Neuro-oncology advances最新文献

Purpose: To implement and evaluate deep learning-based methods for the classification of pediatric brain tumors (PBT) in magnetic resonance (MR) data.

Methods: A subset of the "Children's Brain Tumor Network" dataset was retrospectively used (n = 178 subjects, female = 72, male = 102, NA = 4, age range [0.01, 36.49] years) with tumor types being low-grade astrocytoma (n = 84), ependymoma (n = 32), and medulloblastoma (n = 62). T1w post-contrast (n = 94 subjects), T2w (n = 160 subjects), and apparent diffusion coefficient (ADC: n = 66 subjects) MR sequences were used separately. Two deep learning models were trained on transversal slices showing tumor. Joint fusion was implemented to combine image and age data, and 2 pre-training paradigms were utilized. Model explainability was investigated using gradient-weighted class-activation mapping (Grad-CAM), and the learned feature space was visualized using principal component analysis (PCA).

Results: The highest tumor-type classification performance was achieved when using a vision transformer model pre-trained on ImageNet and fine-tuned on ADC images with age fusion (Matthews correlation coefficient [MCC]: 0.77 ± 0.14, Accuracy: 0.87 ± 0.08), followed by models trained on T2w (MCC: 0.58 ± 0.11, Accuracy: 0.73 ± 0.08) and T1w post-contrast (MCC: 0.41 ± 0.11, Accuracy: 0.62 ± 0.08) data. Age fusion marginally improved the model's performance. Both model architectures performed similarly across the experiments, with no differences between the pre-training strategies. Grad-CAMs showed that the models' attention focused on the brain region. PCA of the feature space showed greater separation of the tumor-type clusters when using contrastive pre-training.

Conclusion: Classification of PBT on MR images could be accomplished using deep learning, with the top-performing model being trained on ADC data, which radiologists use for the clinical classification of these tumors.

Background: Awake craniotomy (AC) is a technique that balances maximum resection and minimal postoperative deficits in patients with intracranial tumors. To aid in the comparability of functional outcomes after awake surgery, this study investigated its international practice and aimed to define categories of postoperative deficits.

Methods: A survey was distributed via neurosurgical networks in Europe (European Association of Neurosurgical Societies, EANS), the Netherlands (Nederlandse Vereniging voor Neurochirurgie, NVVN), Belgium (Belgian Society of Neurosurgery, BSN), and the United States (Congress of Neurological Surgeons, CNS) between April 2022 and April 2023. Questions involved decision-making, including patient selection, anxiety assessment, and termination of resection. Interpretation of "major" and "minor" deficits, respectively labeled "level I" and "level II," was assessed.

Results: Three hundred and ninety-five neurosurgeons from 46 countries completed the survey. Significant heterogeneity was found in the domains of indications, anxiety assessment, seizure management, and termination of resection. Moreover, the interpretation of "major" deficits mainly included language and motor impairments. Analysis across deficit categories showed significant overlap in the domains of executive function, social cognition, and vision. Secondly, "minor" deficits and "minor cognitive" deficits showed vast overlap.

Conclusions: This survey demonstrates high variability between neurosurgeons in AC practice across multiple domains, inviting international efforts to reach a consensus regarding the standardization and grading of postoperative deficits. The proposed categories of "level I" and "level II" deficits may aid in this standardization. It allows for systematic assessment of the benefit of surgery in neuro-oncology patients and allows for comparison of surgical outcomes between institutions and surgeons. This may help to optimize international guidelines for surgical neuro-oncology, including AC.

The phase-3 INDIGO trial demonstrated that the isocitrate dehydrogenase (IDH) inhibitor vorasidenib significantly prolonged progression-free survival and delayed intervention in patients with CNS WHO grade 2 gliomas. However, conventional MRI showed limited response, with only 11% of patients having objective responses. Studies suggest that serial PET imaging with radiolabeled amino acids, such as O -(2-[¹⁸ F]-fluoroethyl)-L-tyrosine (FET) PET, may provide earlier and more informative assessments of treatment response than MRI. In an initial experience with FET PET, 3 out of 5 patients showed metabolic response to vorasidenib. This highlights FET PET's potential to guide decision-making, though further trials are needed to confirm outcome benefits.

