was 345 days. The 2-year survival rate was 21.5%. Registered presenting symptoms were focal neurological deficits, cognitive dysfunction, headache, epilepsy, signs of raised intracranial pressure, and cranial nerve symptoms, with some patients having multiple symptoms. Patients with initial cognitive dysfunction had significantly shorter survival than patients without; 265 days (245-285) vs. 409 days (365-453; P < .001). The reduced survival remained after Cox regression adjusting for known prognostic factors. Patients presenting with seizures and patients with headaches had significantly longer overall survival compared to patients without these symptoms, but the difference was not retained in multivariate analysis. Patients with cognitive deficits were less likely to have radical surgery and to receive extensive anti-neoplastic nonsurgical treatment.

Conclusions

This extensive real-life study reveals that initial cognitive impairment acts as an independent negative predictive factor for treatment decisions and adversely affects survival outcomes in glioblastoma patients.

Impact of body composition parameters, age, and tumor staging on gastric cancer prognosis

BACKGROUND

Research studies on gastric cancer have not investigated the combined impact of body composition, age, and tumor staging on gastric cancer prognosis. To address this gap, we used machine learning methods to develop reliable prediction models for gastric cancer.

METHODS

This study included 1,132 gastric cancer patients, with preoperative body composition and clinical parameters recorded, analyzed using Cox regression and machine learning models.

RESULTS

The multivariate analysis revealed that several factors were associated with recurrence-free survival (RFS) and overall survival (OS) in gastric cancer. These factors included age (≥65 years), tumor-node-metastasis (TNM) staging, low muscle attenuation (MA), low skeletal muscle index (SMI), and low visceral to subcutaneous adipose tissue area ratios (VSR). The decision tree analysis for RFS identified six subgroups, with the TNM staging I, II combined with high MA subgroup showing the most favorable prognosis and the TNM staging III combined with low MA subgroup exhibiting the poorest prognosis. For OS, the decision tree analysis identified seven subgroups, with the subgroup featuring high MA combined with TNM staging I, II showing the best prognosis and the subgroup with low MA, TNM staging II, III, low SMI, and age ≥65 years associated with the worst prognosis.

CONCLUSION

Cox regression identified key factors associated with gastric cancer prognosis, and decision tree analysis determined prognoses across different risk factor subgroups. Our study highlights that the combined use of these methods can enhance intervention planning and clinical decision-making in gastric cancer.

Recursive partitioning analysis for survival stratification and early imaging prediction of molecular biomarker in glioma patients

BACKGROUND

Glioma is the most common primary brain tumor with high mortality and disability rates. Recent studies have highlighted the significant prognostic consequences of subtyping molecular pathological markers using tumor samples, such as IDH, 1p/19q, and TERT. However, the relative importance of individual markers or marker combinations in affecting patient survival remains unclear. Moreover, the high cost and reliance on postoperative tumor samples hinder the widespread use of these molecular markers in clinical practice, particularly during the preoperative period. We aim to identify the most prominent molecular biomarker combination that affects patient survival and develop a preoperative MRI-based predictive model and clinical scoring system for this combination.

METHODS

A cohort dataset of 2,879 patients was compiled for survival risk stratification. In a subset of 238 patients, recursive partitioning analysis (RPA) was applied to create a survival subgroup framework based on molecular markers. We then collected MRI data and applied Visually Accessible Rembrandt Images (VASARI) features to construct predictive models and clinical scoring systems.

RESULTS

The RPA delineated four survival groups primarily defined by the status of IDH and TERT mutations. Predictive models incorporating VASARI features and clinical data achieved AUC values of 0.85 for IDH and 0.82 for TERT mutations. Nomogram-based scoring systems were also formulated to facilitate clinical application.

CONCLUSIONS

The combination of IDH-TERT mutation status alone can identify the most distinct survival differences in glioma patients. The predictive model based on preoperative MRI features, supported by clinical assessments, offers a reliable method for early molecular mutation prediction and constitutes a valuable scoring tool for clinicians in guiding treatment strategies.

Immunotherapeutic Strategies for the Treatment of Glioblastoma: Current Challenges and Future Perspectives

Despite decades of research and the best up-to-date treatments, grade 4 Glioblastoma (GBM) remains uniformly fatal with a patient median overall survival of less than 2 years. Recent advances in immunotherapy have reignited interest in utilizing immunological approaches to fight cancer. However, current immunotherapies have so far not met the anticipated expectations, achieving modest results in their journey from bench to bedside for the treatment of GBM. Understanding the intrinsic features of GBM is of crucial importance for the development of effective antitumoral strategies to improve patient life expectancy and conditions. In this review, we provide a comprehensive overview of the distinctive characteristics of GBM that significantly influence current conventional therapies and immune-based approaches. Moreover, we present an overview of the immunotherapeutic strategies currently undergoing clinical evaluation for GBM treatment, with a specific emphasis on those advancing to phase 3 clinical studies. These encompass immune checkpoint inhibitors, adoptive T cell therapies, vaccination strategies (i.e., RNA-, DNA-, and peptide-based vaccines), and virus-based approaches. Finally, we explore novel innovative strategies and future prospects in the field of immunotherapy for GBM.

Development of preoperative and postoperative models to predict recurrence in postoperative glioma patients: a longitudinal cohort study

BACKGROUND

Glioma recurrence, subsequent to maximal safe resection, remains a pivotal challenge. This study aimed to identify key clinical predictors influencing recurrence and develop predictive models to enhance neurological diagnostics and therapeutic strategies.

METHODS

This longitudinal cohort study with a substantial sample size (n = 2825) included patients with non-recurrent glioma who were pathologically diagnosed and had undergone initial surgical resection between 2010 and 2018. Logistic regression models and stratified Cox proportional hazards models were established with the top 15 clinical variables significantly influencing outcomes screened by the least absolute shrinkage and selection operator (LASSO) method. Preoperative and postoperative models predicting short-term (within 6 months) postoperative recurrence in glioma patients were developed to explore the risk factors associated with short- and long-term recurrence in glioma patients.

RESULTS

Preoperative and postoperative logistic models predicting short-term recurrence had accuracies of 0.78 and 0.87, respectively. A range of biological and early symptomatic characteristics linked to short- and long-term recurrence have been pinpointed. Age, headache, muscle weakness, tumor location and Karnofsky score represented significant odd ratios (t > 2.65, p < 0.01) in the preoperative model, while age, WHO grade 4 and chemotherapy or radiotherapy treatments (t > 4.12, p < 0.0001) were most significant in the postoperative period. Postoperative predictive models specifically targeting the glioblastoma and IDH wildtype subgroups were also performed, with an AUC of 0.76 and 0.80, respectively. The 50 combinations of distinct risk factors accommodate diverse recurrence risks among glioma patients, and the nomograms visualizes the results for clinical practice. A stratified Cox model identified many prognostic factors for long-term recurrence, thereby facilitating the enhanced formulation of perioperative care plans for patients, and glioblastoma patients displayed a median progression-free survival (PFS) of only 11 months.

CONCLUSION

The constructed preoperative and postoperative models reliably predicted short-term postoperative glioma recurrence in a substantial patient cohort. The combinations risk factors and nomograms enhance the operability of personalized therapeutic strategies and care regimens. Particular emphasis should be placed on patients with recurrence within six months post-surgery, and the corresponding treatment strategies require comprehensive clinical investigation.

