Surgery for primary filum terminale ependymomas: outcome and prognostic factors

Convolutional neural network-based magnetic resonance image differentiation of filum terminale ependymomas from schwannomas

PURPOSE

Preoperative diagnosis of filum terminale ependymomas (FTEs) versus schwannomas is difficult but essential for surgical planning and prognostic assessment. With the advancement of deep-learning approaches based on convolutional neural networks (CNNs), the aim of this study was to determine whether CNN-based interpretation of magnetic resonance (MR) images of these two tumours could be achieved.

METHODS

Contrast-enhanced MRI data from 50 patients with primary FTE and 50 schwannomas in the lumbosacral spinal canal were retrospectively collected and used as training and internal validation datasets. The diagnostic accuracy of MRI was determined by consistency with postoperative histopathological examination. T1-weighted (T1-WI), T2-weighted (T2-WI) and contrast-enhanced T1-weighted (CE-T1) MR images of the sagittal plane containing the tumour mass were selected for analysis. For each sequence, patient MRI data were randomly allocated to 5 groups that further underwent fivefold cross-validation to evaluate the diagnostic efficacy of the CNN models. An additional 34 pairs of cases were used as an external test dataset to validate the CNN classifiers.

RESULTS

After comparing multiple backbone CNN models, we developed a diagnostic system using Inception-v3. In the external test dataset, the per-examination combined sensitivities were 0.78 (0.71-0.84, 95% CI) based on T1-weighted images, 0.79 (0.72-0.84, 95% CI) for T2-weighted images, 0.88 (0.83-0.92, 95% CI) for CE-T1 images, and 0.88 (0.83-0.92, 95% CI) for all weighted images. The combined specificities were 0.72 based on T1-WI (0.66-0.78, 95% CI), 0.84 (0.78-0.89, 95% CI) based on T2-WI, 0.74 (0.67-0.80, 95% CI) for CE-T1, and 0.81 (0.76-0.86, 95% CI) for all weighted images. After all three MRI modalities were merged, the receiver operating characteristic (ROC) curve was calculated, and the area under the curve (AUC) was 0.93, with an accuracy of 0.87.

CONCLUSIONS

CNN based MRI analysis has the potential to accurately differentiate ependymomas from schwannomas in the lumbar segment.

Factors Affecting Long-Term Surgical Outcomes of Spinal Extramedullary Ependymomas: A Retrospective Study

BACKGROUND

Spinal intradural-extramedullary ependymomas (IEEs) most commonly affect the lumbosacral spine. Because of their neural axis dissemination and adherence to neural structures, managing IEEs is still a controversial neurosurgical challenge. The study aimed to investigate the potential prognostic factors that influence long-term surgical outcomes by evaluating consecutively operated patients with IEEs.

METHODS

During the study period, medical records of all diagnosed patients with spinal tumors were reviewed retrospectively. This study included all patients with consecutive IEEs who underwent surgical intervention in 3 neurosurgical institutions in different periods (February 2004 to December 2020).

RESULTS

In 3 neurosurgical institutions, 64 (28 women, 36 men) patients were operated upon for IEE. The mean age of the patients at diagnosis was 38.9 years. The mean preoperative symptom duration was 17.8 months. Radicular pain was the most common symptom, observed in 53 patients. Gross total resection was applied to 48 patients. 54 patients had good clinical outcomes at their last follow-up after 106.9 months on average. Myxopapillary ependymoma was the most common histopathological type (n = 43). The progression was observed in 5 patients. The presence of preoperative neurological deficits, IEEs extending >2 spinal levels, bone-eroded IEEs, surgical complication, capsule violation, and incomplete resection were associated with poor prognosis.

CONCLUSIONS

Several clinical and surgical factors can affect the functional outcomes of the surgical treatment of IEEs. Some radiological features can make neurosurgeons aware of the prognosis of IEEs. In such cases, the neurosurgeons should plan to remove tumors without violating their integrities (capsules) to obtain satisfactory functional outcomes.

Filum Terminale: A Comprehensive Review with Anatomical, Pathological, and Surgical Considerations

The conus medullaris is the distal tapering end of the spinal cord, and the filum terminale (FT) is regarded as a bundle of non-functional fibrous tissue; therefore, some scholars call it the spinal ligament, while others describe the human FT as "remnants of the spinal cord." It was later found that in the human spinal cord, the FT is composed of an intradural segment and an epidural segment, and the end of the FT is connected to the coccyx periosteum. Because some nerve tissue is also found in the FT, as research progresses, FT may have the potential for transplantation. A lack of exhaustive overviews on the FT in the present literature prompted us to conduct this review. Considering that a current comprehensive review seemed to be the need of the hour, herein, we attempted to summarize previous research and theories on the FT, elucidate its anatomy, and understand its pathological involvement in various diseases.

