Prognostic Analysis of Clinical and Immunohistochemical Factors for Patients with Spinal Schwannoma

Oncologic Applications of Artificial Intelligence and Deep Learning Methods in CT Spine Imaging-A Systematic Review

In spinal oncology, integrating deep learning with computed tomography (CT) imaging has shown promise in enhancing diagnostic accuracy, treatment planning, and patient outcomes. This systematic review synthesizes evidence on artificial intelligence (AI) applications in CT imaging for spinal tumors. A PRISMA-guided search identified 33 studies: 12 (36.4%) focused on detecting spinal malignancies, 11 (33.3%) on classification, 6 (18.2%) on prognostication, 3 (9.1%) on treatment planning, and 1 (3.0%) on both detection and classification. Of the classification studies, 7 (21.2%) used machine learning to distinguish between benign and malignant lesions, 3 (9.1%) evaluated tumor stage or grade, and 2 (6.1%) employed radiomics for biomarker classification. Prognostic studies included three (9.1%) that predicted complications such as pathological fractures and three (9.1%) that predicted treatment outcomes. AI's potential for improving workflow efficiency, aiding decision-making, and reducing complications is discussed, along with its limitations in generalizability, interpretability, and clinical integration. Future directions for AI in spinal oncology are also explored. In conclusion, while AI technologies in CT imaging are promising, further research is necessary to validate their clinical effectiveness and optimize their integration into routine practice.

Long-term postoperative outcomes of spinal cellular schwannoma: study of 93 consecutive cases

BACKGROUND CONTEXT

Cellular schwannoma (CS) is a rare tumor that accounts for 2.8%-5.2% of all benign schwannomas. There is a dearth of up-to-date information on spinal CS in the literature.

PURPOSE

The aims of this study were to identify the proportion of CS cases amongst spinal benign schwannoma, describe the clinical features of spinal CS, and identify prognostic factors for local recurrence by analyzing data from 93 consecutive CS cases.

STUDY DESIGN

Retrospective review.

PATIENT SAMPLE

We analyzed 93 PSGCT screened from 1,706 patients with spine CS who were treated at our institute between 2008 and 2021.

OUTCOME MEASURES

Demographic, radiographic, operative and postoperative data were recorded and analyzed.

METHODS

We compared the clinical features of spinal CS from the cervical, thoracic, lumbar and sacral segments. Prognostic factors for local recurrence-free survival (RFS) were identified by the Kaplan-Meier method. Factors with p≤.05 in univariate analysis were subjected to multivariate analysis by Cox regression analysis.

RESULTS

The proportion of spinal CS in all benign schwannomas was 6.7%. The mean and median follow-up times for the 93 patients in this study were 92.2 and 91.0 months respectively (range 36-182 months). Local recurrence was detected in 11 cases, giving an overall recurrence rate of 11.7%, with one patient death. Statistical analysis revealed that tumor size ≥5 cm, intralesional resection, and Ki-67 ≥5% were independent negative prognostic factors for RFS in spinal CS.

CONCLUSIONS

Whenever possible, en bloc resection is recommended for spinal CS. Long-term follow-up should be carried out for patients with tumor size ≥5 cm and postoperative pathological Ki-67 ≥5%.

Long-Term Functional Outcomes Following Surgical Treatment of Spinal Schwannomas: A Population-Based Cohort Study

Spinal schwannomas are the second most common form of primary intradural spinal tumor. Despite being benign, they may cause spinal cord compression and subsequently acute or chronic neurological dysfunction. The primary treatment is surgical resection. The aim of this study was to identify pre- and postoperative predictors of favorable outcomes after surgical treatment for spinal schwannoma. All adult patients surgically treated for spinal schwannoma between 2006 and 2020 were eligible for inclusion. Medical records and imaging data were retrospectively reviewed. The primary outcome measures were neurological improvement according to the modified McCormick Scale (mMC) and changes in motor deficit, sensory deficit, gait disturbance, bladder dysfunction, and pain at long-term follow-up. In total, 180 patients with a median follow-up time of 4.4 years were included. Pain was the most common presenting symptom (87%). The median time between symptom presentation and surgery was 12 months, while the median time between diagnosis (first MRI) and surgery was 3 months. Gross total resection (GTR) was achieved in 150 (83%) patients and the nerve root could be preserved in 133 (74%) patients. A postoperative complication occurred in 10 patients (5.6%). There were significant postoperative improvements in terms of motor, sensory, gait, and bladder functions, as well as pain (p < 0.001). Of these symptoms, bladder dysfunction was the one most often improved, with complete symptom resolution in all cases. However, no other predictors of improvement could be identified. There were three cases of recurrence after GTR and nine cases of regrowth after STR. Reoperation was performed in six (3.3%) cases. GTR was associated with a significant improvement in neurological status at long-term follow-up and increased the chance of progression-free survival.

