Introduction

Primary spinal canal tumors comprise approximately 15% of all primary central nervous system (CNS) tumors [1]. Ependymomas are the most common neuroepithelial neoplasm in the spinal canal, comprising 50–60% of spinal gliomas [2]. Spinal canal ependymomas have long been characterized as slow-growing tumors with a predominantly local growth pattern, a high rate of local recurrence and a favorable long term survival. Ependymomas are classified by histologic grade as subendymoma (WHO Grade I), myxopapillary ependymoma (WHO Grade I), ependymoma (WHO Grade II); and anaplastic ependymoma (WHO Grade III) [3].

Without prospective randomized trials on this rare tumor, management of primary spinal canal ependymomas is largely based on single institution historical data. Surgery is generally the first line of therapy, and serves the dual purpose of tissue diagnosis and gross tumor excision. The use of adjuvant therapy varies by institution due to uncertainty with regard to the need for radiation after gross total resection, the influence of histology on recurrence patterns, the optimal radiation dose, and the role of chemotherapy. A retrospective study was performed to evaluate the long term efficacy and safety of post-operative radiation therapy in the management of spinal canal ependymoma at our institution.

Methods and materials

This retrospective study was conducted with approval from the Human Studies Committee of the Washington University School of Medicine. Between 1954 and 1997, 22 patients with spinal canal ependymoma were treated with post-operative radiation therapy at our institution.

The median age at diagnosis was 34.7 years (range 9.8–56.1 years). There were 8 (36%) male and 16 (64%) female patients. Duration of symptoms ranged from 1 to 48 months, with a median of 10 months. Common symptoms included back pain (91%), numbness (55%), gait disturbance (32%), radiculopathy (32%), paresthesias (27%) and urinary retention (27%). Common clinical signs included paresis (77%) and hyperreflexia (36%).

Diagnostic evaluation included conventional myelogram only (12 patients), conventional and CT myelogram (7 patients), myelogram and MRI (3 patients) and MRI alone (2 patients). CSF evaluation was negative in 9 patients and not performed in 13 patients. All patients underwent open biopsy with histologic diagnosis: 13 patients (59%) had ependymoma (WHO Grade II) and 9 patients (41%) had myxopapillary ependymoma (WHO Grade I) [3]. The median tumor size was 4.0 cm (range 1.5–15.0 cm). The tumor locations and characteristics are listed in Table 1. Twenty patients (90%) received subtotal resection (STR) and 2 patients (10%) received gross-total resection (GTR). Twenty patients (91%) received radiation therapy after surgical treatment. Two patients (9%) received salvage radiation therapy for recurrence after treatment with surgery alone (GTR in one case and STR in the other). Median dose of radiation prescribed was 45.0 Gy (range 30.0–54.0 Gy). Median daily fraction size was 1.8 Gy (range 1.5–2.5 Gy). Radiation treatment parameters are listed in Table 2. None of the patients received chemotherapy as part of initial management.

Table 1 Subsites of involved disease
Full size table
Table 2 Radiation treatment parameters
Full size table

After completion of treatment, patients were followed at 3 month intervals for the first 2 years, then every 6–12 months for 5 years and sporadically thereafter. Evaluations at the time of follow-up consisted of a history and physical examination. Computed tomography (CT) scans or magnetic resonance imaging (MRI) of the spinal canal were only conducted if indicated by patient symptoms or signs. Patients were considered to have local failure if there were clinical, radiographic, or histologic evidence of recurrence. Duration for endpoints was calculated from the date of completion of radiation therapy.

StatView software (SAS Institute, Cary, NC) was used to calculate survival rates based on the Kaplan–Meier method. Univariate analyses were conducted by the log-rank test. A P value of ≤0.05 was considered statistically significant.

Results

The median follow up for all patients was 10 years (range 0.4–37.0 years). The median follow up for surviving patients was 11.4 years (range 0.6–37.0 years). An 80% progression free survival (PFS) was observed for all patients at 5-, 10- and 15-year endpoints (Fig. 1). Of the 4 patients (18.1%) who recurred: 2 patients recurred within the radiation fields 17- and 28-months after treatment; one patient recurred in the spine outside of the treatment field 20 months after treatment; and one patient recurred in the treatment field as well as in the untreated cranium 5 months after treatment. Mean time to recurrence was 17 months. All recurrences were within 3 years of treatment. The 5-, 10- and 15-year overall survivals (OS) for all patients were 85%, 78% and 64%, respectively (Fig. 2). Four patients died of disease, 2 patients died of inter-current disease, and 16 patients were censored at last follow up without evidence of disease.

Fig. 1
figure 1

Progression-free survival for all patients

Full size image
Fig. 2
figure 2

Overall survival for all patients

Full size image

Six patients (27%) demonstrated long term neurologic deficits after treatment. Symptoms included paresis (2 patients), urinary retention (2 patients), urinary incontinence (1 patient) and arachnoiditis (1 patient). All patients had complaints prior to the start of radiation, suggesting that the symptoms were sequelae of tumor invasion or surgical resection, however contribution from radiation cannot be excluded.

