Bacterial meningitis is not rare in post-neurosurgical patients. If patients are not treated promptly, the mortality rate can reach 20 to 50 %. The concentration of cerebrospinal fluid (CSF) lactate has been reported to be helpful in the diagnosis of bacterial meningitis; however, no systematic evaluations have investigated CSF from a postoperative perspective. In this study, we performed a systematic evaluation and meta-analysis of the efficacy of using CSF lactate concentrations in the diagnosis of post-neurosurgical bacterial meningitis. We retrieved studies that investigated the diagnostic value of CSF lactate for the diagnosis of post-neurosurgical bacterial meningitis by searching PubMed, EBSCO, the Cochrane Library and ClinicalTrials.gov. All these databases were searched from inception to November 2015. We used Quality Assessment of Diagnostic Accuracy Studies (QUADAS), a tool for the quality assessment of diagnostic accuracy, to evaluate the quality of the included studies. The Meta-DiSc 1.4 and Review Manager 5.3 software programs were used to analyze the included studies. Forest plots and summary receiver operating characteristics (SROC) curves were also drawn. Five studies, involving a total of 404 post-neurosurgical patients, were selected from 1,672 articles according to the inclusion criteria. The quality of the five included studies was assessed using QUADAS, and the related results are presented in tables. The meta-analysis revealed the following diagnostic values regarding CSF lactate for post-neurosurgical bacterial meningitis: a pooled sensitivity of 0.92 (95 % CI 0.85–0.96), a pooled specificity of 0.88 (95 % CI 0.84–0.92 with significant heterogeneity), a diagnostic odds ratio of 83.09 (95 % CI 36.83–187.46), an area under the curve (AUCSROC) of 0.9601, an SE(AUC) of 0.0122, a Q* of 0.9046 and an SE(Q*) of 0.0179. The meta-analysis indicated that the CSF lactate concentration has relatively high sensitivity and specificity for the diagnosis of post-neurosurgical bacterial meningitis and thus has relatively good efficacy.

Background: The radiological differential diagnosis between tumor recurrence and radiation-induced necrosis (ie, pseudoprogression) is of paramount importance in the management of glioma patients.Objective: This research aims to develop a deep learning methodology for automated differentiation of tumor recurrence from radiation necrosis based on routine magnetic resonance imaging (MRI) scans.Methods: In this retrospective study, 146 patients who underwent radiation therapy after glioma resection and presented with suspected recurrent lesions at the follow-up MRI examination were selected for analysis. Routine MRI scans were acquired from each patient, including T1, T2, and gadolinium-contrast-enhanced T1 sequences. Of those cases, 96 (65.8%) were confirmed as glioma recurrence on postsurgical pathological examination, while 50 (34.2%) were diagnosed as necrosis. A light-weighted deep neural network (DNN) (ie, efficient radionecrosis neural network [ERN-Net]) was proposed to learn radiological features of gliomas and necrosis from MRI scans. Sensitivity, specificity, accuracy, and area under the curve (AUC) were used to evaluate performance of the model in both image-wise and subject-wise classifications. Preoperative diagnostic performance of the model was also compared to that of the state-of-the-art DNN models and five experienced neurosurgeons.Results: DNN models based on multimodal MRI outperformed single-modal models. ERN-Net achieved the highest AUC in both image-wise (0.915) and subject-wise (0.958) classification tasks. The evaluated DNN models achieved an average sensitivity of 0.947 (SD 0.033), specificity of 0.817 (SD 0.075), and accuracy of 0.903 (SD 0.026), which were significantly better than the tested neurosurgeons (P=.02 in sensitivity and P<.001 in specificity and accuracy).Conclusions: Deep learning offers a useful computational tool for the differential diagnosis between recurrent gliomas and necrosis. The proposed ERN-Net model, a simple and effective DNN model, achieved excellent performance on routine MRI scans and showed a high clinical applicability.

Objectives. To compare 3D-CT-guided and C-arm-guided percutaneous balloon compression (PBC) in terms of effectiveness and safety. Methods. The medical records and follow-up data of patients with idiopathic trigeminal neuralgia who underwent 3D-CT-guided or C-arm-guided PBCs in Beijing Tiantan Hospital and the Characteristic Medical Center of the Chinese People’s Armed Police Force between February 2018 and March 2020 were retrospectively reviewed and analysed. Results. A total of 291 patients were included. Among them, 212 patients underwent PBC treatment with 3D-CT and others with C-arm. One (0.5%) patient in 3D-CT group and 4 (5.1%) patients in C-arm group failed to receive PBC treatment because of failure of foramen ovale (FO) puncture ( ). Among patients with successful attempts, 5 (2.4%) patients in the 3D-CT group and 11 (14.7%) patients in the C-arm group received more than one needle pass during the procedure ( ). The 3D-CT group required less time than the C-arm group for puncture ( ) and for the whole operation ( ). The groups shared similar initial relief rates ( ) and similar recurrence-free survival during follow-ups for a median of 22 months ( ). No puncture-related complications occurred in either group and the two groups had similar incidences of compression-related complications. Conclusion. 3D-CT facilitated FO puncture and improved success rate of PBC. The overall time efficiency of PBC was also increased with 3D-CT. Thus, 3D-CT is a potentially useful image guidance technology for treating idiopathic trigeminal neuralgia by PBC.