Scoring model to predict postoperative neurological deterioration in spinal schwannoma

Background

Spinal schwannomas (SSs) are benign tumors affecting the nerve sheath, accounting for 25% of spinal nerve root tumors. Surgery represents the mainstay of treatment for SS patients. Following surgery, approximately 30% of patients experienced developed new or worsening neurological deterioration, which probably represented an inevitable complication of nerve sheath tumor surgery. The objective of this study was to identify the rates of new or worsening neurological deterioration in our center and accurately predict the neurological outcomes of patients with SS by developing a new scoring model.

Methods

A total of 203 patients were retrospectively enrolled at our center. Risk factors associated with postoperative neurological deterioration were identified by multivariate logistic regression analysis. β–coefficients for independent risk factors were used to define a numerical score to generate a scoring model. The validation cohort at our center was used to verify the accuracy and reliability of the scoring model. Receiver operating characteristic (ROC) curve analysis was used to evaluate the performance of the scoring model.

Results

In this study, five measured variables were selected for the scoring model: duration of preoperative symptoms (1 point), radiating pain (2 points), tumor size (2 points), tumor site (1 point), and dumbbell tumor (1 point). The scoring model divided the spinal schwannoma patients into three categories: low risk (0-2 points), intermediate risk (3-5 points), and high risk (6-7 points), with predicted risks of neurological deterioration of 8.7%, 36%, and 87.5%, respectively. And the validation cohort confirmed the model with the predicted risks of 8.6%, 46.4%, and 66.6%, respectively.

Conclusion

The new scoring model might intuitively and individually predict the risk of neurological deterioration and may aid individualized treatment decision-making for SS patients.

Differentiation of benign and malignant spinal schwannoma using guided attention inference networks on multi-source MRI: comparison with radiomics method and radiologist-based clinical assessment

BACKGROUND

Differentiating diagnosis between the benign schwannoma and the malignant counterparts merely by neuroimaging is not always clear and remains still confounding in many cases because of atypical imaging presentation encountered in clinic and the lack of specific diagnostic markers.

PURPOSE

To construct and validate a novel deep learning model based on multi-source magnetic resonance imaging (MRI) in automatically differentiating malignant spinal schwannoma from benign.

MATERIAL AND METHODS

We retrospectively reviewed MRI imaging data from 119 patients with the initial diagnosis of benign or malignant spinal schwannoma confirmed by postoperative pathology. A novel convolutional neural network (CNN)-based deep learning model named GAIN-CP (Guided Attention Inference Network with Clinical Priors) was constructed. An ablation study for the fivefold cross-validation and cross-source experiments were conducted to validate the novel model. The diagnosis performance among our GAIN-CP model, the conventional radiomics model, and the radiologist-based clinical assessment were compared using the area under the receiver operating characteristic curve (AUC) and balanced accuracy (BAC).

RESULTS

The AUC score of the proposed GAIN method is 0.83, which outperforms the radiomics method (0.65) and the evaluations from the radiologists (0.67). By incorporating both the image data and the clinical prior features, our GAIN-CP achieves an AUC score of 0.95. The GAIN-CP also achieves the best performance on fivefold cross-validation and cross-source experiments.

CONCLUSION

The novel GAIN-CP method can successfully classify malignant spinal schwannoma from benign cases using the provided multi-source MR images exhibiting good prospect in clinical diagnosis.