Integrating imaging and genomic data for the discovery of distinct glioblastoma subtypes: a joint learning approach

Glioblastoma is a highly heterogeneous disease, with variations observed at both phenotypical and molecular levels. Personalized therapies would be facilitated by non-invasive in vivo approaches for characterizing this heterogeneity. In this study, we developed unsupervised joint machine learning between radiomic and genomic data, thereby identifying distinct glioblastoma subtypes. A retrospective cohort of 571 IDH-wildtype glioblastoma patients were included in the study, and pre-operative multi-parametric MRI scans and targeted next-generation sequencing (NGS) data were collected. L21-norm minimization was used to select a subset of 12 radiomic features from the MRI scans, and 13 key driver genes from the five main signal pathways most affected in glioblastoma were selected from the genomic data. Subtypes were identified using a joint learning approach called Anchor-based Partial Multi-modal Clustering on both radiomic and genomic modalities. Kaplan-Meier analysis identified three distinct glioblastoma subtypes: high-risk, medium-risk, and low-risk, based on overall survival outcome (p < 0.05, log-rank test; Hazard Ratio = 1.64, 95% CI 1.17-2.31, Cox proportional hazard model on high-risk and low-risk subtypes). The three subtypes displayed different phenotypical and molecular characteristics in terms of imaging histogram, co-occurrence of genes, and correlation between the two modalities. Our findings demonstrate the synergistic value of integrated radiomic signatures and molecular characteristics for glioblastoma subtyping. Joint learning on both modalities can aid in better understanding the molecular basis of phenotypical signatures of glioblastoma, and provide insights into the biological underpinnings of tumor formation and progression.

Carcinoembryonic antigen-expressing oncolytic measles virus derivative in recurrent glioblastoma: a phase 1 trial

Measles virus (MV) vaccine strains have shown significant preclinical antitumor activity against glioblastoma (GBM), the most lethal glioma histology. In this first in human trial (NCT00390299), a carcinoembryonic antigen-expressing oncolytic measles virus derivative (MV-CEA), was administered in recurrent GBM patients either at the resection cavity (Group A), or, intratumorally on day 1, followed by a second dose administered in the resection cavity after tumor resection on day 5 (Group B). A total of 22 patients received study treatment, 9 in Group A and 13 in Group B. Primary endpoint was safety and toxicity: treatment was well tolerated with no dose-limiting toxicity being observed up to the maximum feasible dose (2×107 TCID50). Median OS, a secondary endpoint, was 11.6 mo and one year survival was 45.5% comparing favorably with contemporary controls. Other secondary endpoints included assessment of viremia, MV replication and shedding, humoral and cellular immune response to the injected virus. A 22 interferon stimulated gene (ISG) diagonal linear discriminate analysis (DLDA) classification algorithm in a post-hoc analysis was found to be inversely (R = -0.6, p = 0.04) correlated with viral replication and tumor microenvironment remodeling including proinflammatory changes and CD8 + T cell infiltration in post treatment samples. This data supports that oncolytic MV derivatives warrant further clinical investigation and that an ISG-based DLDA algorithm can provide the basis for treatment personalization.

Pyrosequencing Analysis of O-6-Methylguanine-DNA Methyltransferase Methylation at Different Cut-Offs of Positivity Associated with Treatment Response and Disease-Specific Survival in Isocitrate Dehydrogenase-Wildtype Grade 4 Glioblastoma

The O-6-methylguanine-DNA methyltransferase (MGMT) gene is a critical guardian of genomic integrity. MGMT methylation in diffuse gliomas serves as an important determinant of patients' prognostic outcomes, more specifically in glioblastomas (GBMs). In GBMs, the absence of MGMT methylation, known as MGMT promoter unmethylation, often translates into a more challenging clinical scenario, tending to present resistance to chemotherapy and a worse prognosis. A pyrosequencing (PSQ) technique was used to analyze MGMT methylation status at different cut-offs (5%, 9%, and 11%) in a sample of 78 patients diagnosed with IDH-wildtype grade 4 GBM. A retrospective analysis was provided to collect clinicopathological and prognostic data. A statistical analysis was used to establish an association between methylation status and treatment response (TR) and disease-specific survival (DSS). The patients with methylated MGMT status experienced progressive disease rates of 84.6%, 80%, and 78.4% at the respective cut-offs of 5%, 9%, and 11%. The number was considerably higher when considering unmethylated patients, as all patients (100%), regardless of the cut-off, presented progressive disease. Regarding disease-specific survival (DSS), the Hazard Ratio (HR) was HR = 0.74 (0.45-1.24; p = 0.251); HR = 0.82 (0.51-1.33; p = 0.425); and HR = 0.79 (0.49-1.29; p = 0.350), respectively. Our study concludes that there is an association between MGMT unmethylation and worse TR and DSS. The 9% cut-off demonstrated a greater potential for patient survival as a function of time, which may shed light on the future need for standardization of MGMT methylation positivity parameters in PSQ.

Evaluation of Temozolomide and Fingolimod Treatments in Glioblastoma Preclinical Models

Glioblastomas are malignant brain tumors which remain lethal due to their aggressive and invasive nature. The standard treatment combines surgical resection, radiotherapy, and chemotherapy using Temozolomide, albeit with a minor impact on patient prognosis (15 months median survival). New therapies evaluated in preclinical translational models are therefore still required to improve patient survival and quality of life. In this preclinical study, we evaluated the effect of Temozolomide in different models of glioblastoma. We also aimed to investigate the efficacy of Fingolimod, an immunomodulatory drug for multiple sclerosis also described as an inhibitor of the sphingosine-1-phosphate (S1P)/S1P receptor axis. The effects of Fingolimod and Temozolomide were analyzed with in vitro 2D and 3D cellular assay and in vivo models using mouse and human glioblastoma cells implanted in immunocompetent or immunodeficient mice, respectively. We demonstrated both in in vitro and in vivo models that Temozolomide has a varied effect depending on the tumor type (i.e., U87MG, U118MG, U138MG, and GL261), demonstrating sensitivity, acquired resistance, and purely resistant tumor phenotypes, as observed in patients. Conversely, Fingolimod only reduced in vitro 2D tumor cell growth and increased cytotoxicity. Indeed, Fingolimod had little or no effect on 3D spheroid cytotoxicity and was devoid of effect on in vivo tumor progression in Temozolomide-sensitive models. These results suggest that the efficacy of Fingolimod is dependent on the glioblastoma tumor microenvironment. Globally, our data suggest that the response to Temozolomide varies depending on the cancer model, consistent with its clinical activity, whereas the potential activity of Fingolimod may merit further evaluation.

Prognosis prediction for glioblastoma multiforme patients using machine learning approaches: Development of the clinically applicable model

BACKGROUND AND PURPOSE

We aimed to develop a clinically applicable prognosis prediction model predicting overall survival (OS) and progression-free survival (PFS) for glioblastoma multiforme (GBM) patients.

MATERIALS AND METHODS

All 467 patients treated with concurrent chemoradiotherapy at Yonsei Cancer Center from 2016 to 2020 were included in this study. We developed a conventional linear regression, Cox proportional hazards (COX), and non-linear machine learning algorithms, random survival forest (RSF) and survival support vector machine (SVM) based on 16 clinical variables. After backward feature selection and hyperparameter tuning using grid search, we repeated 100 times of cross-validations to combat overfitting and enhance the model performance. Harrell's concordance index (C-index) and integrated brier score (IBS) were employed as quantitative performance metrics.

RESULTS

In both predictions, RSF performed much better than COX and SVM. (For OS prediction: RSF C-index = 0.72 90%CI [0.71-0.72] and IBS = 0.12 90%CI [0.10-0.13]; For PFS prediction: RSF C-index = 0.70 90%CI [0.70-0.71] and IBS = 0.12 90%CI [0.10-0.14]). Permutation feature importance confirmed that MGMT promoter methylation, extent of resection, age, cone down planning target volume, and subventricular zone involvement are significant prognostic factors for OS. The importance of the extent of resection and MGMT promoter methylation was much higher than other selected input factors in PFS. Our final models accurately stratified two risk groups with root mean square errors less than 0.07. The sensitivity analysis revealed that our final models are highly applicable to newly diagnosed GBM patients.

