Giant cauda equina neuroendocrine tumor with superficial siderosis: illustrative case

Abstract

BACKGROUND

Cauda equina neuroendocrine tumor (CENET) is a rare spinal neuroendocrine tumor, and its coexistence with superficial siderosis (SS) is even more rare. CENETs are usually small and most commonly located below L2, and their imaging often overlaps with that of myxopapillary ependymomas or schwannomas, thus complicating their preoperative differentiation. The coexistence of SS and intratumoral fluid-fluid level (FFL) in CENET has not been documented.

OBSERVATIONS

The authors describe a giant thoracolumbar CENET (T12–L3) with intracranial SS in a patient with progressive paraparesis and bladder dysfunction. MRI revealed a hypervascular intradural extramedullary mass with an intratumoral FFL and neuraxial hemosiderin deposition, suggesting recurrent hemorrhage. Because of dense cauda equina adhesions, function-preserving subtotal resection was performed under intraoperative neurophysiological monitoring, resulting in marked neurological improvement.

LESSONS

In CENETs, the coexistence of SS, intratumoral FFL, and marked hypervascularity on preoperative imaging may indicate a chronic hemorrhagic phenotype with severe adhesions to the cauda equina, particularly in large multisegmental tumors. In such cases, nerve-sparing subtotal resection planned with preoperative vascular assessment and performed under intraoperative neurophysiological monitoring can be an appropriate strategy to balance tumor control with preservation of neurological function when safe dissection planes are limited.

https://thejns.org/doi/10.3171/CASE251015

Cauda equina neuroendocrine tumors (CENETs) are rare intradural extramedullary neoplasms, which were formerly termed “cauda equina paraganglioma.”¹ In the 2021 WHO classification, the term was redefined as CENET to reflect the distinct molecular and immunophenotypic features that separate these tumors from extra-CNS paragangliomas.^2,3 CENETs most commonly arise below the L2 level and are typically small (mean diameter approximately 2.5 cm) and slow-growing.^4–6 Because their MRI appearance often overlaps with that of schwannomas and myxopapillary ependymomas, preoperative identification remains challenging.⁴

Superficial siderosis (SS) is caused by subpial hemosiderin deposition due to chronic low-grade subarachnoid bleeding and is best visualized using MR susceptibility-weighted imaging (SWI). Dural defects and CSF leaks are the most common causes, whereas tumor-related bleeding is uncommon.⁷ Intratumoral fluid-fluid levels (FFLs) similarly indicate layered blood products, suggesting ongoing or recurrent intralesional hemorrhage.⁸ However, the coexistence of SS and FFL in CENET is exceedingly rare, and its clinical and surgical implications remain unclear because of limited reporting.

Here, we present the case of a giant CENET measuring approximately 11 cm, extending from T12 to L3, causing progressive neurological dysfunction and SS involving both the brain and spinal cord, attributed to chronic hemorrhage. The tumor displayed marked hypervascularity and firm adhesion to cauda equina roots, necessitating function-preserving subtotal resection under intraoperative neurophysiological monitoring.

This report emphasizes that hemorrhagic imaging features such as SS and FFL may serve as preoperative indicators of a bleeding-prone, adhesively growing CENET and should influence the surgical strategy to prioritize neurological preservation.

Illustrative Case

A 54-year-old woman presented with an 8-year history of intermittent buttock pain followed by progressive weakness of the left lower limb about 1 year before admission. Six months prior to admission, she developed bilateral calf paresthesia and tightness, and 1 month prior, she experienced rapid worsening of bilateral leg weakness with gait disturbance. On admission, manual muscle testing revealed grade 4/5 weakness in both lower limbs, absent patellar and Achilles reflexes, and sensory impairment below the inguinal ligament. Neurogenic bladder and bowel dysfunction were observed. No prior trauma, medication, or spinal interventions were reported.

Spinal MRI revealed a giant intradural extramedullary tumor spanning T12–L3 (approximately 11 cm), with T1 mild hyperintensity (Fig. 1A), heterogeneous T2 hypointensity (Fig. 1B), contrast enhancement, intratumoral flow voids, intratumoral FFL (Fig. 1C and D), heterogeneous hypointensity on sagittal STIR images (Fig. 1E), and intratumoral and peritumoral flow voids on sagittal SWI (Fig. 1F). A multiplanar reconstruction image demonstrating the left L3 segmental artery is shown (Fig. 1G). Selective spinal angiography revealed marked hypervascularity with diffuse tumor blush supplied by feeders from the left L2–3 radiculopial arteries (Fig. 1H and I). Cranial SWI showed diffuse hypointensity along the supratentorial and infratentorial cortical sulci, whereas thoracic T2-weighted MRI demonstrated hypointensity along the dorsal cord, consistent with SS (Fig. 2).

FIG. 1.