Prognostic significance of preoperative neutrophil-to-lymphocyte ratio in surgically resected schwannomas

Objective

The goal of schwannoma resection is to control the tumor while preserving neurological function. Schwannomas have a variable postoperative growth pattern, therefore preoperative prediction of a schwannoma's growth pattern is favorable. This study aimed to examine the relationship between preoperative neutrophil-to-lymphocyte ratio (NLR) and postoperative recurrence and retreatment in patients with schwannoma.

Methods

We retrospectively examined 124 patients who underwent schwannoma resection in our institution. Associations between preoperative NLR, other patient and tumor characteristics, and tumor recurrence and retreatment were analyzed.

Results

Median follow-up was 2569.5 days. Postoperative recurrence occurred in 37 patients. Recurrence that required retreatment occurred in 22. Treatment-free survival (TFS) was significantly shorter in patients with NLR ≥2.21 (P = 0.0010). Multivariate Cox proportional hazards regression showed that NLR and neurofibromatosis type 2 were independent predictors of retreatment (P = 0.0423 and 0.0043, respectively). TFS was significantly shorter in patients with NLR ≥2.21 in the following subgroups: sporadic schwannoma, primary schwannoma, schwannoma ≥30 mm in size, subtotal resection, vestibular schwannoma, and postoperative recurrence.

Conclusions

Preoperative NLR ≥2.21 before surgery was significantly associated with retreatment after schwannoma resection. NLR may be a novel predictor of retreatment and assist surgeons in preoperative surgical decision making.

Successful management of a rare radius schwannoma mimicking malignant bone tumors: A case report and literature review

Background

Schwannomas are benign tumors originating from Schwann cells, frequently occurring in the spinal cord and peripheral nerves. Intraosseous schwannomas, a rare subset, account for approximately 0.2% of schwannomas. Intraosseous schwannomas commonly impinge the mandible, followed by the sacrum and the spine. By far, only three cases of radius intraosseous schwannomas have been reported in PubMed. The tumor was treated differently in all three cases, resulting in different outcomes.

Case presentation

A 29-year-old male construction engineer who complained of a painless mass on the radial aspect of the right forearm was diagnosed with an intraosseous schwannoma of the radius based on radiography, three-dimensional computed tomography reconstruction, magnetic resonance imaging, pathological examination, and immunohistochemistry. A different surgical approach was employed to reconstruct the radial graft defect using bone microrepair techniques, resulting in more reliable bone healing and early functional recovery. Meanwhile, no clinical and radiographic findings suggestive of recurrence were observed at the 12-month follow-up.

Conclusion

Vascularized bone flap transplantation combined with three-dimensional imaging reconstruction planning might yield better results for repairing small segmental bone defects of the radius caused by intraosseous schwannomas.

Risk factors for recurrence and regrowth of spinal schwannoma

BACKGROUND

Spinal schwannoma recurs after initial surgery at a rate of 4%-6%, with known risk factors including subtotal resection, multilevel involvement, large tumor size, and malignant histopathology. This study examined risk factors for schwannoma recurrence and residual tumor regrowth.

METHODS

Sixty-five patients who underwent resection of spinal schwannoma in our department between July 2010 and December 2018 and were followed up for more than 1 year were retrospectively analyzed for age, sex, follow-up duration, imaging and surgical data, recurrence, reoperation, and Japanese Orthopaedic Association scores before and 1 year after surgery. Patients with postoperative recurrence or residual tumor regrowth of >10% at the final visit (R+ group) were compared with patients without recurrence or regrowth (R- group). Multivariate logistic regression analysis was performed to analyze concurrent effects of risk factors on recurrence and regrowth.

RESULTS

The 65 patients (mean age 52.4 years at surgery) had schwannomas involving cervical (n = 14), thoracic (n = 25), and lumbar (n = 26) spinal levels. Mean follow-up duration was 58 months. Location was intradural in 65%, extradural in 17%, and both intradural and extradural in 18%. There were 4 recurrences (6.2%), and the mean interval between surgery and recurrence was 18.8 months. Seven patients (10.8%) experienced regrowth. Comparing group R+ (n = 11) and group R- (n = 54), univariate analysis showed significant differences in Sridhar tumor classification, giant tumor (Sridhar classification II, IVb, and V), left-right and cranial-caudal tumor size, largest diameter, operative time, blood loss, subtotal resection, reoperation, fusion surgery, and follow-up duration. Multivariate logistic regression analysis revealed giant tumor (Sridhar classification types II, IVb, and V) as an independent risk factor for recurrence and regrowth.

