Metabolic signatures derived from whole-brain MR-spectroscopy identify early tumor progression in high-grade gliomas using machine learning.

PURPOSE: Recurrence for high-grade gliomas is inevitable despite maximal safe resection and adjuvant chemoradiation, and current imaging techniques fall short in predicting future progression. However, we introduce a novel whole-brain magnetic resonance spectroscopy (WB-MRS) protocol that delves into the intricacies of tumor microenvironments, offering a comprehensive understanding of glioma progression to inform expectant surgical and adjuvant intervention. METHODS: We investigated five locoregional tumor metabolites in a post-treatment population and applied machine learning (ML) techniques to analyze key relationships within seven regions of interest: contralateral normal-appearing white matter (NAWM), fluid-attenuated inversion recovery (FLAIR), contrast-enhancing tumor at time of WB-MRS (Tumor), areas of future recurrence (AFR), whole-brain healthy (WBH), non-progressive FLAIR (NPF), and progressive FLAIR (PF). Five supervised ML classification models and a neural network were developed, optimized, trained, tested, and validated. Lastly, a web application was developed to host our novel calculator, the Miami Glioma Prediction Map (MGPM), for open-source interaction. RESULTS: Sixteen patients with histopathological confirmation of high-grade glioma prior to WB-MRS were included in this study, totaling 118,922 whole-brain voxels. ML models successfully differentiated normal-appearing white matter from tumor and future progression. Notably, the highest performing ML model predicted glioma progression within fluid-attenuated inversion recovery (FLAIR) signal in the post-treatment setting (mean AUC = 0.86), with Cho/Cr as the most important feature. CONCLUSIONS: This study marks a significant milestone as the first of its kind to unveil radiographic occult glioma progression in post-treatment gliomas within 8 months of discovery. These findings underscore the utility of ML-based WB-MRS growth predictions, presenting a promising avenue for the guidance of early treatment decision-making. This research represents a crucial advancement in predicting the timing and location of glioblastoma recurrence, which can inform treatment decisions to improve patient outcomes.

Traumatic middle meningeal arteriovenous fistulas (MMAVFs): an exploratory systematic review.

This study aims to systematically review case reports and case series in order to compare the postoperative course of conservative, endovascular and surgical treatments for traumatic dural arteriovenous fistulas predominantly supplied by the middle meningeal artery (MMAVFs), which usually occur following head trauma or iatrogenic causes. We conducted a comprehensive search of PubMed, Embase, Scopus, Web of Science, and Google Scholar until June 23rd, 2024. Three cohorts were defined based on the treatment modality employed. The primary outcomes were the rates of overall obliteration and postoperative complications, with all-cause mortlality considered as secondary outcome. A total of 61 studies encompassing 78 pooled MMAVFs were included in the qualitative analysis. The predominant demographic consisted of males (53.9%) with a median age of 50.5 (IQR: 33.5-67.5) years. The main etiologies for fistula formation were head trauma (75.6%), cranial neurosurgical procedures (11.5%) and endovascular embolization (8.97%). Venous drainage patterns were categorized as follows based on anatomical confluence: Class I (16.7%), II (14.1%), III (12.8%), IV (14.1%), V (7.7%), and VI (3.9%). Regarding treatment efficacy, the overall obliteration rate was 89.74%, achieved through endovascular (95.83%), surgical (64.29%) or conservative (93.75%) approaches. In terms of safety, the overall postoperative complication rate was 6.49% with an all-cause mortality rate of 8.97%, predominantly observed in the surgical group (35.71%). Our systematic review highlights the challenging management of traumatic MMAVFs, frequently associated with head injuries. Endovascular therapy has emerged as the predominant treatment modality, demonstrating markedly higher rates of fistula obliteration, reduced all-cause mortality, and fewer postoperative complications.

Structure-sparing resection for the management of cervical chordomas: a retrospective institutional series.