Thoracolumbar spine MR images and spinal angiograms. A: Sagittal T1-weighted image showing a giant intradural tumor extending from T12 to L3. B: Sagittal contrast-enhanced T1-weighted image demonstrating heterogeneous enhancement of the tumor. C and D: Sagittal (C) and axial (D) T2-weighted images demonstrating an intratumoral FFL (arrow). E:Sagittal STIR image showing heterogeneous hypointensity within the lesion. F: Sagittal susceptibility-weighted image revealing intratumoral and peritumoral flow voids (circles). G: Multiplanar reconstruction image of the left L3 segmental artery. H and I: Selective anteroposterior (H) and lateral (I) angiograms of the left L3 segmental artery demonstrating diffuse tumor blush corresponding to the T12–L3 lesion, supplied by small feeder branches arising from the left L2 and L3 radiculopial arteries.

FIG. 2.

Thoracic and cranial MR images demonstrating SS. A: Sagittal T2-weighted image showing linear hypointense bands along the dorsal surface of the thoracic spinal cord (T4–7), above the tumor (T11–12), and over the cerebellar surface (circles). B: Axial T2-weighted image at the T6 level demonstrating a hypointense coating along the entire circumference of the spinal cord (arrowheads). C and D:Cranial susceptibility-weighted images showing diffuse hypointensity along the cortical sulci in the infratentorial (C) and supratentorial (D) regions (circles).

A T12–L3 laminectomy was performed. The soft, encapsulated tumor was firmly adherent to multiple cauda equina roots, particularly at the inferior pole (Fig. 3). Although gross-total resection was initially considered preferable, dense adhesions and neurophysiological changes prompted function-preserving subtotal resection, prioritizing neurological safety.

FIG. 3.

Intraoperative findings. A: Surgical exposure after T12–L3 laminectomy demonstrating the intradural tumor. B: Intraoperative view showing a well-encapsulated tumor compressing the cauda equina. C:Postresection image demonstrating subtotal tumor resection with adequate decompression of the cauda equina.

MRI on postoperative day 3 showed no definite residual tumors (Fig. 4A and B). At 12 and 18 months, the tiny enhancing nodule remained stable without any hemorrhagic changes (Fig. 4C–F). The patient experienced marked pain relief, recovered ambulation, and regained independence in activities of daily living.

FIG. 4.

Postoperative follow-up MR images. A and B: Sagittal gadolinium-enhanced T1-weighted (A) and sagittal T2-weighted (B) MR images obtained 3 days after surgery, showing postoperative changes without definite residual tumor. C and D: Sagittal gadolinium-enhanced T1-weighted (C) and sagittal T2-weighted (D) MR images obtained 2 months after surgery, demonstrating a small enhancing residual tumor nodule at the inferior pole (arrows). E and F: Sagittal gadolinium-enhanced T1-weighted (E) and sagittal T2-weighted (F) MR images obtained 18 months after surgery, showing a stable enhancing residual nodule without progression (arrows).

Hematoxylin and eosin staining revealed a nested growth pattern of uniform tumor cells with delicate fibrovascular septa, without significant atypia, mitotic figures, or necrosis (Fig. 5A). Immunohistochemistry demonstrated synaptophysin and chromogranin A positivity in addition to cytokeratin expression (CAM5.2 and AE1/AE3) (Fig. 5B–E). S100 staining was present but lacked a sustained rim pattern (Fig. 5F). Staining for glial fibrillary acidic protein (GFAP), epithelial membrane antigen (EMA), and CD34 was negative (Fig. 5G–I). The Ki-67 labeling index was 3% (Fig. 5J). These features supported the diagnosis of CENET (WHO grade 1) rather than paraganglioma.

FIG. 5.

Histopathological and immunohistochemical findings. A: Hematoxylin and eosin staining showing tumor cells arranged in a nested growth pattern. B:Synaptophysin immunostaining showing strong diffuse cytoplasmic positivity in tumor cells. C: Chromogranin A immunostaining demonstrating granular cytoplasmic positivity. D: CAM5.2 immunostaining showing cytoplasmic positivity in tumor cells. E: AE1/AE3 immunostaining confirming cytokeratin positivity. F:S100 immunostaining showing scattered positive cells without a continuous sustentacular rim pattern. G:GFAP immunostaining showing no expression in tumor cells. H: EMA immunostaining showing no expression in tumor cells. I: CD34 immunostaining highlighting intratumoral vascular structures, with tumor cells remaining negative. J:Ki-67 immunostaining demonstrating a low proliferative index of approximately 3%. Original magnification ×40 (A–J).

Informed Consent

The necessary informed consent was obtained in this study.

Discussion

Observations

This case describes a giant thoracolumbar CENET (T12–L3, approximately 11 cm) accompanied by intracranial SS. Preoperative imaging also demonstrated an intratumoral FFL and marked hypervascularity on spinal angiography. These features have rarely been reported together in CENET, and its surgical implications remain unclear due to limited reporting.