Prognostic factors for progression-free survival of the filum terminale ependymomas in adults

OBJECTIVE

To define the prognostic factors for progression and to determine the impact of the histological grading (according to the World Health Organization) on the progression-free survival (PFS) of filum terminale ependymomas.

METHODS

A retrospective chart review of 38 patients with ependymoma of the filum terminale was performed, focusing on demographic data, preoperative symptoms, tumor size, quality of resection, presence of a tumor capsule, and histological grade.

RESULTS

Gross total resection (GTR) was achieved in 30 patients (78.9%). Histopathological analysis found 21 (55.3%) myxopapillary grade I ependymoma (MPE), 16 (42.1%) ependymoma grade II (EGII), and 1 (2.6%) ependymoma grade III. There was no significant difference between the mean ± SD volume of MPE (5840.5 ± 5244.2 mm³) and the one of EGII (7220.3 ± 6305.9 mm³, p=0.5). The mean ± SD follow-up was 54.1 ± 38.4 months. At last follow-up, 30 (78.9%) patients were free of progression. In multivariate analysis, subtotal resection (p=0.015) and infiltrative tumor (p=0.03) were significantly associated with progression. The PFS was significantly higher in patients with encapsulated tumor than in patients with infiltrative tumor (log-rank p=0.01) and in patients who had a GTR in comparison with those who had an incomplete resection (log-rank p=0.05). There was no difference in PFS between patient with MPE and EGII (p=0.1).

CONCLUSION

The progression of ependymoma of the filum terminale highly depends on the quality of resection, and whether the tumor is encapsulated. Except for anaplastic grade, histopathological type does not influence progression.

Cauda equina ependymomas: surgical treatment and long-term outcomes in a series of 125 patients

OBJECTIVE

Cauda equina ependymoma (CEE) is a rare tumor for which little information is available on the oncological and clinical outcomes of patients. In this study the authors aimed to address functional, oncological, and quality-of-life (QOL) outcomes in a large series of consecutive patients operated on at their institution during the past 20 years.

METHODS

The records of 125 patients who underwent surgery between January 1998 and September 2018 were reviewed. Analyzed variables included demographic, clinical, radiological, surgical, and histopathological features. Neurological outcomes were graded according to the McCormick and Kesselring scales. The QOL at follow-up was evaluated by administering the EQ-5DL questionnaire.

RESULTS

On admission, 84% of patients had a McCormick grade of I and 76.8% had a Kesselring score of 0. At follow-up (clinical 8.13 years; radiological 5.87 years) most scores were unchanged. Sacral level involvement (p = 0.029) and tumor size (p = 0.002) were predictors of poor functional outcome at discharge. Tumor size (p = 0.019) and repeated surgery (p < 0.001) were predictors of poor outcome. A preoperative McCormick grade ≥ III and Kesselring grade ≥ 2 were associated with worse outcomes (p = 0.035 and p = 0.002, respectively). Myxopapillary ependymoma (MPE) was more frequent than grade II ependymoma (EII). The overall rate of gross-total resection (GTR) was 91.2% and rates were significantly higher for patients with EII (98%) than for those with MPE (84%) (p = 0.0074). On multivariate analysis, the only factor associated with GTR was the presence of a capsule (p = 0.011). Seventeen patients (13.7%) had recurrences (13 MPE, 4 EII; 76.4% vs 23.6%; p = 0.032). The extent of resection was the only factor associated with recurrence (p = 0.0023) and number of surgeries (p = 0.006). Differences in progression-free survival (PFS) were seen depending on the extent of resection at first operation (p < 0.001), subarachnoid seeding (p = 0.041), piecemeal resection (p = 0.004), and number of spine levels involved (3 [p = 0.016], 4 [p = 0.011], or ≥ 5 [p = 0.013]). At follow-up a higher proportion of EII than MPE patients were disease free (94.7% vs 77.7%; p = 0.007). The QOL results were inferior in almost all areas compared to a control group of subjects from the Italian general population. A McCormick grade ≥ 3 and repeated surgeries were associated with a worse QOL (p = 0.006 and p = 0.017).

