Novel prognostication of patients with spinal and pelvic chondrosarcoma using deep survival neural networks.

BACKGROUND: We used the Surveillance, Epidemiology, and End Results (SEER) database to develop and validate deep survival neural network machine learning (ML) algorithms to predict survival following a spino-pelvic chondrosarcoma diagnosis. METHODS: The SEER 18 registries were used to apply the Risk Estimate Distance Survival Neural Network (RED_SNN) in the model. Our model was evaluated at each time window with receiver operating characteristic curves and areas under the curves (AUCs), as was the concordance index (c-index). RESULTS: The subjects (n = 1088) were separated into training (80%, n = 870) and test sets (20%, n = 218). The training data were randomly sorted into training and validation sets using 5-fold cross validation. The median c-index of the five validation sets was 0.84 (95% confidence interval 0.79-0.87). The median AUC of the five validation subsets was 0.84. This model was evaluated with the previously separated test set. The c-index was 0.82 and the mean AUC of the 30 different time windows was 0.85 (standard deviation 0.02). According to the estimated survival probability (by 62 months), we divided the test group into five subgroups. The survival curves of the subgroups showed statistically significant separation (p < 0.001). CONCLUSIONS: This study is the first to analyze population-level data using artificial neural network ML algorithms for the role and outcomes of surgical resection and radiation therapy in spino-pelvic chondrosarcoma.

Subtotal resection of cervical schwannomas and growth rate of residual tumors.

OBJECTIVEThe objective of this study was to elucidate the features and surgical outcomes of cervical schwannomas.METHODSThe authors retrospectively reviewed the records of 90 patients who underwent surgically treated cervical schwannomas from January 1995 to December 2017, with an emphasis on MRI findings such as tumor location, tumor size, extent of tumor resection, and growth of a residual tumor.RESULTSThis study included 51 men (56.7%) and 39 women (43.3%) with a mean age of 44.5 years (range 7-77 years). Dumbbell-shaped tumors comprised 62 (68.9%) of 90 cases and gross-total resection (GTR) was achieved in 59 (65.6%) of 90 cases. All nondumbbell tumors (n = 28) underwent GTR. Only 1 case of recurrence in the GTR group showed a gradual increase in size (by 8.9 mm) during the 150-month follow-up period. For the regrowth patients in the subtotal resection group, the mean percentage increase in tumor size was 47.5% ± 33.1% and the mean growth rate was 5.8 ± 4.6 mm/year during the 20.3-month follow-up period. However, the size of residual tumor spontaneously decreased by a mean of 8.3% ± 11.1% during the 48.4-month follow-up period in the nonregrowth group.CONCLUSIONSThese findings suggested that frequent MRI follow-up examinations are required for residual schwannomas in the cervical spine for at least 2 years, and continuous MRI follow-ups are also required thereafter.

Subtotal Resection of Cervical Dumbbell Schwannomas: Radiographic Predictors for Surgical Considerations.

OBJECTIVE: Currently, radiologic predictors for the resectability of cervical dumbbell schwannomas remain unknown. To identify radiologic predictors for resectability, we retrospectively reviewed data from 72 patients. METHODS: From January 1995 to June 2017, 72 patients who underwent surgical treatment for cervical dumbbell schwannomas were enrolled. We focused on the relationship between preoperative magnetic resonance imaging (MRI) features and the extent of tumor removal. The MRI features evaluated were tumor size, tumor level, Eden classification, degree of vertebral artery (VA) involvement, and signal intensity (SI) on T2-weighted images (WIs). RESULTS: Among the 72 patients, gross total resection (GTR) and subtotal resection (STR) were achieved in 37 (51.4%) and 35 (48.6%) patients, respectively. Mean maximal tumor size (P = 0.011), mean size of foraminal and extraforaminal portion (P = 0.017), tumor level (P < 0.001), VA involvement (P < 0.001), and SI on T2-WIs (P = 0.006) were significantly different between the GTR and STR groups. Univariate analyses demonstrated that maximal tumor size (odds ratio [OR]: 0.93, P = 0.012), high cervical level (OR: 11.37, P < 0.001), pushed VA (OR: 0.11, P = 0.002), encased VA (OR: 0.02, P < 0.001), and hyper-SI on T2-WIs (OR: 12.46, P = 0.020) were significant predictors for GTR. In the multivariate analysis, only high cervical level (OR: 5.48, P = 0.033) and encased VA (OR: 0.07, P = 0.014) were significant predictors for GTR. CONCLUSIONS: The resectability of cervical dumbbell schwannomas may be predicted by MRI features, including tumor size, tumor level, and degree of VA involvement.

Risk of Recurrent Chronic Subdural Hematoma Associated with Early Warfarin Resumption: A Matched Cohort Study.