The objective of this study was to investigate IL13Ra2 expression in brainstem glioma (BSG) and its correlation with key markers, functions, and prognostic implications, evaluating its therapeutic potential. A total of 80 tumor samples from BSG patients were analyzed. Multiplex immunofluorescence was used to examine six markers—IL13Ra2, H3.3K27M, CD133, Ki67, HLA-1, and CD4—establishing relationships between IL13Ra2 and these markers. Survival analysis, employing Kaplan–Meier and Cox proportional hazard regression models, encompassed 66 patients with complete follow-up. RNA-Seq data from a previously published study involving 98 patients were analyzed using the DESeq2 library to determine differential gene expression between groups. Gene Ontology (GO) enrichment and single-sample gene set enrichment analysis (ssGSEA) via the clusterProfiler library were used to delineate the gene functions of differentially expressed genes (DEGs). Nearly all the BSG patients displayed varying IL13Ra2 expression, with 45.0% (36/80) exhibiting over a 20% increase. Elevated IL13Ra2 levels were notably observed in pontine gliomas, diffuse intrinsic pontine gliomas (DIPGs), H3F3A-mutant gliomas, and WHO IV gliomas. IL13Ra2 expression was strongly correlated with H3.3K27M mutant protein, Ki67, and CD133. Patients with IL13Ra2 expression >20% showed shorter overall survival compared to those with ≤20% IL13Ra2 expression. The Cox proportional hazard regression model identified H3F3A mutations, rather than IL13Ra2 expression, as an independent prognostic factor. Analysis of RNA-Seq data from our prior cohort confirmed IL13Ra2’s correlation with H3.3, CD133, and Ki67 levels. Widespread IL13Ra2 expression in BSG, particularly elevated in the H3F3A mutant group, was strongly correlated with H3F3A mutations, increased proliferation, and heightened tumor stemness. IL13Ra2 represents a promising therapeutic target for BSGs, potentially benefiting patients with H3K27M mutations, DIPGs, WHO Grade IV, and pontine location-specific BSGs, particularly those with H3K27M mutations.

The inter-tumor heterogeneity of the tumor microenvironment (TME) and how it correlates with clinical profiles and biological characteristics in brainstem gliomas (BSGs) remain unknown, dampening the development of novel therapeutics against BSGs. The TME status was determined with a list of pan-cancer conserved gene expression signatures using a single-sample gene set enrichment analysis (ssGSEA) and was subsequently clustered via consensus clustering. BSGs exhibited a high inter-tumor TME heterogeneity and were classified into four clusters: “immune-enriched, fibrotic”, “immune-enriched, non-fibrotic”, “fibrotic”, and “depleted”. The “fibrotic” cluster had a higher proportion of diffuse intrinsic pontine gliomas (p = 0.041), and “PA-like” tumors were more likely to be “immune-enriched, fibrotic” (p = 0.044). The four TME clusters exhibited distinct overall survival (p < 0.001) and independently impacted BSG outcomes. A four-gene panel as well as a radiomics approach were constructed to identify the TME clusters and achieved high accuracy for determining the classification. Together, BSGs exhibited high inter-tumor heterogeneity in the TME and were classified into four clusters with distinct clinical outcomes and tumor biological properties. The TME classification was accurately identified using a four-gene panel that can potentially be examined with the immunohistochemical method and a non-invasive radiomics method, facilitating its clinical application.

BACKGROUND The radiological differential diagnosis between tumor recurrence and radiation-induced necrosis (ie, pseudoprogression) is of paramount importance in the management of glioma patients. OBJECTIVE This research aims to develop a deep learning methodology for automated differentiation of tumor recurrence from radiation necrosis based on routine magnetic resonance imaging (MRI) scans. METHODS In this retrospective study, 146 patients who underwent radiation therapy after glioma resection and presented with suspected recurrent lesions at the follow-up MRI examination were selected for analysis. Routine MRI scans were acquired from each patient, including T1, T2, and gadolinium-contrast-enhanced T1 sequences. Of those cases, 96 (65.8%) were confirmed as glioma recurrence on postsurgical pathological examination, while 50 (34.2%) were diagnosed as necrosis. A light-weighted deep neural network (DNN) (ie, efficient radionecrosis neural network [ERN-Net]) was proposed to learn radiological features of gliomas and necrosis from MRI scans. Sensitivity, specificity, accuracy, and area under the curve (AUC) were used to evaluate performance of the model in both image-wise and subject-wise classifications. Preoperative diagnostic performance of the model was also compared to that of the state-of-the-art DNN models and five experienced neurosurgeons. RESULTS DNN models based on multimodal MRI outperformed single-modal models. ERN-Net achieved the highest AUC in both image-wise (0.915) and subject-wise (0.958) classification tasks. The evaluated DNN models achieved an average sensitivity of 0.947 (SD 0.033), specificity of 0.817 (SD 0.075), and accuracy of 0.903 (SD 0.026), which were significantly better than the tested neurosurgeons (P=.02 in sensitivity and P<.001 in specificity and accuracy). CONCLUSIONS Deep learning offers a useful computational tool for the differential diagnosis between recurrent gliomas and necrosis. The proposed ERN-Net model, a simple and effective DNN model, achieved excellent performance on routine MRI scans and showed a high clinical applicability.