CONCLUSION

Our final models can provide a reliable outcome prediction for individual GBM. The final OS and PFS predicting models we developed accurately stratify high-risk groups up to 5-years, and the sensitivity analysis confirmed that both final models are clinically applicable.

Gal-1 Expression Analysis in the GLIOCAT Multicenter Study: Role as a Prognostic Factor and an Immune-Suppressive Biomarker

Glioblastoma (GBM) is the most frequent primary malignant brain tumor and has a dismal prognosis. Unfortunately, despite the recent revolution of immune checkpoint inhibitors in many solid tumors, these have not shown a benefit in overall survival in GBM patients. Therefore, new potential treatment targets as well as diagnostic, prognostic, and/or predictive biomarkers are needed to improve outcomes in this population. The β-galactoside binding protein Galectin-1 (Gal-1) is a protein with a wide range of pro-tumor functions such as proliferation, invasion, angiogenesis, and immune suppression. Here, we evaluated Gal-1 expression by immunohistochemistry in a homogenously treated cohort of GBM (the GLIOCAT project) and correlated its expression with clinical and molecular data. We observed that Gal-1 is a negative prognostic factor in GBM. Interestingly, we observed higher levels of Gal-1 expression in the mesenchymal/classical subtypes compared to the less aggressive proneural subtype. We also observed a Gal-1 expression correlation with immune suppressive signatures of CD4 T-cells and macrophages, as well as with several GBM established biomarkers, including SHC1, PD-L1, PAX2, MEOX2, YKL-40, TCIRG1, YWHAG, OLIG2, SOX2, Ki-67, and SOX11. Moreover, Gal-1 levels were significantly lower in grade 4 IDH-1 mutant astrocytomas, which have a better prognosis. Our results confirm the role of Gal-1 as a prognostic factor and also suggest its value as an immune-suppressive biomarker in GBM.

New natural compound inhibitors of PDGFRA (platelet-derived growth factor receptor α) based on computational study for high-grade glioma therapy

Background

High-grade glioma (HGG) is a malignant brain tumor that is common and aggressive in children and adults. In the current medical paradigm, surgery and radiotherapy are the standard treatments for HGG patients. Despite this, the overall prognosis is still very bleak. Studies have shown that platelet-derived growth factor receptor α (PDGFRA) is an essential target to treat tumors and inhibiting the activity of PDGFRA can improve the prognosis of HGG. Thus, PDGFRA inhibitors are critical to developing drugs and cancer treatment.

Objective

The purpose of this study was to screen lead compounds and candidate drugs with potential inhibitors against platelet-derived growth factor receptor α (PDGFRA) from the drug library (ZINC database) in order to improve the prognosis of patients with high-grade glioma (HGG).

Materials and methods

In our study, we selected Imatinib as the reference drug. A series of computer-aided technologies, such as Discovery Studio 2019 and Schrodinger, were used to screen and assess potential inhibitors of PDGFRA. The first step was to calculate the LibDock scores and then analyze the pharmacological and toxicological properties. Following this, we docked the small molecules selected in the previous steps with PDGFRA to study their docking mechanism and affinity. In addition, molecular dynamics simulation was used to determine whether the ligand-PDGFRA complex was stable in nature.

Results

Two novel natural compounds 1 and 2 (ZINC000008829785 and ZINC000013377891) from the ZINC database were found binding to PDGFRA with more favorable interaction energy. Also, they were predicted with less Ames mutagenicity, rodent carcinogenicity, non-developmental toxic potential, and tolerant with cytochrome P450 2D6 (CYP2D6). The dynamic simulation analysis demonstrated that ZINC000008829785-PDGFRA and ZINC000013377891-PDGFRA dimer complex had more favorable potential energy compared with Imatinib, and they can exist in natural environments stably.

Conclusion

ZINC000008829785 and ZINC000013377891 might provide a solid foundation for drugs that inhibit PDGFRA in HGG. In addition to being safe drug candidates, these compounds had important implications for improving drugs targeting PDGFRA.

Survival analysis in high-grade glioma: The role of salvage surgery

OBJECTIVES

This study addresses the survival of consecutive patients with high-grade gliomas (HGG) treated at the same institution over a period of 10 years. We analyse the importance of associated factors and the role of salvage surgery at the time of progression.

METHODS

We retrospectively analysed a series of patients with World Health Organization (WHO) grade III/IV gliomas treated between 2008 and 2017 at Hospital Gregorio Marañón (Madrid, Spain). Clinical, radiological, and anatomical pathology data were obtained from patient clinical histories.

RESULTS

Follow-up was completed in 233 patients with HGG. Mean age was 62.2 years. The median survival time was 15.4 months. Of 133 patients (59.6%) who had undergone surgery at the time of diagnosis, 43 (32.3%) underwent salvage surgery at the time of progression. This subgroup presented longer overall survival and survival after progression. Higher Karnofsky Performance Status score at diagnosis, a greater extent of surgical resection, and initial diagnosis of WHO grade III glioma were also associated with longer survival.

CONCLUSIONS

About one-third of patients with HGG may be eligible for salvage surgery at the time of progression. Salvage surgery in this subgroup of patients was significantly associated with longer survival.

Impact of fractionated stereotactic radiotherapy on activity of daily living and performance status in progressive/recurrent glioblastoma: a retrospective study

BACKGROUND

The prognosis of recurrent glioblastoma (GBM) is poor, with limited options of palliative localized or systemic treatments. Survival can be improved by a second localized treatment; however, it is not currently possible to identify which patients would benefit from this approach. This study aims to evaluate which factors lead to a lower Karnofsky performance status (KPS) score after fractionated stereotactic RT (fSRT).

METHODS

We retrospectively collected data from patients treated with fSRT for recurrent GBM at the Institut de Cancérologie de Lorraine between October 2010 and November 2017 and analyzed which factors were associated with a lower KPS score.

RESULTS

59 patients received a dose of 25 Gy in 5 sessions spread over 5-7 days (80% isodose). The median time from the end of primary radiotherapy to the initiation of fSRT was 10.7 months. The median follow-up after fSRT initiation was 8.8 months. The incidence of KPS and ADL impairment in all patients were 51.9% and 37.8% respectively with an adverse impact of PTV size on KPS (HR = 1.57 [95% CI 1.19-2.08], p = 0.028). Only two patients showed early grade 3 toxicity and none showed grade 4 or late toxicity. The median overall survival time, median overall survival time after fSRT, median progression-free survival and institutionalization-free survival times were 25.8, 8.8, 3.9 and 7.7 months, respectively. Initial surgery was associated with better progression-free survival (Hazard ratio (HR) = 0.48 [95% CI 0.27-0.86], p = 0.013).

CONCLUSIONS

A larger PTV should predicts lower KPS in the treatment of recurrent GBM using fSRT.

Management of Recurrent Glioblastomas: What Can We Learn from the French Glioblastoma Biobank?

