Atypical Imaging of Hemorrhagic Lumbosacral Myxopapillary Ependymoma with Histopathological Correlation: A Case Report

Background

Spinal myxopapillary ependymoma (MPE) is a slow-growing tumor arising from ependymal cells of the central nervous system. It was first described by James Watson Kernohan in 1932 [1]. Patients are usually middle age and average 36 years old, and there is a slight male predominance. Common clinical manifestations may include back pain and progressive indolent course of nerve root-related impingement symptomatology, such as lower extremity stiffness, gait ataxia, sensory loss, and paresthesia [2]. The World Health Organization (WHO) classifies the histologic subtypes of ependymoma into 3 grades, and myxopapillary ependymoma and subependymoma are considered grade I tumors [3]. MPE rarely presents with acute neurological compromise [4], and to date, only a few cases have been reported of patients presenting acutely with hemorrhagic MPE [5,6].

Herein, we present a case report of a 16-year-old boy who presented acutely with spontaneous hemorrhagic MPE at the lumbosacral region.

Case Report

A healthy 16-year-old boy without previous medical problems presented with a 3-day history of sudden onset of pain in the low back radiating to both lower limbs. The pain extended from the buttocks to the back of the thighs, reaching to the back of knees, and it was associated with an inability to walk. He also reported constipation for over 2 days. No history of sphincter incontinence, trauma, fever, recent flu symptoms, or previous similar problems. Individual muscle examination while the patient was lying down did not reveal any motor deficit. However, he could not walk with a steady gait, which seemed to be related to the challenge of carrying his bodyweight when standing and walking. Sensory examination was unremarkable.

Magnetic resonance imaging (MRI) of the lumbosacral spine revealed a T2 hypointense intradural mass with corresponding heterogeneous diffusion restriction. The diffusion restriction was thought to be due to a hemorrhagic component, as illustrated by gradient-weighed images. Postcontrast images showed a lack of enhancement in most of the mass, except for an enhancing focus at the posterior aspect of the mass (Figure 1). MRI showed a slight vertebral bone scalloping. No bone erosions or foraminal extension was seen. Computed tomography scan of the lumbosacral spine was performed (not shown), but it did not extend the MRI findings.

The patient was placed in a prone position for operation. L4–L5 laminectomy and durotomy were performed. An intradural encapsulated hemorrhagic mass was visualized (Figure 2). No bone erosion or foraminal extension was demonstrated. Gross total resection (GTR) of the tumor was achieved with the aid of intraoperative neuromonitoring. The neuromonitoring modalities used were somatosensory evoked potentials, motor evoked potentials, and free-running electromyography with direct nerve stimulation. Resection was achieved by microscopic technique and was done piecemeal and intralesion. The tumor was dissected off the nerve roots. After all visible tumor was resected, the last structure appeared to be filum terminale that had firmer tumor attachment; a positive electromyography stimulation was suggestive of a nerve root. Therefore, this structure was not resected with the tumor that was removed from its surface. Primary closure of the dura was then done, and it was covered with fibrin glue medical sealant. No cerebrospinal fluid leak occurred during the postoperative period.

Histopathological examination showed an organized hematoma with an ependymal tumor that had a rich myxoid background. The tumor cells were well differentiated, cuboidal to elongated in shape, and radially oriented around vascularized myxoid cores (Figure 3). Immunohistochemical studies showed that the tumor cells were strongly positive for glial fibrillary acidic protein, pan-cytokeratin, and occasionally for S-100 protein. The Ki-67 proliferative index was low (Figure 4). The findings were consistent with WHO grade I ependymoma of the myxopapillary type.

Postoperatively, the patient was fully ambulatory with no deficits, and the postoperative course was unremarkable. No radio-therapy was given postoperatively because GTR was achieved. Postoperative MRI was not performed.