CONCLUSIONS

This retrospective review of 65 consecutive patients with spinal schwannoma in a single institution demonstrated that 16.9% had recurrence or regrowth, demonstrating that this potential risk should be kept in mind.

Do Tumour Size, Type and Localisation Affect Resection Rate in Patients with Spinal Schwannoma?

Background and Objectives: Spinal schwannomas are benign tumours that can present with various symptoms such as pain, radiculopathy and neurological deficit. Gross total resection (GTR) is of key importance for local recurrence. The aim of this study is to describe the clinical characteristics, resection rate, clinical outcome, as well as tumour recurrence, in patients with non-syndromic spinal schwannomas and to clarify which factors affect the resection rate. Materials and Methods: Patients with non-syndromic spinal schwannomas that underwent surgical resection between January 2009 and December 2018 at a single institution were included. Demographic parameters, clinical symptoms, tumour localisation and size, surgical approach and complications were noted. Factors influencing the extent of resection, the surgeon’s decision regarding the approach and the occurrence of new postoperative deficits were evaluated. Results: Fifty patients (18 females) were included. The most common presenting symptom was radiculopathy (88%). The lumbar spine was the most commonly affected site (58%). Laminotomy (72%) was the preferred surgical approach overall and specifically for exclusively intraspinal schwannomas (p = 0.02). GTR was achieved in 76.0% (n = 38). In multivariate analysis, only tumour localisation within the spinal canal (p = 0.014) independently predicted GTR, whereas the type of approach (p = 0.50) and tumour volume (p = 0.072) did not. New postoperative persisting deficits could not be predicted by any factor, including the use and alteration of intraoperative neuromonitoring. Recurrence was observed in four cases (8%) and was significantly higher in cases with STR (p = 0.04). Conclusions: In this retrospective study, GTR was solely predicted by tumour localisation within the spinal canal. The decision regarding the utilisation of different surgical approaches was solely influenced by the same factor. No factor could predict new persisting deficits. Tumour recurrence was higher in STR.

Comparison of pathological, radiological, and prognostic features between cellular schwannoma and non-cellular schwannoma

PURPOSE

To investigate the differences of pathological, radiological, and prognostic features between cellular schwannoma (CS) and non-cellular schwannoma (NCS).

METHODS

CT and MRI images of 24 patients with CSs and 30 patients with NCSs were reviewed retrospectively. Clinico-pathological characteristics of CSs and NCSs and tumor radiological features including location, shape, size, border, cystic-solid components, hemorrhage, calcification, bone remodeling, pattern of CT/MRI precontrast scan, degree of enhancement, target sign, and tumor vessels were recorded. Statistical analyses were performed with Chi square or Fisher's exact test, independent sample t test, and logistic regression analysis to compare the differences between CSs and NCSs.

RESULTS

Four CSs showed mitotic activity, which was not found in the NCS group (P = 0.034). The CS group showed higher MIB-1 index than that in the NCS group (P = 0.002). Two patients with CS presented with tumor recurrence. Compared to NCSs, CSs were often located in spinal area (P = 0.028) and irregular (P = 0.013) with larger size (P = 0.005). Target sign, a common finding in NCSs (7/22, 31.8 %), was not seen in CSs (P = 0.014). The tumor vessels were only seen in CS group (4/22, 18.2 %; P = 0.027). Regression analysis revealed that location (P = 0.048) and size (P = 0.012) were independent indicators in differentiating CSs from NCSs.

CONCLUSIONS

CS is a rare subtype of schwannoma with some significant radiological features including a predilection for the spinal area, irregular shape, large tumor size, absent target sign, tumor vessels, and potential risk of recurrence. Location and size of the schwannomas were the most useful indicators in differentiating CSs from NCSs.