Safe maximal resection followed by radiotherapy plus concomitant and adjuvant temozolomide (TMZ) is universally accepted as the first-line treatment for glioblastoma (GB), but no standard of care has yet been defined for managing recurrent GB (rGB). We used the French GB biobank (FGB) to evaluate the second-line options currently used, with a view to defining the optimal approach and future directions in GB research. We retrospectively analyzed data for 338 patients with de novo isocitrate dehydrogenase (IDH)-wildtype GB recurring after TMZ chemoradiotherapy. Cox proportional hazards models and Kaplan-Meier analyses were used to investigate survival outcomes. Median overall survival after first surgery (OS1) was 19.8 months (95% CI: 18.5-22.0) and median OS after first progression (OS2) was 9.9 months (95% CI: 8.8-10.8). Two second-line options were noted for rGB patients in the FGB: supportive care and treatments, with systemic treatment being the treatment most frequently used. The supportive care option was independently associated with a shorter OS2 (p < 0.001). None of the systemic treatment regimens was unequivocally better than the others for rGB patients. An analysis of survival outcomes based on time to first recurrence (TFR) after chemoradiotherapy indicated that survival was best for patients with a long TFR (≥18 months; median OS1: 44.3 months (95% CI: 41.7-56.4) and median OS2: 13.0 months (95% CI: 11.2-17.7), but that such patients constituted only a small proportion of the total patient population (13.0%). This better survival appeared to be more strongly associated with response to first-line treatment than with response to second-line treatment, indicating that the recurring tumors were more aggressive and/or resistant than the initial tumors in these patients. In the face of high rates of treatment failure for GB, the establishment of well-designed large cohorts of primary and rGB samples, with the help of biobanks, such as the FGB, taking into account the TFR and survival outcomes of GB patients, is urgently required for solid comparative biological analyses to drive the discovery of novel prognostic and/or therapeutic clinical markers for GB.

Apatinib for recurrent/progressive glioblastoma multiforme: A salvage option

Purpose: The recurrent/progressive glioblastoma multiforme (GBM) carries a dismal prognosis and the definitive treatment strategy has not yet been established. This study aimed to assess the efficacy and safety of apatinib in recurrent/progressive GBM patients.

Materials and methods: The clinical data of 19 recurrent/progressive GBM patients who received apatinib treatment from November 2015 to December 2019 at Sun Yat-sen University Cancer Center were collected retrospectively in this study. Objective response rate (ORR), disease control rate (DCR), progression-free survival (PFS), overall survival (OS), and treatment-related adverse events (AEs) were reviewed and assessed.

Results: The overall ORR was 52.6%, and the DCR was 73.7%. Median PFS and OS were 5.1 and 10.4 months, respectively. The 6-month PFS and OS rates were 38.9% and 68.4%, respectively. The 12-month PFS and OS rates were 16.7% and 36.8%, respectively. The treatment-related toxicities were generally well-tolerated. The most common grade 3/4 AEs were hand-foot syndrome (36.8%) and hypertension (21.1%).

Conclusion: Our study showed that apatinib therapy provided a better salvaging option for recurrent/progressive GBM patients and the toxicity was manageable.

Construction of an immune-related gene signature for the prognosis and diagnosis of glioblastoma multiforme

Background

Increasing evidence has suggested that inflammation is related to tumorigenesis and tumor progression. However, the roles of immune-related genes in the occurrence, development, and prognosis of glioblastoma multiforme (GBM) remain to be studied.

Methods

The GBM-related RNA sequencing (RNA-seq), survival, and clinical data were acquired from The Cancer Genome Atlas (TCGA), Genotype-Tissue Expression (GTEx), Chinese Glioma Genome Atlas (CGGA), and Gene Expression Omnibus (GEO) databases. Immune-related genes were obtained from the Molecular Signatures Database (MSigDB). Differently expressed immune-related genes (DE-IRGs) between GBM and normal samples were identified. Prognostic genes associated with GBM were selected by Kaplan-Meier survival analysis, Least Absolute Shrinkage and Selection Operator (LASSO)-penalized Cox regression analysis, and multivariate Cox analysis. An immune-related gene signature was developed and validated in TCGA and CGGA databases separately. The Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) analyses were performed to explore biological functions of the signature. The correlation between immune cell infiltration and the signature was analyzed by single-sample gene set enrichment analysis (ssGSEA), and the diagnostic value was investigated. The gene set enrichment analysis (GSEA) was performed to explore the potential function of the signature genes in GBM, and the protein-protein interaction (PPI) network was constructed.

Results

Three DE-IRGs [Pentraxin 3 (PTX3), TNFSF9, and bone morphogenetic protein 2 (BMP2)] were used to construct an immune-related gene signature. Receiver operating characteristic (ROC) curves and Cox analyses confirmed that the 3-gene-based prognostic signature was a good independent prognostic factor for GBM patients. We found that the signature was mainly involved in immune-related biological processes and pathways, and multiple immune cells were disordered between the high- and low-risk groups. GSEA suggested that PTX3 and TNFSF9 were mainly correlated with interleukin (IL)-17 signaling pathway, nuclear factor kappa B (NF-κB) signaling pathway, tumor necrosis factor (TNF) signaling pathway, and Toll-like receptor signaling pathway, and the PPI network indicated that they could interact directly or indirectly with inflammatory pathway proteins. Quantitative real-time PCR (qRT-PCR) indicated that the three genes were significantly different between target tissues.

Conclusion

The signature with three immune-related genes might be an independent prognostic factor for GBM patients and could be associated with the immune cell infiltration of GBM patients.

Integrin-αvβ3 as a Therapeutic Target in Glioblastoma: Back to the Future?

Glioblastoma (GBM), the most common primary malignant brain tumor, is associated with a dismal prognosis. Standard therapies including maximal surgical resection, radiotherapy, and temozolomide chemotherapy remain poorly efficient. Improving GBM treatment modalities is, therefore, a paramount challenge for researchers and clinicians. GBMs exhibit the hallmark feature of aggressive invasion into the surrounding tissue. Among cell surface receptors involved in this process, members of the integrin family are known to be key actors of GBM invasion. Upregulation of integrins was reported in both tumor and stromal cells, making them a suitable target for innovative therapies targeting integrins in GBM patients, as their impairment disrupts tumor cell proliferation and invasive capacities. Among them, integrin-αvβ3 expression correlates with high-grade GBM. Driven by a plethora of preclinical biological studies, antagonists of αvβ3 rapidly became attractive therapeutic candidates to impair GBM tumorigenesis. In this perspective, the advent of nuclear medicine is currently one of the greatest components of the theranostic concept in both preclinical and clinical research fields. In this review, we provided an overview of αvβ3 expression in GBM to emphasize the therapeutic agents developed. Advanced current and future developments in the theranostic field targeting αvβ3 are finally discussed.

Tracking glioblastoma progression after initial resection with minimal reaction-diffusion models

We describe a preliminary effort to model the growth and progression of glioblastoma multiforme, an aggressive form of primary brain cancer, in patients undergoing treatment for recurrence of tumor following initial surgery and chemoradiation. Two reaction-diffusion models are used: the Fisher-Kolmogorov equation and a 2-population model, developed by the authors, that divides the tumor into actively proliferating and quiescent (or necrotic) cells. The models are simulated on 3-dimensional brain geometries derived from magnetic resonance imaging (MRI) scans provided by the Barrow Neurological Institute. The study consists of 17 clinical time intervals across 10 patients that have been followed in detail, each of whom shows significant progression of tumor over a period of 1 to 3 months on sequential follow up scans. A Taguchi sampling design is implemented to estimate the variability of the predicted tumors to using 144 different choices of model parameters. In 9 cases, model parameters can be identified such that the simulated tumor, using both models, contains at least 40 percent of the volume of the observed tumor. We discuss some potential improvements that can be made to the parameterizations of the models and their initialization.