Background: Alterations in cellular metabolism affect cancer survival and can manifest in metrics of body composition. We investigated the effects of various body composition metrics on survival in patients with glioblastoma (GBM).

Methods: We retrospectively analyzed patients who had an abdominal and pelvic computed tomography (CT) scan performed within 1 month of diagnosis of GBM (178 participants, 102 males, 76 females, median age: 62.1 years). Volumetric body composition metrics were derived using automated CT segmentation of adipose tissue, skeletal muscle, and aortic calcification from L1 to L5. Univariable and multivariable Cox proportional hazards models were performed separately in males and females using known predictors of GBM overall survival (OS) as covariates. A sex-specific composite score of predisposing and protective factors was constructed using the relative importance of each metric in GBM OS.

Results: Higher skeletal muscle volume and lower skeletal muscle fat fraction were associated with better OS in the entire dataset. A robust and independent effect on GBM OS was seen specifically for fraction of inter/intramuscular adipose tissue to total adipose tissue after correction for known survival predictors and comorbidities. Worse OS was observed with increased abdominal aortic calcification volume in both sexes. There was a significant difference in GBM OS among participants stratified into quartiles based on sex-specific composite predisposing and protective scores.

Conclusion: The relationship between body composition and GBM OS provides an actionable advancement toward precision medicine in GBM management, as lifestyle and dietary regimens can alter body composition and metabolism and from there GBM survival.

Background: The EF-14 clinical trial demonstrated the safety and efficacy of tumor-treating fields (TTFields) for newly diagnosed glioblastoma. This study aimed to clarify the current status, safety, and efficacy of TTFields in Japanese patients who meet the EF-14 inclusion criteria.

Methods: This was a multicenter retrospective cohort study. Background, treatment, and outcome data of patients who satisfied the inclusion criteria of the EF-14 trial were collected from 45 institutions across Japan. The rate, determinants, and current status of TTField use, including its safety and efficacy in terms of progression and survival, were retrospectively investigated. This study was conducted in accordance with the STROBE checklist.

Results: Among the 607 patients enrolled, 70 were excluded due to progressive disease during radiation and temozolomide therapy, age > 80 years old, and Karnofsky Performance Status score of <70. Among the remaining 537 patients, 210 (39%) underwent TTField treatment. Multivariate analysis revealed younger age and spouse as a caregiver as significant factors for TTField use. The compliance rate of TTField use exceeded 75% in 60% of patients, with a median TTField usage duration of 11 months. Skin disorders requiring medical treatment occurred in 56% of patients. Multivariate Cox proportional hazards analysis in the whole series and propensity score-matched analysis revealed that TTField use was not a prognostic factor for progression-free survival (PFS) or overall survival (OS).

Conclusions: TTField use did not have a substantial effect on either PFS or OS in Japanese patients with glioblastoma, despite compliance rates comparable to those observed in the EF-14.

Background: Glioblastoma (GBM) has a median survival of <2 years. Pexidartinib (PLX3397) is a small-molecule inhibitor of CSF1R, KIT, and oncogenic FTL3, which are implicated in GBM treatment resistance. Results from glioma models indicate that combining radiation therapy (RT) and pexidartinib reduces radiation resistance. We added pexidartinib to standard-of-care RT/temozolomide (TMZ) in patients with newly diagnosed GBM to assess the therapeutic benefit of altering the tumor microenvironment with pexidartinib.