Discussion

MPE represents 13% to 27% of all spinal ependymomas [1] and 30% to 80% of ependymomas arising from the filum terminale and conus medullaris. MPE may present as an intradural or extradural lesion. Most MPEs present as intradural extramedullary masses, reflecting the usual origin of these tumors and the typical exophytic growth that often fills the distal spinal canal. MPE is typically isointense in relation to the spinal cord on T1-weighted images and hyperintense on T2-weighted images. The tumor typically demonstrates homogeneous postcontrast enhancement [1,7]. In our case, the tumor was T2 hypoin-tense with corresponding heterogeneous diffusion restriction. In addition, postcontrast images showed a lack of typical enhancement in most of the mass, except for an enhancing focus at its posterior aspect. These features were attributed to the hemorrhagic content, which was illustrated by blooming in gradient-weighted images.

Spontaneous hemorrhagic MPE is rare. To date, only 14 case reports describing acute neurological compromise after hemorrhage of spinal MPEs have been published in the English medical literature [8,9]. In our case, the MPE was hemorrhagic and located at the L5–S2 level. A lumbosacral location of hemorrhagic MPE is rare. Among the previously published case reports, only 1 case report described a hemorrhagic MPE at the lumbosacral level [4,10]. MPE is a highly vascular tumor, and it has the highest risk of intratumoral bleeding relative to the other subtypes of ependymal tumors [9,11]. Multiple theories have been proposed to explain the tendency of an MPE to bleed [8,9]. High mobility, physical tension, and muscular mechanical compression to the vessel walls in the lower lumbosacral region may predispose disruption and bleeding of tumoral blood vessels. Additionally, activity associated with elevated intrathoracic pressure, such as heavy lifting, has an effect similar to the Valsalva maneuver on the cardiovascular system. The pressure is transmitted to spinal veins, which can lead to stasis, rupture of tumoral blood vessels, necrosis, and tumoral hemorrhage [8]. Furthermore, vascular insult has been suggested to be predisposed by abuse of anabolic steroids [8]. The average age of presentation has been reported as 34.9 years (range, 15–65 years), and the main presenting symptom is predominantly severe back pain [11]. Our patient was relatively young and physically active, with no history of any proposed risk factors, such as trauma, heavy lifting, or medication use.

The timing of surgical intervention relative to the timing of presenting symptoms is critical. Multiple studies reported a favorable prognosis for early surgical decompression [12]. The mainstay of treatment of spinal MPE is GTR, avoiding violation of the tumor capsule as much as possible. MPE has a high risk of central nervous system dissemination and may recur locally or distally even after a prolonged follow-up period [1]. The risk of recurrence is higher with subtotal removal of the tumor or tumoral capsule disruption. Marginal en bloc resection (i.e., GTR without tumoral capsule disruption) is the strongest predictor of overall survival, long-term control, and progression-free survival [1,13,14]. No consensus has been developed regarding adjuvant radiation therapy (RT) for MPEs [12]. In the majority of cases, RT is used for recurrence or subtotal tumor removal [1]. In a study conducted between 1972 and 2005 by Nakamura and his group, no tumor recurrence was found during a mean postoperative follow-up period of 10.4 years among 14 of 15 patients who underwent GTR followed by either whole brain and spinal cord radiation or local irradiation [11,15]. Feldman et al. [16] found that the recurrence rate of MPE after GTR alone was 15.9%, which was similar to the recurrence rate for patients who underwent GTR with adjuvant RT. However, the recurrence rate for patients who underwent subtotal resection was 35.1% versus 29.3% for patients who underwent subtotal resection with adjuvant RT [11,16]. A similar conclusion was found by Tsai and his group [17]. Moreover, Abdulaziz et al. [14], found a positive trend for increased progression-free survival for patients who underwent adjuvant radiotherapy after either GTR or subtotal resection. Also, they found no recurrence among patients who underwent en bloc resection. In contrast, multiple other studies demonstrated that RT improves local control and progression-free survival regardless of the extent of resection [12,18].

Conclusions

We described a rare case of hemorrhagic MPE in the lumbosacral region in a patient who presented with acute neurological compromise and atypical imaging features.