The role of artificial intelligence in paediatric neuroradiology

Imaging plays a fundamental role in the managing childhood neurologic, neurosurgical and neuro-oncological disease. Employing multi-parametric MRI techniques, such as spectroscopy and diffusion- and perfusion-weighted imaging, to the radiophenotyping of neuroradiologic conditions is becoming increasingly prevalent, particularly with radiogenomic analyses correlating imaging characteristics with molecular biomarkers of disease. However, integration into routine clinical practice remains elusive. With modern multi-parametric MRI now providing additional data beyond anatomy, informing on histology, biology and physiology, such metric-rich information can present as information overload to the treating radiologist and, as such, information relevant to an individual case can become lost. Artificial intelligence techniques are capable of modelling the vast radiologic, biological and clinical datasets that accompany childhood neurologic disease, such that this information can become incorporated in upfront prognostic modelling systems, with artificial intelligence techniques providing a plausible approach to this solution. This review examines machine learning approaches than can be used to underpin such artificial intelligence applications, with exemplars for each machine learning approach from the world literature. Then, within the specific use case of paediatric neuro-oncology, we examine the potential future contribution for such artificial intelligence machine learning techniques to offer solutions for patient care in the form of decision support systems, potentially enabling personalised medicine within this domain of paediatric radiologic practice.

Initial cognitive impairment predicts shorter survival of patients with glioblastoma

OBJECTIVES

Seizures as presenting symptom of glioblastoma (GBM) are known to predict prolonged survival, whereas the clinical impact of other initial symptoms is less known. Our main objective was to evaluate the influence of different presenting symptoms on survival in a clinical setting. We also assessed lead times, tumour size and localization.

METHODS

Medical records of 189 GBM patients were reviewed regarding the first medical appointment, presenting symptom/s, date of diagnostic radiology and survival. Tumour size, localization and treatment data were retrieved. Overall survival was calculated using Kaplan-Meier and Mann-Whitney U test. Cox regression was used for risk estimation.

RESULTS

Cognitive impairment as the initial symptom was often misinterpreted in primary health care leading to a delayed diagnosis. Initial global symptoms (66% of all patients) were associated with reduced survival compared to no global symptoms (median 8.4 months vs. 12.6 months). Those with the most common cognitive dysfunctions: change of behaviour, memory impairment and/or disorientation had a reduced median survival to 6.4 months. In contrast, seizures (32%) were associated with longer survival (median 11.2 months vs. 8.3 months). Global symptoms were associated with larger tumours than seizures, but tumour size had no linear association with survival. The setting of the first medical appointment was evenly distributed between primary health care and emergency units.

CONCLUSION

Patients with GBM presenting with cognitive symptoms are challenging to identify, have larger tumours and reduced survival. In contrast, epileptic seizures as the first symptom are associated with longer survival and smaller tumours.

A Discussion of Machine Learning Approaches for Clinical Prediction Modeling

While machine learning has occupied a niche in clinical medicine for decades, continued method development and increased accessibility of medical data have led to broad diversification of approaches. These range from humble regression-based models to more complex artificial neural networks; yet, despite heterogeneity in foundational principles and architecture, the spectrum of machine learning approaches to clinical prediction modeling have invariably led to the development of algorithms advancing our ability to provide optimal care for our patients. In this chapter, we briefly review early machine learning approaches in medicine before delving into common approaches being applied for clinical prediction modeling today. For each, we offer a brief introduction into theory and application with accompanying examples from the medical literature. In doing so, we present a summarized image of the current state of machine learning and some of its many forms in medical predictive modeling.

Tyrosine Kinase Inhibitors in Adult Glioblastoma: An (Un)Closed Chapter?

Glioblastoma (GBM) is the most common and lethal form of malignant brain tumor. GBM patients normally undergo surgery plus adjuvant radiotherapy followed by chemotherapy. Numerous studies into the molecular events driving GBM highlight the central role played by the Epidermal Growth Factor Receptor (EGFR), as well as the Platelet-derived Growth Factor Receptors PDGFRA and PDGFRB in tumor initiation and progression. Despite strong preclinical evidence for the therapeutic potential of tyrosine kinase inhibitors (TKIs) that target EGFR, PDGFRs, and other tyrosine kinases, clinical trials performed during the last 20 years have not led to the desired therapeutic breakthrough for GBM patients. While clinical trials are still ongoing, in the medical community there is the perception of TKIs as a lost opportunity in the fight against GBM. In this article, we review the scientific rationale for the use of TKIs targeting glioma drivers. We critically analyze the potential causes for the failure of TKIs in the treatment of GBM, and we propose alternative approaches to the clinical evaluation of TKIs in GBM patients.

Measurements of Functional Network Connectivity Using Resting State Arterial Spin Labeling During Neurosurgery

In neurosurgery, an exact delineation of functional areas is of great interest to spare important regions to ensure the best possible outcome for the patient (i.e., maximum removal while maintaining the highest possible quality of life). Preoperative imaging is routinely performed, including the visualization of not only structural but also functional information. During surgery, however, brain shift can occur, leading to an offset between the previously defined and the real position. Real-time imaging during the procedure is therefore desired to obtain this information while performing surgery. In this study 15 patients suffering from glioblastoma multiforme were included. These patients underwent structural and perfusion imaging using arterial spin labeling during the procedure. The latter has been used for gathering information about tumor residual perfusion. However, special postprocessing of this data allows for additional mapping of resting state networks and is intended to be used to gather deeper insights to aid the surgeon in planning the procedure. The data of each patient could be successfully postprocessed and used to map different resting state networks alongside the default mode network. On the basis of this study, it is feasible to use the information obtained from perfusion imaging to visualize not only vascular signal but also functional activation of resting state networks without acquiring any additional data besides the already available information. This may help guide the neurosurgeon in real time to adjust the surgical plan.

Discriminating surgical bed cysts from bacterial brain abscesses after Carmustine wafer implantation in newly diagnosed IDH-wildtype glioblastomas

Carmustine wafers can be implanted in the surgical bed of high-grade gliomas, which can induce surgical bed cyst formation, leading to clinically relevant mass effect. An observational retrospective monocentric study was conducted including 122 consecutive adult patients with a newly diagnosed supratentorial glioblastoma who underwent a surgical resection with Carmustine wafer implantation as first line treatment (2005-2018). Twenty-two patients (18.0%) developed a postoperative contrast-enhancing cyst within the surgical bed: 16 surgical bed cysts and six bacterial abscesses. All patients with a surgical bed cyst were managed conservatively, all resolved on imaging follow-up, and no patient stopped the radiochemotherapy. Independent risk factors of formation of a postoperative surgical bed cyst were age ≥ 60 years (p = 0.019), number of Carmustine wafers implanted ≥ 8 (p = 0.040), and partial resection (p = 0.025). Compared to surgical bed cysts, the occurrence of a postoperative bacterial abscess requiring surgical management was associated more frequently with a shorter time to diagnosis from surgery (p = 0.009), new neurological deficit (p < 0.001), fever (p < 0.001), residual air in the cyst (p = 0.018), a cyst diameter greater than that of the initial tumor (p = 0.027), and increased mass effect and brain edema compared to early postoperative MRI (p = 0.024). Contrast enhancement (p = 0.473) and diffusion signal abnormalities (p = 0.471) did not differ between postoperative bacterial abscesses and surgical bed cysts. Clinical and imaging findings help discriminate between surgical bed cysts and bacterial abscesses following Carmustine wafer implantation. Surgical bed cysts can be managed conservatively. Individual risk factors will help tailor their steroid therapy and imaging follow-up.

Repeat Resection in Recurrent Glioblastoma (3rGBM) Trial: a randomized care trial

BACKGROUND

The prognosis for patients with recurrent glioblastoma (GBM) is dismal, and the question of repeat surgery at time of recurrence is common. Re-operation in the management of these patients remains controversial, as there is no randomized evidence of benefit. An all-inclusive pragmatic care trial is needed to evaluate the role of repeat resection.