OBJECTIVE: Chordomas are a rare and relatively slow-growing malignancy of notochordal origin with a nearly 50% recurrence rate. Chordomas of the cervical spine are particularly challenging tumors given surrounding vital anatomical structures. Although standard in other areas of the spine, en bloc resection of cervical chordomas is exceedingly difficult and carries the risk of significant postoperative morbidity. Here, the authors present their institutional experience with 13 patients treated with a structure-sparing radical resection and adjuvant radiation for cervical chordomas. METHODS: Records of the standing senior author and institutional database of spinal surgeries were retrospectively reviewed for surgically managed cervical and high thoracic chordomas between 1997 and 2022. Chordomas whose epicenter was cervical but touched the clivus or had extension to the thoracic spine were included in this series. Clinical and operative data were gathered and analyzed for the index surgery and any revisions needed. Outcome metrics such as recurrence rates, complication rates, functional status, progression-free interval (PFI) and overall survival (OS) were evaluated. RESULTS: The median patient age at diagnosis was 57 (range 32-80) years. The median modified Rankin Scale (mRS) score at the time of presentation was 1 (range 0-4). Approximately 40% of tumors were located in the upper cervical spine (occiput-C2). The median time from diagnosis to surgery was 74.5 (range 10-483) days. Gross-total resection was achieved in just under 40% of patients. All patients received adjuvant radiotherapy. The mean duration of follow-up was 4.09 years, with a mean PFI of 3.80 (range 1.16-13.1) years. Five patients experienced recurrence (38.5%). The mean OS was 3.44 years. Three patients died during the follow-up period; 2 due to disease progression and 1 died in the immediate postoperative period. One patient was lost to follow-up. A significant positive relationship was identified between high cervical tumor location and disease recurrence (p = 0.021). CONCLUSIONS: While en bloc resection is appropriate and feasible for tumors in the sacral spine, the cervical region poses a significant technical challenge and is associated with increased postoperative morbidity. Radical resection may allow for achievement of negative operative margins and, along with sparing postoperative morbidity following resection of cervical chordomas, maintaining a similar rate of recurrence when compared with en bloc resection while preserving quality of life.

Leveraging machine learning for preoperative prediction of supramaximal ablation in laser interstitial thermal therapy for brain tumors.

OBJECTIVE: Maximizing safe resection in neuro-oncology has become paramount to improving patient survival and outcomes. Laser interstitial thermal therapy (LITT) offers similar survival benefits to traditional resection, alongside shorter hospital stays and faster recovery times. The extent of ablation (EOA) achieved using LITT is linked to patient outcomes, with greater EOA correlating with improved outcomes. However, the preoperative predictors for achieving supramaximal ablation (EOA ≥ 100%) are not well understood. By leveraging machine learning (ML) techniques, this study aimed to identify these predictors to enhance patient selection and therefore outcomes. The objective was to explore preoperative predictors for supramaximal EOA using ML in patients with glioblastoma. METHODS: A retrospective study was conducted on the medical records of 254 patients undergoing LITT from 2013 to 2023 at a single tertiary center. Cohort criteria included age ≥ 18 years, diagnosis of glioblastoma, single-trajectory ablation, and a complete dataset. The study assessed preoperative clinical and radiographic factors, using EOA ≥ 100% as the endpoint. Five ML models were used: logistic regression, random forest (RF), gradient boosting, Gaussian naive Bayes, and support vector machine. Training and testing cohorts were subsequently assessed across ML models with fivefold cross-validation. Models were optimized using hyperparameter tuning. Performance was primarily quantified using the area under the curve (AUC) of the receiver operating characteristic curve. RESULTS: The final cohort consisted of 72 patients. Among the ML models, RF achieved the highest AUC (mean ± SD 0.94 ± 0.06). The leading models identified that lower preoperative volume, history of prior radiation therapy, history of prior craniotomy, preoperative neurological deficits, history of preoperative seizures, and distance from intracranial heat sinks were predictive of successful ablations in patients. Additionally, RF had the best mean metrics: accuracy 0.88, precision 0.87, specificity 0.87, and sensitivity 0.89. CONCLUSIONS: This is the first study to investigate the role of ML for optimizing ablation volumes in LITT. These ML models suggest that low preoperative volumes, previous craniotomy, previous radiation therapy, no previous neurological deficits, larger catheter-heat sink distance, and the presence of preoperative seizures are important prognostic factors for predicting successful supramaximal ablations with LITT.