Methods: In this open-label, dose-escalation, multicenter, Phase 1b/2 trial, pexidartinib was administered in combination with RT/TMZ followed by adjuvant pexidartinib + TMZ. During Phase 1b, pexidartinib was given 5 or 7 days/week at multiple dosing levels. The primary Phase 1b endpoint was the recommended Phase 2 dose (RP2D). Phase 2 patients received the RP2D with the primary endpoint of median progression-free survival (mPFS). Secondary objectives were median overall survival (mOS), pharmacokinetics, and safety.

Results: The RP2D of pexidartinib was 800 mg/day for 5 days/week during RT/TMZ, followed by 800 mg/day for 7 days/week with adjuvant TMZ. mPFS was 6.7 months (90% CI: 4.5, 11.5) for the modified intention-to-treat population. The actual mOS was 13.1 months (90% CI: 11.5, 24.5), and the mOS corrected for comparison with matched historical controls was 18.8 months (95% CI: 12.6, 28.0).

Conclusions: This trial established the RP2D of pexidartinib in combination with RT/TMZ and adjuvant TMZ. Pexidartinib was generally safe and well tolerated. Although the study regimen with pexidartinib was not efficacious, pharmacodynamic studies showed modulation of systemic markers that could lead to alteration of the tumor microenvironment.

Background: Glioblastoma, a lethal high-grade glioma, has not seen improvements in clinical outcomes in nearly 30 years. Ion channels are increasingly associated with tumorigenesis, and there are hundreds of brain-penetrant drugs that inhibit ion channels, representing an untapped therapeutic resource. The aim of this exploratory drug study was to screen an ion channel drug library against patient-derived glioblastoma cells to identify new treatments for brain cancer.

Methods: Seventy-two ion channel inhibitors were screened in patient-derived glioblastoma cells, and cell viability was determined using the ViaLight Assay. Cell cycle and apoptosis analysis were determined with flow cytometry using PI and Annexin V staining, respectively. Protein and phosphoprotein expression was determined using mass spectrometry and analyzed using gene set enrichment analysis. Kaplan-Meier survival analyses were performed using intracranial xenograft models of GBM6 and WK1 cells.

Results: The voltage-gated sodium channel modulator, DPI-201-106, was revealed to reduce glioblastoma cell viability in vitro by inducing cell cycle arrest and apoptosis. Phosphoproteomics indicated that DPI-201-106 may impact DNA damage response (DDR) pathways. Combination treatment of DPI-201-106 with the CHK1 inhibitor prexasertib or the PARP inhibitor niraparib demonstrated synergistic effects in multiple patient-derived glioblastoma cells both in vitro and in intracranial xenograft mouse models, extending survival of glioblastoma-bearing mice.

Conclusions: DPI-201-106 enhances the efficacy of DDR inhibitors to reduce glioblastoma growth. As these drugs have already been clinically tested in humans, repurposing DPI-201-106 in novel combinatorial approaches will allow for rapid translation into the clinic.

Background: Following large resection, proposing a watch-and-wait strategy in selected grade 3 glioma, isocitrate dehydrogenase (IDH)-mutant patients is an emerging practice. We compared the watch-and-wait approach to the standard postoperative adjuvant oncological treatment for grade 3 gliomas, IDH-mutant.

Methods: Observational, retrospective, single-institution cohort (2011-2023) of 106 consecutive adult patients harboring supratentorial grade 3 gliomas, IDH-mutant treated by maximal awake resection and who received a watch-and-wait approach (surgery group) or an adjuvant oncological treatment (oncological group) postoperatively. Case-matched analysis (1:1) criteria between the surgery group and oncological group: extent of resection, tumor volume, Karnofsky Performance Status (KPS) score, tumor location and size, and age.

Results: Patients of the surgery group (n = 26) had significantly better KPS scores, less preoperative neurological and/or neurocognitive deficits, less hyperperfusion, less corpus callosum infiltration, smaller tumor volume, higher rate of total resection, and smaller residual tumor than patients of the oncological group (n = 80). The 5-year progression-free survival (66.2 vs. 77.9 months, P = .713) and the 5-year overall survival (88.9 vs. 83.9 months, P = .291) did not differ between surgery and oncological groups. In the whole series, a preoperative KPS score >70, a total resection, and the oligodendroglioma subtype were independent predictors of longer progression-free survival and overall survival. After case matching, no difference in survival was observed between watch-and-wait and oncological treatment both in astrocytomas (n = 14 per group) and oligodendrogliomas (n = 12 per group).

