Prediction Model for Unfavorable Outcome in Spontaneous Intracerebral Hemorrhage Based on Machine Learning.

OBJECTIVE: The spontaneous intracerebral hemorrhage (ICH) remains a significant cause of mortality and morbidity throughout the world. The purpose of this retrospective study is to develop multiple models for predicting ICH outcomes using machine learning (ML). METHODS: Between January 2014 and October 2021, we included ICH patients identified by computed tomography or magnetic resonance imaging and treated with surgery. At the 6-month check-up, outcomes were assessed using the modified Rankin Scale. In this study, four ML models, including Support Vector Machine (SVM), Decision Tree C5.0, Artificial Neural Network, Logistic Regression were used to build ICH prediction models. In order to evaluate the reliability and the ML models, we calculated the area under the receiver operating characteristic curve (AUC), specificity, sensitivity, accuracy, positive likelihood ratio (PLR), negative likelihood ratio (NLR), diagnostic odds ratio (DOR). RESULTS: We identified 71 patients who had favorable outcomes and 156 who had unfavorable outcomes. The results showed that the SVM model achieved the best comprehensive prediction efficiency. For the SVM model, the AUC, accuracy, specificity, sensitivity, PLR, NLR, and DOR were 0.91, 0.92, 0.92, 0.93, 11.63, 0.076, and 153.03, respectively. For the SVM model, we found the importance value of time to operating room (TOR) was higher significantly than other variables. CONCLUSION: The analysis of clinical reliability showed that the SVM model achieved the best comprehensive prediction efficiency and the importance value of TOR was higher significantly than other variables.

Attenuates reactive oxygen species: induced pyroptosis via activation of the Nrf2/HO-1 signal pathway in models of trigeminal neuralgia.

In this study, we examined the impact of demyelinating and neuroinflammation on trigeminal neuralgia (TN) by utilizing models of chronic constriction injury to the infraorbital nerve (CCI). The CCI rats were treated with either VX-765 (an inhibitor of caspase-1) or a control solution of PBS/DMSO to observe the effects on neurobehavioral and neuropathological outcomes. The histochemical changes, pyroptosis-related proteins were assessed using immunohistochemistry, Elisa, and western blotting. RSC96 cells were pretreated with belnacasan (VX-765, an inhibitor of caspase-1), Gasdermin D(GSDMD)-targeting siRNAs, cobalt protoporphyrin (CoPP) or zinc protoporphyrin (Znpp) before being exposed to H2O2. Following these treatments, the Reactive oxygen species (ROS) level, cell viability, percentage of pyroptosis, pyroptosis-related proteins, nuclear factor erythroid 2-related factor 2 (Nrf2) and HO-1 level was measured. The scanning electron microscopy revealed increased ball-like bulge and membrane pore formation in the CCI group. In the CCI and CCI+ Vehicle groups, we found ROS level and expression of pyroptosis-related proteins increased. While, treatment with VX-765resulted in a decreased expression of GSDMD, IL-1, IL-18, and caspase-1 decreased. In the in-vitro study, RSC96 cells showed mild pyroptosis and overall mild edema after being exposed to H2O2. The ROS level, percentage of pyroptosis, pyroptosis-related proteins, Nrf2 and HO-1 level increased significantly in the H2O2 group. While, the percentage of pyroptosis and pyroptosis-related proteins decreased significantly in the H2O2 + VX-765 group, H2O2 + siRNA group, and H2O2 + VX-765 + siRNA group. After treatment with HO-1-inhibitor Znpp and HO-1-activator Copp, the percentage of pyroptosis and pyroptosis-related proteins increased and decreased significantly respectively. In conclusions, the pyroptosis of Schwann cell in the CCI model generated the demyelination of TN nerve. The ROS is an upstream event of NLRP3 inflammasome activation which induced eventual pyroptosis. The Nrf2/HO-1 signaling pathway could protect the H2O2-induced pyroptosis in RSC96 cells.

Predictive nomogram for deep brain stimulation-related infections.

OBJECTIVE: Infection is one of the important and frequent complications following implantable pulse generator and deep brain stimulation (DBS) electrode insertion. The goal of this study was to retrospectively evaluate and identify potential risk factors for DBS infections. METHODS: From January 2015 to January 2021 in Qingdao municipal hospital (training cohort) and The First Affiliated Hospital of the University of Science and Technology of China (validation cohort), the authors enrolled patients with Parkinson disease who had undergone primary DBS placement or implantable pulse generator replacement. The cases were divided into infection or no-infection groups according to the 6-month follow-up. The authors used the logistic regression models to determine the association between the variables and DBS infection. Depending on the results of logistic regression, the authors established a nomogram. The calibration curves, receiver operating characteristic curve analysis, and decision curves were used to evaluate the reliability of the nomogram. RESULTS: There were 191 cases enrolled in the no-infection group and 20 cases in the infection group in the training cohort. The univariate logistic regression showed that BMI, blood glucose, and albumin were all significant predictors of infection after DBS surgery (OR 0.832 [p = 0.009], OR 1.735 [p < 0.001], and OR 0.823 [p = 0.001], respectively). In the crude, adjust I, and adjust II models, the three variables stated above were all considered to be significant predictors of infection after DBS surgery. The calibration curves in both training and validation cohorts showed that the predicted outcome fitted well to the observed outcome (p > 0.05). The decision curves showed that the nomogram had more benefits than the "All or None" scheme. The areas under the curve were 0.93 and 0.83 in the training and validation cohorts, respectively. CONCLUSIONS: The nomogram included BMI, blood glucose, and albumin, which were significant predictors of infection in patients with DBS surgery. The nomogram was reliable for clinical application.