Various patient, tumor and treatment factors were examined to determine their influence on prognosis. A worse outcome was observed with larger tumors (Fig. 3). Patients with tumors greater than 6 cm at time of presentation demonstrated 10-year PFS of 58.3% compared to 92.3% for patients with tumors 6 cm or smaller. This difference was statistically significant (P = 0.047). There was no significant correlation between tumor size and OS. In this retrospective series, no prognostic value was noted for gender, age, dose prescribed, volume of irradiation, histologic grade, extent of surgery, timing of radiation or era of treatment.

Fig. 3
figure 3

Progression-free survival for patients with tumors ≤6 cm or >6 cm

Full size image

Discussion

Reported survival rates for patients with spinal canal ependymoma after surgery and post-operative radiation range from 68 to 95% at 10 years [4–13]. The median follow up of 11.4 years obtained with this series is quite lengthy with respect to prior studies and provides further evidence of a sustained favorable outcome for these patients.

Institutional reports suggest the potential for excellent control rates with surgery alone for low grade lesions that are completely removed [14–19]. However, progression rates after partial or subtotal tumor removal range from 20 to 50% at 5 years [10, 13, 19–21]. Despite the fact that 90% of the patients in our study received only subtotal resection, an 80% local control rate was maintained at 15 years with the use of radiation therapy, suggesting that post-operative radiation is effective and should be considered after incomplete resection of tumor. Recurrence rates in series that include high grade tumors (current WHO Grade III) range from 16 to 37% even after documented GTR [9–11], supporting the use of adjuvant radiation for high grade lesions irrespective of the degree of resection.

It is difficult to draw conclusions on the prognostic value of patient, tumor or treatment variables given the small sample size in our series. Our data suggest a PFS advantage with tumors 6 cm or less. Other reports have suggested improved outcome with younger age [13], smaller tumor size [9], distal spinal disease [22], myxopapillary histology [12], low tumor grade [13, 23], gross total resection [8, 10], post-operative radiation [12] and radiation dose above 50 Gy [9].

Our study does not demonstrate a dose response relationship for tumor control. Some investigators have observed a trend towards improvement with doses of 50 Gy or higher and advocate for treatment to 55 Gy, with the last 5 Gy given to a boost volume [9]. A dose range of 45–50 Gy has been used historically as the threshold dose beyond which the incidence of radiation myelopathy is thought to increase significantly. Current models of spinal cord tolerance suggest that up to 55 Gy in conventional fractions (2 Gy or less per day) can be delivered safely with a less than 2% risk of causing radiation myelopathy [24–29]. Nevertheless, in the absence of strong evidence for a dose–response, most institutions remain cautious about escalating dose beyond 50 Gy and continue to recommend doses in the range of 40–50 Gy [11–13, 22, 30–32]. Only 2 patients in our series were treated beyond 50 Gy (both received 54 Gy in 1.8 Gy fractions). Radiation therapy did not seem to cause treatment related late effects within our population, suggesting that the doses used in our study (range 30 Gy–54 Gy; median 45 Gy) can be delivered safely.

Only 1 patient in our series failed outside of the localized treatment field. The vast majority of spinal ependymoma recurrences occur at or near the primary site. Of those patients who fail at distant sites in the CNS, many do so despite the addition of cranio-spinal irradiation (CSI) [13, 30]. Whereas the increased morbidity associated with CSI is well established, there is little evidence in the literature that whole-CNS or whole-spinal irradiation adds tumor control or survival advantage for non-disseminated lesions. The role of large volume irradiation should therefore be limited to patients with disseminated disease.

Chemotherapy has a limited role in the management of spinal ependymomas. There is no data to suggest a benefit for chemotherapy in the initial treatment of adults. Treatment of very young patients is individualized and sometimes utilizes chemotherapy in an attempt to delay radiation. Several prospective randomized trials of chemotherapy in intracranial ependymoma have failed to demonstrate a local control or survival advantage [33–35]. The efficacy of chemotherapy continues to be investigated in clinical trials.

Improvement in both surgical and radiation treatments is expected to have occurred over the time course of this study. Although we did not find a difference in outcome of our patients by year of treatment, other investigators have shown improved outcome with later eras of treatment [13]. Improved microsurgical techniques and earlier diagnosis through CT and MR imaging have contributed to improved chances of GTR at first presentation. The use of three dimensional imaging for radiation treatment planning allows for more conformal radiation delivery in the modern era. New treatment modalities such as intensity modulated radiation therapy, image guided radiation therapy, stereotactic radiosurgery and helical tomotherapy will theoretically allow for improvement in the therapeutic ratio.

Conclusions

Post-operative radiation after subtotal resection is safe and offers durable tumor control and long term patient survival.