METHODS

3rGBM is a multicenter, pragmatic, prospective, parallel-group randomized care trial, with 1:1 allocation to repeat resection or standard care with no repeat resection. To test the hypothesis that repeat resection can improve overall survival by at least 3 months (from 6 to 9 months), 250 adult patients with prior resection of pathology-proven glioblastoma for whom the attending surgeon believes repeat resection may improve quality survival will be enrolled. A surrogate measure of quality of life, the number of days outside of hospital/nursing/palliative care facility, will also be compared. Centers are invited to participate without financial compensation and without contracts. Clinicians may apply to local authorities to approve an investigator-led in-house trial, using a common protocol, web-based randomization platform, and simple standardized case report forms.

DISCUSSION

The 3rGBM trial is a modern transparent care research framework with no additional risks, tests, or visits other than what patients would encounter in normal care. The burden of proof remains on repeat surgical management of recurrent GBM, because this management has yet to be shown beneficial. The trial is designed to help patients and surgeons manage the uncertainty regarding optimal care.

CLINICAL TRIAL REGISTRATION

http://www.clinicaltrials.gov. Unique identifier: NCT04838782.

Autologous adoptive immune-cell therapy elicited a durable response with enhanced immune reaction signatures in patients with recurrent glioblastoma: An open label, phase I/IIa trial

Glioblastoma multiforme (GBM) is an aggressive malignancy classified by the World Health Organization as a grade IV glioma. Despite the availability of aggressive standard therapies, most patients experience recurrence, for which there are currently no effective treatments. We aimed to conduct a phase I/IIa clinical trial to investigate the safety and efficacy of adoptive, ex-vivo-expanded, and activated natural killer cells and T lymphocytes from peripheral blood mononuclear cells of patients with recurrent GBM. This study was a single-arm, open-label, investigator-initiated trial on 14 patients recruited between 2013 and 2017. The immune cells were administered via intravenous injection 24 times at 2-week intervals after surgical resection or biopsy. The safety and clinical efficacy of this therapy was examined by assessing adverse events and comparing 2-year overall survival (OS). Transcriptomic analysis of tumor tissues was performed using NanoString to identify the mechanism of therapeutic efficacy. No grade 4 or 5 severe adverse events were observed. The most common treatment-related adverse events were grade 1 or 2 in severity. The most severe adverse event was grade 3 fever. Median OS was 22.5 months, and the median progression-free survival was 10 months. Five patients were alive for over 2 years and showed durable response with enhanced immune reaction transcriptomic signatures without clinical decline until the last follow-up after completion of the therapy. In conclusion, autologous adoptive immune-cell therapy was safe and showed durable response in patients with enhanced immune reaction signatures. This therapy may be effective for recurrent GBM patients with high immune response in their tumor microenvironments. Trial registration: The Korea Clinical Research Information Service database: KCT0003815, Registered 18 April 2019, retrospectively registered.

Developing a Model for Prediction of Serum 25-Hydroxyvitamin D Level: The Use of Linear Regression and Machine Learning Methods

OBJECTIVE

Because vitamin D status affects many organs and tissues of the body, it is important to determine the factors affecting it. The purpose of this study was to develop a model for predicting the serum 25-hydroxyvitamin D [25(OH)D] level in healthy young adults.

METHOD

This cross-sectional study was conducted on 201 healthy individuals aged 20 to 40 years old in Shiraz, Iran. Data regarding demographic characteristics, vitamin D intake through supplements, and sun exposure habits were gathered. Serum 25(OH)D concentration was also measured. Data were analyzed with R software using linear regression and different machine learning methods such as conditional tree, conditional forest and random forest.

RESULTS

Based on the linear regression, male sex (p < 0.001), taking 50,000 IU vitamin D₃ supplement monthly (p < 0.001), and lower waist circumference (p = 0.018) were identified as effective factors in increasing serum 25(OH)D levels. According to the conditional tree, taking 50,000 IU vitamin D₃ supplement monthly (p < 0.001) and sex (p < 0.001) were two main factors in the classification of individuals in terms of serum 25(OH)D levels. Besides, conditional forest and random forest results showed that the most important variable was taking 50,000 IU vitamin D₃ supplement monthly.

CONCLUSIONS

Supplement use is the first and most important predictor of 25(OH)D levels and other factors, including sex and waist circumference, are ranked thereafter, and the importance of these factors is greater in those who do not take vitamin D₃ supplements.

Dynamic Susceptibility Perfusion Imaging for Differentiating Progressive Disease from Pseudoprogression in Diffuse Glioma Molecular Subtypes

Rationale and Objectives: Advanced adjuvant therapy of diffuse gliomas can result in equivocal findings in follow-up imaging. We aimed to assess the additional value of dynamic susceptibility perfusion imaging in the differentiation of progressive disease (PD) from pseudoprogression (PsP) in different molecular glioma subtypes. Materials and Methods: 89 patients with treated diffuse glioma with different molecular subtypes (IDH wild type (Astro-IDH^wt), IDH mutant astrocytomas (Astro-IDH^mut) and oligodendrogliomas), and tumor-suspect lesions on post-treatment follow-up imaging were classified into two outcome groups (PD or PsP) retrospectively by histopathology or clinical follow-up. The relative cerebral blood volume (rCBV) was assessed in the tumor-suspect FLAIR and contrast-enhancing (CE) lesions. We analyzed how a multilevel classification using a molecular subtype, the presence of a CE lesion, and two rCBV histogram parameters performed for PD prediction compared with a decision tree model (DTM) using additional rCBV parameters. Results: The PD rate was 69% in the whole cohort, 86% in Astro-IDH^wt, 52% in Astro-IDH^mut, and 55% in oligodendrogliomas. In the presence of a CE lesion, the PD rate was higher with 82%, 94%, 59%, and 88%, respectively; if there was no CE lesion, however, the PD rate was only 44%, 60%, 40%, and 33%, respectively. The additional use of the rCBV parameters in the DTM yielded a prediction accuracy for PD of 99%, 100%, 93%, and 95%, respectively. Conclusion: Utilizing combined information about the molecular tumor type, the presence or absence of CE lesions and rCBV parameters increases PD prediction accuracy in diffuse glioma.

Management, functional outcomes and survival in a French multicentric series of 118 adult patients with cerebellar glioblastoma

PURPOSE

To analyze the outcomes and predictors in a large series of cerebellar glioblastomas in order to guide patient management.

METHODS

The French brain tumor database and the Club de Neuro-Oncologie of the Société Française de Neurochirurgie retrospectively identified adult patients with cerebellar glioblastoma diagnosed between 2003 and 2017. Diagnosis was confirmed by a centralized neuropathological review.

RESULTS

Data from 118 cerebellar glioblastoma patients were analyzed (mean age 55.9 years, 55.1% males). The clinical presentation associated raised intracranial pressure (50.8%), static cerebellar syndrome (68.6%), kinetic cerebellar syndrome (49.2%) and/or cranial nerve disorders (17.8%). Glioblastomas were hemispheric (55.9%), vermian (14.4%) or both (29.7%). Hydrocephalus was present in 49 patients (41.5%). Histologically, tumors corresponded either to IDH-wild-type or to K27-mutant glioblastomas. Surgery consisted of total (12.7%), subtotal (35.6%), partial resection (33.9%) or biopsy (17.8%). The postoperative Karnofsky performance status was improved, stable and worsened in 22.4%, 43.9% and 33.7% of patients, respectively. Progression-free and overall survivals reached 5.1 months and 9.1 months, respectively. Compared to other surgical strategies, total or subtotal resection improved the Karnofsky performance status (33.3% vs 12.5%, p < 0.001), prolonged progression-free and overall survivals (6.5 vs 4.3 months, p = 0.015 and 16.7 vs 6.2 months, p < 0.001, respectively) and had a comparable complication rate (40.4% vs 31.1%, p = 0.29). After total or subtotal resection, the functional outcomes were correlated with age (p = 0.004) and cerebellar hemispheric tumor location (p < 0.001) but not brainstem infiltration (p = 0.16).

