Spinal ependymomas--the value of postoperative radiotherapy for residual disease control.

Spinal ependymomas are rare intramedullary tumours, usually with a benign biological nature. They have, however, a known potential for recurrence and metastatic spread through the CSF pathways. Traditionally, radiotherapy has been employed as an adjuvant to surgery. Recently, however, this view has been increasingly questioned. Thirty-eight adult patients operated in the West Midlands between 1956 and 1994 were reviewed. Mean age at presentation was 37.8 years. Of all tumours 45% (17) were situated in the conus/cauda equina region, 26% (10) in the thoracic and 29% (11) in the cervical region. All patients underwent surgery as primary treatment. Postoperative mortality was 2.6% (one patient). Complete macroscopic excision of their tumours was performed in 14 patients; of these three received postoperative radiotherapy. Subtotal excision was performed in 18 patients, seven receiving adjuvant radiotherapy. Decompressive laminectomy and biopsy was performed in five patients with either no radiotherapy in one, radiotherapy in three or radiotherapy and chemotherapy in one. The overall survival rate was 87% at 5 years and 73% at 10 years for the whole series. In patients treated after the 1970s, the cumulative 10-year survival was 95%. Patients over the age of 40 years at diagnosis had a 10-year survival of 61% compared to 78% of those under the age of 40. Intramedullary tumours had a worse outcome with 68% survival at 10 years, compared to 89% of tumours at the cauda equina. Total excision carried 86% survival, subtotal excision 80% and biopsy only 25% at 10 years. In those patients treated by administration of postoperative radiotherapy the 10-year survival was 48%, compared to 96% in those in whom radiotherapy was not given postoperatively. Analysis of the recurrence and survival rates demonstrated that radiotherapy did not confer any significant protection against recurrence or disease progression. Whilst complete excision offers the maximal potential for cure, an expectant policy with possible repeat surgery may be appropriate when total excision is not achieved. The main conclusion to be drawn is that a randomized controlled trial of the value of adjuvant radiotherapy is greatly needed.