Conclusions: Watch-and-wait following radical resection appears to be feasible in highly selected grade 3 gliomas, IDH-mutant patients without impairing survival both in astrocytoma and in oligodendroglioma subgroups.

Background: Magnetic resonance imaging (MRI) cerebral blood volume (CBV) measurements improve the diagnosis of recurrent gliomas. The study investigated the prognostic value of dynamic contrast-enhanced (DCE) CBV imaging in treated IDH wildtype glioblastoma when added to MRI or amino acid positron emission tomography (PET).

Methods: Hybrid [¹⁸F]FET PET/MRI with 2CXM (2-compartment exchange model) DCE from 86 adult patients with suspected recurrent or residual glioblastoma were retrospectively analyzed. High CBV tumor volume (VOL_CBV), and contrast-enhancing (VOL_CE) and [¹⁸F]FET active tumor (VOL_FET) volumes were delineated. Absolute and fractional high CBV subvolumes within VOL_CE and VOL_FET were determined. Associations with overall survival (OS) were assessed by Cox analysis.

Results: Adjusted for methyltransferase gene status and steroid use all total tumor volumes were individually associated with shorter OS. Adding VOL_CBV to VOL_CE or VOL_FET only the effect of VOL_CBV was prognostic of OS (hazard ratio [HR] 1.327, P = .042 and 1.352, P = .011, respectively). High CBV subvolumes within both VOL_CE and VOL_FET were associated with shorter survival (HR 1.448, P = .042 and 1.416, P = .011, respectively), and the low CBV subvolumes with longer survival (HR 0.504, P = .002 and .365, P = .001, respectively). The fraction of VOL_CE and VOL_FET with high CBV was a strong predictor of OS with shorter median OS in upper versus lower tertiles (8.3 vs 14.5 months and 7.1 vs 15.6 months, respectively, both P < .001).

Conclusions: The high CBV tumor volume was a strong prognosticator of survival and allowed for the separation of high- and low-risk subvolumes underlining the heterogeneous physiological environment represented in the contrast-enhancing or metabolically active tumor volumes of treated glioblastoma.

Background: Brain metastasis invasion pattern (BMIP) is an emerging biomarker associated with recurrence-free and overall survival in patients, and differential response to therapy in preclinical models. Currently, BMIP can only be determined from the histopathological examination of surgical specimens, precluding its use as a biomarker prior to therapy initiation. The aim of this study was to investigate the potential of machine learning (ML) approaches to develop a noninvasive magnetic resonance imaging (MRI)-based biomarker for BMIP determination.

Methods: From an initial cohort of 329 patients, a subset of 132 patients met the inclusion criteria for this retrospective study. We evaluated the ability of an expert neuroradiologist to reliably predict BMIP. Thereafter, the dataset was randomly divided into training/validation (80% of cases) and test subsets (20% of cases). The ground truth for BMIP was the histopathologic evaluation of resected specimens. Following MRI sequence co-registration, advanced feature extraction techniques deriving hand-crafted radiomic features with traditional ML classifiers and convolution-based deep learning (CDL) models were trained and evaluated. Different ML approaches were used individually or using ensembling techniques to determine the model with the best performance for BMIP prediction.

Results: Expert evaluation of brain MRI scans could not reliably predict BMIP, with an accuracy of 44%-59% depending on the semantic feature used. Among the different ML and CDL models evaluated, the best-performing model achieved an accuracy of 85% and an F1 score of 90%.

Conclusions: ML approaches can effectively predict BMIP, representing a noninvasive MRI-based approach to guide the management of patients with brain metastases.