CONCLUSION

In selected patients, maximal resection of cerebellar glioblastoma is associated with improved onco-functional outcomes, compared with less invasive procedures.

Recurrent Glioblastoma: From Molecular Landscape to New Treatment Perspectives

Glioblastoma is the most frequent and aggressive form among malignant central nervous system primary tumors in adults. Standard treatment for newly diagnosed glioblastoma consists in maximal safe resection, if feasible, followed by radiochemotherapy and adjuvant chemotherapy with temozolomide; despite this multimodal treatment, virtually all glioblastomas relapse. Once tumors progress after first-line therapy, treatment options are limited and management of recurrent glioblastoma remains challenging. Loco-regional therapy with re-surgery or re-irradiation may be evaluated in selected cases, while traditional systemic therapy with nitrosoureas and temozolomide rechallenge showed limited efficacy. In recent years, new clinical trials using, for example, regorafenib or a combination of tyrosine kinase inhibitors and immunotherapy were performed with promising results. In particular, molecular targeted therapy could show efficacy in selected patients with specific gene mutations. Nonetheless, some molecular characteristics and genetic alterations could change during tumor progression, thus affecting the efficacy of precision medicine. We therefore reviewed the molecular and genomic landscape of recurrent glioblastoma, the strategy for clinical management and the major phase I-III clinical trials analyzing recent drugs and combination regimens in these patients.

Survival prediction of glioblastoma patients-are we there yet? A systematic review of prognostic modeling for glioblastoma and its clinical potential

Glioblastoma is associated with a poor prognosis. Even though survival statistics are well-described at the population level, it remains challenging to predict the prognosis of an individual patient despite the increasing number of prognostic models. The aim of this study is to systematically review the literature on prognostic modeling in glioblastoma patients. A systematic literature search was performed to identify all relevant studies that developed a prognostic model for predicting overall survival in glioblastoma patients following the PRISMA guidelines. Participants, type of input, algorithm type, validation, and testing procedures were reviewed per prognostic model. Among 595 citations, 27 studies were included for qualitative review. The included studies developed and evaluated a total of 59 models, of which only seven were externally validated in a different patient cohort. The predictive performance among these studies varied widely according to the AUC (0.58-0.98), accuracy (0.69-0.98), and C-index (0.66-0.70). Three studies deployed their model as an online prediction tool, all of which were based on a statistical algorithm. The increasing performance of survival prediction models will aid personalized clinical decision-making in glioblastoma patients. The scientific realm is gravitating towards the use of machine learning models developed on high-dimensional data, often with promising results. However, none of these models has been implemented into clinical care. To facilitate the clinical implementation of high-performing survival prediction models, future efforts should focus on harmonizing data acquisition methods, improving model interpretability, and externally validating these models in multicentered, prospective fashion.

A Comparison Study of Machine Learning (Random Survival Forest) and Classic Statistic (Cox Proportional Hazards) for Predicting Progression in High-Grade Glioma after Proton and Carbon Ion Radiotherapy

Background

Machine learning (ML) algorithms are increasingly explored in glioma prognostication. Random survival forest (RSF) is a common ML approach in analyzing time-to-event survival data. However, it is controversial which method between RSF and traditional cornerstone method Cox proportional hazards (CPH) is better fitted. The purpose of this study was to compare RSF and CPH in predicting tumor progression of high-grade glioma (HGG) after particle beam radiotherapy (PBRT).

Methods

The study enrolled 82 consecutive HGG patients who were treated with PBRT at Shanghai Proton and Heavy Ion Center between 6/2015 and 11/2019. The entire cohort was split into the training and testing set in an 80/20 ratio. Ten variables from patient-related, tumor-related and treatment-related information were utilized for developing CPH and RSF for predicting progression-free survival (PFS). The model performance was compared in concordance index (C-index) for discrimination (accuracy), brier score (BS) for calibration (precision) and variable importance for interpretability.

Results

The CPH model demonstrated a better performance in terms of integrated C-index (62.9%) and BS (0.159) compared to RSF model (C-index = 61.1%, BS = 0.174). In the context of variable importance, CPH model indicated that age (P = 0.024), WHO grade (P = 0.020), IDH gene (P = 0.019), and MGMT promoter status (P = 0.040) were significantly correlated with PFS in the univariate analysis; multivariate analysis showed that age (P = 0.041), surgical completeness (P = 0.084), IDH gene (P = 0.057), and MGMT promoter (P = 0.092) had a significant or trend toward the relation with PFS. RSF showed that merely IDH and age were of positive importance for predicting PFS. A final nomogram was developed to predict tumor progression at the individual level based on CPH model.

Conclusions

In a relatively small dataset with HGG patients treated with PBRT, CPH outperformed RSF for predicting tumor progression. A comprehensive criterion with accuracy, precision, and interpretability is recommended in evaluating ML prognostication approaches for clinical deployment.

Pretreatment inflammatory indices predict Bevacizumab response in recurrent Glioma

Aim: It remains unclear what the best therapeutic option for recurrent glioma patients after Stupp treatment is. Bevacizumab (BVZ) is commonly administered in progression, but it appears that only some patients benefit. It would be useful to find biomarkers that determine beforehand who these patients are.

Methods: The protocol included 31 high-risk progressing glioma patients after Stupp treatment who received BVZ 5-10 mg/kg every 14 days and temozolomide (3-19 cycles, 150-200 mg five days each 28-day cycle) during a mean of eight cycles of BVZ or until tumor progression or unacceptable toxicity. We analyzed the clinical outcome values of inflammatory indices measured before BVZ administration.

Results: Lymphocyte level before BVZ administration was the best independent predictor of overall survival (HR = 0.34; 95%CI: 0.145-0.81; P = 0.015). The area under the receiver operating characteristic (ROC) curve was 0.823, with 1.645 being the optimal cut-off value, and 0.80 and 0.85 the sensitivity and specificity values, respectively. Responder and non-responder survival curves were also significantly different, considering the first and second tertiles as cut-off points. The number of BVZ cycles was not related to lymphopenia. Pretreatment neutrophil, platelet levels, platelet-to-lymphocyte ratio (PLR), and neutrophil-to-lymphocyte ratio (NLR) did not have independent predictive value. Inflammatory variables were not correlated with each other. However, patients with high NLR and PLR simultaneously (double positive PLR-NLR) showed a worse clinical outcome than the rest (P = 0.043).

Conclusion: Pretreatment lymphocyte levels and double positive PLR-NLR could be used as non-invasive hematological prognostic markers for recurrent gliomas treated with bevacizumab. A close relationship emerged between inflammation and angiogenesis.

Survival analysis in high-grade glioma: the role of salvage surgery

OBJECTIVES

This study addresses the survival of consecutive patients with high-grade gliomas treated at the same institution over a period of 10 years. We analyse the importance of associated factors and the role of salvage surgery at the time of progression.

METHODS

We retrospectively analysed a series of patients with World Health Organization (WHO) grade III/IV gliomas treated between 2008 and 2017 at Hospital Gregorio Marañón (Madrid, Spain). Clinical, radiological, and anatomical pathology data were obtained from patient clinical histories.