Hearing preservation and quality of life in small to medium sized vestibular schwannomas after a wait and scan approach or stereotactic radiosurgery: a systematic review and meta-analysis.

BACKGROUND: The management of vestibular schwannomas (VS) encompasses a choice between conservative "wait-and-scan" (WAS) approach, stereotactic radiosurgery (SRS) or open microsurgical resection. Currently, there is no consensus on the optimal management approach for small to medium sized VS. This study aims to compared outcomes related to hearing in patients with small and medium sized VS who underwent initial treatment with WAS versus SRS. METHODS: A systematic review of the available literature was conducted using PubMed/MEDLINE, Embase, and Cochrane up December 08, 2023. Meta-analysis was performed using a random-effect model to calculate mean difference (MD) and relative risk (RR). A leave-one-out analysis was conducted. The risk of bias was assessed via the Risk of Bias in Non-randomized Studies-Interventions (ROBINS-I) and Cochrane Risk of Bias assessment tool (RoB-2). Ultimately, the certainty of evidence was evaluated using the GRADE assessment. The primary outcomes were serviceable hearing, and pure-tone average (PTA). The secondary outcome was the Penn Acoustic Neuroma Quality of Life Scale (PANQOL) total score. RESULTS: Nine studies were eligible for inclusion, comprising a total of 1,275 patients. Among these, 674 (52.86%) underwent WAS, while 601 patients (47.14%) received SRS. Follow-up duration ranged from two to eight years. The meta-analysis indicated that WAS had a better outcome for serviceable hearing (0.47; 95% CI: 0.32 - 0.68; p < 0.001), as well as for postoperative functional measures including PTA score (MD 13.48; 95% CI 3.83 - 23.13; p < 0.01), and PANQOL total score (MD 3.83; 95% CI 0.42 - 7.25; p = 0.03). The overall certainty of evidence ranged from "very low" to "moderate". CONCLUSIONS: Treating small to medium sized VS with WAS increases the likelihood of preserving serviceable hearing and optimized PANQOL overall postoperative score compared to SRS. Nevertheless, the limited availability of literature and the methodological weakness observed in existing studies outline the need for higher-quality studies.

Towards the definition of progressive disease in brain metastasis treated with laser ablation: an evidence-based study.

PURPOSE: The postoperative period after laser interstitial thermal therapy (LITT) is marked by a temporary increase in volume, which can impact the accuracy of radiographic assessment. The current criteria for progressive disease (PD) suggest that a 20% increase in size of brain metastasis (BM) assessed in 6-12 weeks intervals should be considered as local progression (LP). However, there is no agreement on how LP should be defined in this context. In this study, we aimed to statistically analyze which tumor volume variations were associated with LP. METHODS: We analyzed 40 BM that underwent LITT between 2013 and 2022. For this study, LP was defined following radiographic features. A ROC curve was generated to evaluate volume change as a predictor of LP and find the optimal cutoff point. A logistic regression analysis and Kaplan Meier curves were performed to assess the impact of various clinical variables on LP. RESULTS: Out of 40 lesions, 12 (30%) had LP. An increase in volume of 25.6% from baseline within 120-180 days after LITT presented a 70% sensitivity and 88.9% specificity for predicting LP (AUC: 0.78, p = 0.041). The multivariate analysis showed a 25% increase in volume between 120 and 180 days as a negative predictive factor (p = 0.02). Volumetric changes within 60-90 days after LITT did not predict LP (AUC: 0.57; p = 0.61). CONCLUSION: Volume changes within the first 120 days after the procedure are not independent indicators of LP of metastatic brain lesions treated with LITT.