OBJECTIVE: Studies on resuming anticoagulation after burr-hole drainage for chronic subdural hematoma (CSDH) are limited. To evaluate the safety for early warfarin resumption after burr-hole drainage, we conducted a retrospective matched cohort study. METHODS: Between January 2008 and April 2015, 36 patients with warfarin-related unilateral CSDH and 151 patients with ordinary unilateral CSDH were enrolled in this study. Patients taking warfarin were managed homogeneously according to the study protocol, and the usual dosage of warfarin was resumed within 2 or 3 days of burr-hole drainage to reach a target international normalized ratio (INR) of 2.1. The primary outcome, defined as recurrent CSDH requiring repeated burr-hole drainage within 3 months of the initial surgery, was compared between the two groups. RESULTS: The primary outcome was observed in 4 (11%) of the 36 patients taking warfarin and in 18 (12%) of the 151 ordinary patients. After propensity score matching, the primary outcome was observed in 3 of 33 patients (9%) in the matched warfarin cohort and 11 of 74 patients (15%) in the matched ordinary cohort. When the results were analyzed using the generalized estimating equation, no significant difference was observed in the rate of recurrent CSDH between the 2 groups (P = 0.411). In addition, we found that recurrent CSDH was not related to postoperative international normalized ratio levels (P = 0.332). CONCLUSIONS: There was no definitive association between postoperative early warfarin resumption and the recurrence rate of CSDH. Patients with warfarin-related CSDH and a strong indication for anticoagulation can be managed by resuming warfarin within 3 days of burr-hole drainage.

Predicting Survival of Patients with Spinal Ependymoma Using Machine Learning Algorithms with the SEER Database.

OBJECTIVE: This study was conducted to understand the clinical and demographic factors influencing the overall survival (OS) of patients with spinal ependymoma and to predict the OS with machine learning (ML) algorithms. METHODS: We compiled spinal ependymoma cases diagnosed between 1973 and 2014 from the Surveillance, Epidemiology, and End Results (SEER) registry. To identify the factors influencing survival, statistical analyses were performed using the Kaplan-Meier method and Cox proportional hazards regression model. In addition, we implemented ML algorithms to predict the OS of patients with spinal ependymoma. RESULTS: In the multivariate analysis model, age ≥65 years, histologic subtype, extraneural metastasis, multiple lesions, surgery, radiation therapy, and gross total resection (GTR) were found to be independent predictors for OS. Our ML model achieved an area under the receiver operating characteristic curve (AUC) of 0.74 (95% confidence interval [CI], 0.72-0.75) for predicting a 5-year OS of spinal ependymoma and an AUC of 0.81 (95% CI, 0.80-0.83) for predicting a 10-year OS. The stepwise logistic regression model showed poorer performance by an AUC of 0.71 (95% CI, 0.70-0.72) for predicting a 5-year OS and an AUC of 0.75 (95% CI, 0.73-0.77) for predicting a 10-year OS. CONCLUSIONS: With SEER data, we reaffirmed that therapeutic factors, such as surgery and GTR, were associated with improved OS. Compared with statistical methods, ML techniques showed satisfactory results in predicting OS; however, the dataset was heterogeneous and complex with numerous missing values.

Optimal treatment of jugular foramen schwannomas: long-term outcome of a multidisciplinary approach for a series of 29 cases in a single institute.

BACKGROUND: The goal of treatment for jugular foramen schwannomas (JFSs) is to achieve complete tumor removal with cranial nerve preservation. However, achieving this goal remains a challenge despite the advances in microsurgical techniques. The aim of this study was to determine optimal treatment strategies for JFSs based on a review of a series of 29 surgical cases in our institute. MATERIALS AND METHODS: Between 1997 and 2013, 29 patients with JFSs underwent surgical treatment by multidisciplinary otoneurosurgical approaches. We retrospectively evaluated various clinical outcomes including the extent of tumor resection, postoperative cranial nerve deficits, and the recurrence rate. Tumor extension was classified using the Kaye and Pellet classification (KPC) system, and the extent of tumor resection was graded as gross total resection (GTR), near total resection (NTR), and subtotal resection (STR). We utilized the House-Brackmann facial nerve grading system (HBFNGS), the average pure-tone audiometry and speech audiometry (PTA/SA) tests, and the American Speech-Language-Hearing Association National Outcome Measurement System (ASHA NOMS) swallowing scale (ASHA level) for assessment of functional outcomes. RESULTS: The extent of tumor resection was not related to the degree of immediate postoperative cranial nerve deficits. However, the surgical approach was significantly related to postoperative hearing status and immediate postoperative facial function. Also, among the ten patients who were below the level of acceptable facial function immediately postoperatively, nine patients (90%) recovered to acceptable facial function by the last follow-up. Concerning postoperative swallowing status, all 21 patients recovered swallowing function by the last follow-up. Postoperative Gamma Knife stereotactic radiosurgery (GKRS) was performed for three recurrent and seven residual tumors, and recurrence was not observed in the mean 36-month follow-up period. CONCLUSIONS: A surgical strategy should be tailored to the individual case, and clinicians should consider the possibility of recurrence and further adjuvant treatment.

Superficial Siderosis of the Central Nervous System Originating from the Thoracic Spine: A Case Report.

Superficial siderosis of the central nervous system(SSCNS) is a rare disease characterized by hemosiderin deposition on the surface of the central nervous system. We report a case of SSCNS originating from the thoracic spine, presenting with neurological deficits including, sensorineuronal hearing loss, ataxia, and corticospinal and dorsal column tract signs. The patient underwent dural repair with an artificial dural patch. Clinical findings were elicited by neurological examination, imaging studies, and intraoperative findings, and these were addressed through literature review.