RESULTS

Follow-up was completed in 233 patients with HGG. Mean age was 62.2 years. The median survival time was 15.4 months. Of 133 patients (59.6%) who had undergone surgery at the time of diagnosis, 43 (32.3%) underwent salvage surgery at the time of progression. This subgroup presented longer overall survival and survival after progression. Higher Karnofsky Performance Status score at diagnosis, a greater extent of surgical resection, and initial diagnosis of WHO grade III glioma were also associated with longer survival.

CONCLUSIONS

About one-third of patients with HGG may be eligible for salvage surgery at the time of progression. Salvage surgery in this subgroup of patients was significantly associated with longer survival.

Sex-specific impact of patterns of imageable tumor growth on survival of primary glioblastoma patients

Background

Sex is recognized as a significant determinant of outcome among glioblastoma patients, but the relative prognostic importance of glioblastoma features has not been thoroughly explored for sex differences.

Methods

Combining multi-modal MR images, biomathematical models, and patient clinical information, this investigation assesses which pretreatment variables have a sex-specific impact on the survival of glioblastoma patients (299 males and 195 females).

Results

Among males, tumor (T1Gd) radius was a predictor of overall survival (HR = 1.027, p = 0.044). Among females, higher tumor cell net invasion rate was a significant detriment to overall survival (HR = 1.011, p < 0.001). Female extreme survivors had significantly smaller tumors (T1Gd) (p = 0.010 t-test), but tumor size was not correlated with female overall survival (p = 0.955 CPH). Both male and female extreme survivors had significantly lower tumor cell net proliferation rates than other patients (M p = 0.004, F p = 0.001, t-test).

Conclusion

Despite similar distributions of the MR imaging parameters between males and females, there was a sex-specific difference in how these parameters related to outcomes.

A Comparison of Functional Outcomes between Patients Admitted to Inpatient Rehabilitation after Initial Diagnosis Versus Recurrence of Glioblastoma Multiforme

BACKGROUND

Prior studies of inpatient rehabilitation of patients with brain tumor demonstrate similar functional gains as compared to other rehabilitation populations. There are few studies specifically examining the rehabilitation of patients with glioblastoma.

OBJECTIVE

To compare functional outcomes between matched patients admitted to acute inpatient rehabilitation after initial diagnosis of glioblastoma (iGBM) and after diagnosis of recurrent glioblastoma (rGBM).

DESIGN

A retrospective, case-matched study using descriptive statistics compared demographic information and functional outcomes as designated by the Functional Independence Measure (FIM) score.

SETTING

A single, freestanding inpatient rehabilitation hospital.

PATIENTS

Over a 20-month period, 25 patients with iGBM were matched with 25 patients admitted to an inpatient rehabilitation facility with rGBM by the following criteria: (1) side of lesion (left/right hemisphere), (2) admission total FIM score within 10 points, (3) age within 10 years, and (4) gender. Nineteen of the 25 patients in each group were matched meeting all criteria, and 6 of the 25 patients were matched meeting three out of four criteria.

INTERVENTIONS

Not applicable.

MAIN OUTCOME MEASURE

The primary outcome measures were differences in functional outcomes as measured by FIM scores.

RESULTS

There were no statistically significant differences (P < .05) between the groups in mean admission FIM scores, discharge FIM scores, FIM gains, and FIM efficiencies. There were no statistically significant differences in the development of complications during acute rehabilitation and transfer rate to acute care hospital. Sixty-four percent of patients in both groups were able to be discharged home.

CONCLUSIONS

This study demonstrated no statistically significant differences in functional outcomes between matched patients admitted with iGBM compared to rGBM. Further studies are indicated to examine the rehabilitation outcomes of patients with rGBM in inpatient rehabilitation.

A decision tree built with parameters obtained by computed tomographic pulmonary angiography is useful for predicting adverse outcomes in non-high-risk acute pulmonary embolism patients

Background

Acute pulmonary embolism (APE) is one of the leading causes of death in cardiovascular disease. The 30-day mortality can still be 1.7–15% in non-high-risk APE patients. Some non-high-risk patients can progress into the high-risk group and even die, which is referred to as an adverse outcome. Promoting the diagnosis and predictive ability of adverse short-term prognosis was still a problem that needed to be solved. Computed tomography pulmonary angiography (CTPA) may be a way to promote the predictive ability. Our aim to develop predictive tools based on parameters obtained by computed tomographic pulmonary angiography (CTPA) in the form of a decision tree for use in non-high-risk acute pulmonary embolism (APE) patients.

Methods

Adverse outcome was defined within 30 days after admission to the hospital. A decision tree was built to predict adverse outcomes based on discriminating factors screened from cardiac volume and clot characteristics from recursive partitioning analysis and compared with simplified pulmonary embolism severity index (sPESI), Bova scores and risk stratification. The area under the receiver operating characteristic curve (ROC-AUC) was used to confirm the predictive ability.

Results

A total of 38 patients with and 303 patients without adverse outcomes were enrolled. Right ventricular/left ventricular (RV/LV) volume ratio, central pulmonary artery (CPA) embolism and right atria/left atria (RA/LA) volume ratio were used as splits in the decision tree to predict adverse outcomes in all patients. The ROC-AUC was 0.858. In CPA embolism patients, a recursive partitioning analysis was performed with cardiac volume and novel clot burden, but only the obstructing area (OA) ratio was included as a discriminating factor to build a second decision tree. The ROC-AUC for the second decision tree was 0.810. The decision trees were superior to those of sPESI, Bova scores and risk stratification, and there were no significant differences between the two decision trees.

Conclusions

A decision tree built by CTPA parameters can predict adverse outcomes in non-high-risk APE patients.

Histopathologic quantification of viable tumor versus treatment effect in surgically resected recurrent glioblastoma

PURPOSE

The prognostic impact of the histopathologic features of recurrent glioblastoma surgical specimens is unknown. We sought to determine whether key histopathologic characteristics in glioblastoma tumors resected after chemoradiotherapy are associated with overall survival (OS).

METHODS

The following characteristics were quantified in recurrent glioblastoma specimens at our institution: extent of viable tumor (accounting for % of specimen comprised of tumor and tumor cellularity), mitoses per 10 high-power fields (0, 1-10, > 10), Ki-67 proliferative index (0-100%), hyalinization (0-6; none to extensive), rarefaction (0-6), hemosiderin (0-6), and % of specimen comprised of geographic necrosis (0-100%; converted to 0-6 scale). Variables associated with OS in univariate analysis, as well as age, eastern cooperative oncology group performance status (ECOG PS), extent of repeat resection, time from initial diagnosis to repeat surgery, and O⁶-methylguanine-DNA methyltransferase promoter methylation, were included in a multivariable Cox proportional hazards model.

RESULTS

37 specimens were assessed. In a multivariate model, high Ki-67 proliferative index was the only histopathologic characteristic associated with worse OS following repeat surgery for glioblastoma (hazard ratio (HR) 1.3, 95% CI 1.1-1.5, p = 0.003). Shorter time interval from initial diagnosis to repeat surgery (HR 1.11, 95% CI 1.02-1.21, p = 0.016) and ECOG PS ≥ 2 (HR 4.19, 95% CI 1.72-10.21, p = 0.002) were also independently associated with inferior OS.

CONCLUSION

In patients with glioblastoma undergoing repeat resection following chemoradiotherapy, high Ki-67 index in the recurrent specimen, short time to recurrence, and poor PS are independently associated with worse OS. Histopathologic quantification of viable tumor versus therapy-related changes has limited prognostic influence.