Predictive Nomogram for Unfavorable Outcome of Spontaneous Intracerebral Hemorrhage.

BACKGROUND: The goal of this retrospective study was to evaluate the effect of surgical timing on patient outcomes after spontaneous intracerebral hemorrhage (ICH). We also identified risk factors associated with poor prognosis. METHODS: We reviewed all patients who underwent surgery for ICH between January 2014 and January 2021. The outcome was measured using the modified Rankin Scale (mRS) score at 6 months after the surgery. Patients with mRS 0-2 were considered having favorable outcomes, and those with mRS 3-5 were considered having unfavorable outcomes. The relationships of surgical timing with the risk of unfavorable outcomes were identified using the interaction and stratified analyses, and generalized additive and logistic regression models. A nomogram was established and evaluated using a receiver operating characteristic curve analysis, plotted decision curve, and calibration curve. RESULTS: We identified 53 patients with favorable outcomes and 144 with unfavorable outcomes. The number of cases who underwent surgery at >12 hours and <36 hours in the favorable outcome group was more than that in the unfavorable outcome group (P < 0.001). When the time to operating room (TOR) was less than 21 hours, a shorter TOR was associated with unfavorable outcomes, using the smoothing spline analysis (odds ratio = 0.8, P < 0.001). Finally, we developed a nomogram using systolic blood pressure, Glasgow Coma Scale, midline shift, hematoma volume, and TOR for predicting the unfavorable outcome. The area under the curve, accuracy, specificity, and sensitivity of nomogram were 0.90, 0.87, 0.72, and 0.93, respectively. CONCLUSION: Surgical timing between 12 and 26 hours after ICH was associated with favorable outcomes. The nomogram including systolic blood pressure, Glasgow Coma Scale, midline shift, hematoma volume, and TOR was reliable for predicting the ICH outcome.

Dihydrotanshinone I inhibits human glioma cell proliferation via the activation of ferroptosis.

The aim of the present study was to investigate the effect of dihydrotanshinone I (DHI) on the survival of human glioma cells and the expression levels of ferroptosis-associated proteins. Human U251 and U87 glioma cells were cultured in vitro and treated with different concentrations of DHI and/or the ferroptosis inhibitor ferrostatin-1. A Cell Counting Kit-8 assay was used to determine the cell survival rate. The cells were further analyzed to determine their 5-, 12- and 15-hydroxyeicosatetraenoic acid (HETE), lactate dehydrogenase (LDH) and malondialdehyde (MDA) levels, and reduced glutathione (GSH)/oxidized glutathione (GSSG) ratios. Western blotting was used to detect ferroptosis-associated glutathione peroxidase 4 (GPX4) and long-chain acyl-CoA synthetase 4 (ACSL-4). Changes in the mitochondrial membrane potential (MMP) were also observed using tetramethylrhodamine methyl ester staining and confocal fluorescence microscopy. The results revealed that DHI inhibited the proliferation of human glioma cells. Following treatment of the U251 and U87 cells with DHI, changes in the expression levels of ferroptosis-associated proteins were observed; the expression level of GPX4 decreased and that of ACSL-4 increased. DHI also increased the levels of LDH and MDA in the human glioma cells and reduced the GSH/GSSG ratio. The DHI-treated cells also exhibited a marked reduction in MMP. Furthermore, ferrostatin-1 blocked the DHI-induced effects in human glioma cells. From these results, it may be concluded that DHI inhibits the proliferation of human glioma cells via the induction of ferroptosis.

Overexpression of HOXC10 promotes glioblastoma cell progression to a poor prognosis via the PI3K/AKT signalling pathway.

OBJECTIVE: The HOX gene is expressed in neoplasias occurred in multiple tissues, such as the colon, lung and breast. However, the effects of the HOX gene on glioblastoma (GBM) remain poorly understood. We examined HOXC10 expression in GBM tissues and cells, analysed its effect on GBM prognosis, and finally assessed its possible underlying mechanisms in this study. METHODS: HOXC10 expression levels and its prognostic effects on GBM tissues were analysed based on The Cancer Genome Atlas (TCGA) and ONCOMINE database. Overall survival (OS) analysis was performed using the Kaplan-Meier method and log rank test. Then, the expression of HOXC10 was detected in four GBM cell lines using quantitative real-time reverse transcription-PCR (qRT-PCR). In addition, small interfering RNA (si-RNA) was utilised in the U87 cell line with the highest HOXC10 expression to facilitate subsequent in vitro cell experiment. Cell proliferation, migration and invasion were assessed using the Cell Counting Kit-8 (CCK-8) and colony formation assay, wound healing, Transwell assay, respectively in GBM U87 cell after HOXC10 knockdown. Key proteins related to the phosphatidylinositol-3-kinase (PI3K)/protein kinase B (AKT) signalling pathway were measured by western blotting. RESULTS: HOXC10 expression was significantly increased in GBM tissues and cell lines, leading a poor OS in GBM patients. Knockdown of HOXC10 could inhibit the GBM U87 cells proliferation, migration and invasion, as well as decreased expression levels of key proteins in PI3K/AKT signalling pathway. CONCLUSION: HOXC10 was overexpressed in GBM tissues and cells, and associated with poor prognosis in GBM patients. Moreover, HOXC10 knockdown inhibited U87 cell proliferation, migration and invasion, which were potentially related to PI3K/AKT signalling pathway activation. Our findings revealed that HOXC10 represent a promising biological target for GBM treatment in the future.