A machine learning-based prediction model of H3K27M mutations in brainstem gliomas using conventional MRI and clinical features.

BACKGROUND: H3K27M is the most frequent mutation in brainstem gliomas (BSGs), and it has great significance in the differential diagnosis, prognostic prediction and treatment strategy selection of BSGs. There has been a lack of reliable noninvasive methods capable of accurately predicting H3K27M mutations in BSGs. METHODS: A total of 151 patients with newly diagnosed BSGs were included in this retrospective study. The H3K27M mutation status was obtained by whole-exome, whole-genome or Sanger's sequencing. A total of 1697 features, including 6 clinical parameters and 1691 imaging features, were extracted from pre- and post-contrast T1-weighted and T2-weighted images. Using a random forest algorithm, 36 selected MR image features were integrated with 3 selected clinical features to generate a model that was predictive of H3K27M mutations. Additionally, a simplified prediction model comprising the Karnofsky Performance Status (KPS) at diagnosis, symptom duration at diagnosis and edge sharpness on T2 was established for practical clinical utility using the least squares estimation method. RESULTS: H3K27M mutation was an independent prognostic factor that conferred a worse prognosis (p = 0.01, hazard ratio = 3.0, 95% confidence interval [CI], 1.57-5.74). The machine learning-based model achieved an accuracy of 84.44% (area under the curve [AUC] = 0.8298) in the test cohort. The simplified model achieved an AUC of 0.7839 in the test cohort. CONCLUSIONS: Using conventional MRI and clinical features, we established a machine learning-based model with high accuracy and a simplified model with improved clinical utility to predict H3K27M mutations in BSGs.

Fluorescein-Guided Surgery for Pediatric Brainstem Gliomas: Preliminary Study and Technical Notes.

Introduction Brainstem gliomas (BsG) account for 10 to 15% of pediatric brain tumors. Surgery is the preferred treatment for focal and exophytic lesions. Sodium fluorescein has been proven safe and effective in resection of malignant brain tumors. Objective The objective was to o analyze the safety and effectiveness of this approach, to evaluate intraoperative fluorescein imaging, and to measure the safety of chosen dose for pediatric patients. Methods Twelve cases were enrolled between March 2014 and September 2016 in Beijing Tiantan Hospital. All of the patients received 2.5 mg/kg of sodium fluorescein before opening the dura; the intraoperative fluorescence enhancement was observed, and the degree of satisfaction and consistency with the neuronavigation were evaluated. Results With a mean age of 7.5 years, there were eight cases located within the pontine, three in the medullary oblongata, and one in the tectal plate. Histological results were astrocytoma, glioblastoma, oligodendroglioma, and pilocytic astrocytoma. Under the fluorescein module of the microscope, the tumors were recognizable enough to help surgeons to discriminate the lesion from non-fluorescent tissue, with a consistency of 83% with the neuronavigation. Total removal was accomplished in nine cases, while the mean percentage of resection of the other cases was 93.7%. The Karnofsky performance score (KPS) showed no significant differences between pre-operation and discharge, but there was a difference between pre-operation and 6-month follow-up. Conclusion The fluorescein-guided surgery is useful for demarcating the tumor margin and works well with other navigation and monitoring devices. A safe dose of sodium fluorescein (2.5 mg/kg) was proven effective for children.

Prognostic indicators of adult medullary gliomas after microsurgical treatment - A retrospective analysis of 54 patients.

Due to the low incidence of medullary gliomas, the special location, and the function of the gliomas in the medulla oblongata, microsurgical treatment is still challenging for neurosurgeons. The aim of this study was to observe the effect of microsurgical treatment of adult medullary gliomas and to explore the prognostic factors after treatment. The clinical data from 54 patients with adult medullary gliomas who received microsurgical treatment at Beijing Tiantan Hospital (China) from April 2008 to April 2014 was retrospectively analyzed. The factors affecting their prognosis were analyzed with log-rank univariate analysis. The factors that affected prognosis included age, gender, duration of preoperative symptoms, Karnofsky Performance Scale (KPS) score, World Health Organization (WHO) grade, extent of tumor resection, and postoperative complications. Those with statistical significance in the univariate analysis were entered into a multivariate Cox regression analysis. WHO grading showed 7 cases of grade I, 30 cases of grade II, 14 cases of grade III, and 3 cases of grade IV tumors. Univariable analysis showed that postoperative nasogastric feeding (P=0.031), WHO pathological grade (P=0.018), extent of resection (P=0.016), and preoperative involvement of ≥3 cranial nerves (CNs) (P=0.014) affected overall survival. The WHO pathological grade of the tumor was an independent risk factor for prognosis. In conclusion, the WHO pathological grade of the tumor was an important prognostic indicator.

Brainstem gangliogliomas: prognostic factors, surgical indications and functional outcomes.

To explore the prognostic factors and discuss the surgical indications of brainstem gangliogliomas. Twenty-one patients with brainstem ganglioglioma were surgically treated at our hospital between 2006 and 2014. The clinical, radiological, operative, and pathological findings of these patients were retrospectively reviewed. The 3-years overall survival and event-free survival (EFS) rates were 90.5 % and 68.4 %, respectively. Four patients (4/18, 22 %) experienced a recurrence with a mean recurrence-free survival of 5.5 months and a mean follow-up of 37 months. Three patients died of surgery-related complications. Three growth patterns were identified: exophytic (6/21), intrinsic (2/21), and endo-exophytic (13/21). Eight patients (8/15, 53 %) harbored a BRAF V600E mutation. All recurrent tumors were endo-exophytic, and except the one without molecular information, were BRAF V600E mutants. A Cox hazard proportion ratio model was used to identify factors influencing EFS, including sex, age, location, growth patterns, extent of resection (EOR), and BRAF V600E mutation status. On univariate analysis, none of these factors reached statistical significance. Among them, EOR and growth patterns were strongly associated with each other (Fisher's exact test, P < 0.01). A multivariate analysis demonstrated that growth patterns were the only factor associated with EFS (P = 0.02; HR 49.05; 95 % CI 1.76-1365.13). Growth patterns may be useful to select surgery candidates and predict prognosis for patients with brainstem gangliogliomas. BRAF V600E was frequently present and appeared to be associated with shorter recurrence-free survival. Studies on BRAF V600E-targeted therapy for patients with high surgical risks are needed.