CENETs are typically small and situated at the L2–5 level. Based on reported cases of CENET since 2021, we summarized 85 cases across 9 studies.^2,4,9–15 Tumor size was documented in 29 cases and ranged from 0.7 to 9.0 cm (mean 3.1 cm) in diameter. The spinal level was specified in 63 cases, spanning from L1 to S2, with 57 of these 63 (90.5%) cases situated between L2 and L5. The tumor extended from T12 to L3 and measured 11 cm, qualifying it as one of the largest CENETs reported to date.¹⁶ This unusually high thoracolumbar location likely contributed to an earlier involvement of the lower lumbar roots, resulting in bilateral weakness, sensory disturbances, and sphincter dysfunction.¹⁷ The multisegmental longitudinal extent also suggested a long-standing lesion and increased the complexity of safe surgical dissection.

Although the MRI appearance of CENET is often nonspecific, cranial SWI hypointensity and thoracic T2 hypointensity, which are indicative of SS, were important diagnostic clues in this case. This pattern was consistent with 6 previously reported cases of cauda equina or filum terminale paragangliomas associated with SS, in which tumor hypervascularity caused chronic low-grade hemorrhage with persistent leakage into the CSF, resulting in hemosiderin deposition along the neuraxis^18–23 (Supplementary Table 1). Notably, SS in this region is uncommon and frequently overlaps radiologically with myxopapillary ependymoma, or less commonly schwannoma, warranting the inclusion of CENET in the differential diagnosis when SS is present.

A distinctive feature of this case was the coexistence of SS and intratumoral FFL, a combination rarely documented in the literature. The FFL reflects the layering of blood products and suggests repeated intratumoral hemorrhage.²⁴ Consistent with this interpretation, spinal angiography demonstrated marked hypervascularity with a diffuse tumor blush and multiple segmental feeders, supporting the notion of a bleeding-prone lesion. Although causality cannot be confirmed in a single case, the triad of SS, FFL, and hypervascularity strongly suggests a chronic hemorrhagic tumor phenotype. Such repeated bleeding-healing cycles may promote fibrosis, leading to tight adhesions between the tumor and the cauda equina roots.²⁵ This interpretation aligned with a previous report of SS-associated paraganglioma, in which a nonresponsive nerve root required sacrifice during an attempted gross-total resection.²⁰ Thus, hemorrhagic imaging markers should alert surgeons of the risk of dense adhesions and potential neurological compromise during dissection.²⁶

In the present case, preoperative concern regarding these dense adhesions was confirmed intraoperatively, as the tumor showed firm attachment to multiple cauda equina roots (particularly at its lower pole). Given these findings, the preservation of neurological function was prioritized over aggressive radical excision, and function-preserving subtotal resection was performed. Selective spinal angiography and continuous intraoperative neurophysiological monitoring support safe decision-making and dissection.^27,28 During the 18-month follow-up, MRI revealed no tumor progression, supporting the effectiveness of subtotal resection combined with surveillance when dense adhesions precluded safe gross-total resection.

As in previous reports, the absence of a peripheral sustentacular S100 rim combined with cytokeratin positivity (CAM5.2 and AE1/AE3) provided key histopathological evidence distinguishing CENET from paraganglioma, which is consistent with the 2021 WHO classification.^2,3

Lessons

This case highlighted that giant multisegmental intradural cauda equina tumors with hemorrhagic imaging features, specifically the triad of SS, intratumoral FFL, and marked hypervascularity, should be suspected as CENET. Recognizing this pattern is clinically essential because it suggests a chronic hemorrhagic phenotype associated with severe adhesions, indicating that safe gross-total resection may not be feasible. In such situations, nerve-sparing subtotal resection planned with preoperative vascular assessment and performed under intraoperative neurophysiological monitoring can be an appropriate and safe strategy that balances neurological preservation with adequate tumor control.

Acknowledgments

We thank Editage for English-language editing.

Disclosures

The authors report no conflict of interest concerning the materials or methods used in this study or the findings specified in this paper.

Author Contributions

Conception and design: Hiu, Lin, Yoshida, Ujifuku, T Matsuo. Acquisition of data: Hiu, Kurohama, Matsunaga, Nakashima. Analysis and interpretation of data: Hiu, Lin, Kurohama, Tsunoda, Nakashima. Drafting the article: Hiu, Lin, Abe, Matsunaga. Critically revising the article: Hiu, Tsunoda, Yoshida, A Matsuo, Baba, Ujifuku. Reviewed submitted version of manuscript: Hiu, Lin, Tsunoda, Uotani, Takeuchi, A Matsuo, Uchida, Matsunaga, Baba, Nakashima. Approved the final version of the manuscript on behalf of all authors: Hiu. Statistical analysis: Lin. Administrative/technical/material support: Hiu, Koike, Shirahama. Study supervision: Hiu.

Supplemental Information

Online-Only Content

  Supplementary Table 1. https://thejns.org/doi/suppl/10.3171/CASE251015.

Correspondence

Takeshi Hiu: Nagasaki University Graduate School of Biomedical Sciences, Nagasaki, Japan. takeshihiu@nagasaki-u.ac.jp.