CONCLUSIONS

An early diagnosis of CEE is important because larger tumors are associated with recurrences and worse functional neurological outcomes. Surgery should be performed with the aim of achieving an en bloc GTR. The histological subtype was not directly associated with recurrences, but some of the features more commonly encountered in MPEs were. The outcomes are in most cases favorable, but the mean QOL perception is inferior to that of the general population.

Ependymomas of Filum Terminale: Clinical Characteristics and Surgical Management of 5 Patients

OBJECTIVE

To retrospective review a series of patients with ependymomas of filum terminale (FTEs) who underwent microsurgical resection and had long-term follow-up.

METHODS

Five adults with FTEs were surgically treated and enrolled in the study. Clinical manifestations, imaging data, intraoperative findings, and postoperative outcome were recorded.

RESULTS

Six FTEs were discovered on preoperative lumbar magnetic resonance imaging, with 1 rare case of a single FTE accompanied by a neurofibroma. Gross total resection was achieved in all patients; en bloc excision was carried out in 2 patients for 2 FTEs and 1 neurofibroma, while piecemeal resection was done in 3 patients for 4 FTEs. Most preoperative symptoms were relieved soon after surgery; 1 patient with urination difficulty achieved complete recovery 1 year postoperatively. Tumor recurrence was found in 1 patient 54 months after her first surgery.

CONCLUSIONS

FTEs are rare subdural extramedullary tumors. Patients usually present initially with back and leg pain. Elaborate microsurgical manipulations and intraoperative electrophysiologic monitoring are helpful in ensuring satisfactory excision and neurological prognosis. Regular long-term follow-up is recommended for all postoperative patients.

Spinal ependymomas. Part 2: Ependymomas of the filum terminale

OBJECT

Ependymomas of the filum terminale provide specific surgical challenges due to their often enormous size, contact with nerve roots of the cauda equina and conus, and potential for subarachnoid dissemination. This study presents treatment results for these tumors over a 30-year period.

METHODS

Among 1447 patients with tumors of the spinal canal treated between 1980 and 2014, 618 patients presented with extramedullary tumors. Of these, 42 patients (25 males, 17 females) demonstrated a myxopapillary ependymoma in the lumbosacral region. Thirty-four patients underwent 36 operations for 39 such tumors. The mean patient age was 38 ± 14 years (range 11−73 years), with an average clinical history of 37 ± 67 months. Patients were followed through outpatient visits and questionnaires, with a mean follow-up of 10 years (127 ± 100 months). Twenty-seven operations were performed to treat de novo tumors and the remainder were undertaken on recurrent tumors. Short-term results were determined for individual symptoms, and tumor recurrence rates were calculated with Kaplan-Meier statistical analyses.

RESULTS

Subarachnoid dissemination was observed in 11 patients and was related to previous surgery in 9 patients and associated with extensive tumors in 2 patients. Gross-total resections (GTR) were achieved in 28 operations (77.7%) and subtotal resections in the remainder. Subtotal resections were restricted to unencapsulated ependymomas (61.5%). Radiotherapy was employed after 6 operations on unencapsulated tumors, with 5 of these also demonstrating subarachnoid seeding. Permanent surgical morbidity affected 3 patients who experienced permanent worsening of bladder function, while 7 patients showed no postoperative changes, and the remaining 26 operations were followed by improvements. Long-term outcome depended on the amount of resection and the presence of a tumor capsule. Eight of 9 tumor recurrences affected unencapsulated tumors, of which 3 had undergone GTR. The overall recurrence rates were 6.6%, 19.0%, and 37.0% after 1, 10, and 20 years, respectively. For unencapsulated ependymomas, the corresponding rates were 15.6%, 32.5%, and 66.2% after 1, 10, and 20 years, respectively, with significantly lower rates of 9.1% after 10 and 20 years for encapsulated tumors. Postoperative radiotherapy tended to prolong the recurrence-free interval for patients with unencapsulated tumors. Five patient deaths occurred during follow-up, of which 2 deaths were tumor related and occurred at 216 and 287 months after surgery.

CONCLUSION

Extramedullary ependymomas are slow-growing tumors in the lumbosacral region, sometimes with an indolent course for long periods of time. Despite their delicate location and often enormous size, surgical morbidity in experienced hands is low, with good chances for postoperative clinical improvements and very low recurrence rates after GTR for encapsulated tumors. The role of postoperative radiotherapy remains controversial. Radiotherapy may be considered after incomplete resections of unencapsulated tumors and/or for patients with subarachnoid dissemination.