Metastatic spinal cord compression: radiotherapy outcome and dose fractionation

Abstract

Background and purpose: No standard dose fractionation has been defined for metastatic spinal cord compression. This retrospective analysis was undertaken to explore the impact of hypofractionated treatment compared to conventional multifraction treatment.

Materials and methods: One hundred and two consecutive patients referred to Mount Vernon Cancer Centre with metastatic spinal canal compression confirmed on MR scan in 95% with median age 68 years (range 32-90) and main primary tumour types breast (28%), prostate (28%) and lung (20%); 51% of patients were fully ambulant at diagnosis, 41% ambulant but with paraparesis and 9% had complete paraplegia. Spinal radiotherapy was given delivering a single dose in 32% and 20 Gy in five fractions in 64%.

Results: The median survival was 3.5 months; survival was significantly related to primary site and motor function at presentation. Normal ambulation was achieved in 58% at 2 weeks and 71% up to 2 months after treatment. No patient who presented with paraplegia regained function. At presentation 59% of patients had severe pain, which fell to 8% at 2 weeks. Comparing those patients who received one or two dose treatments with those who received protracted fractionation, the two groups were matched for age, sex, primary site and site of compression. Relatively more patients treated with one or two doses had paraplegia; 19% vs. 3%. Despite this outcome in the two groups was equivalent for motor and sphincter function and pain control.

Conclusions: Metastatic spinal canal compression carries a poor prognosis. Urgent treatment will maintain and improve motor function in patients presenting ambulant but those who have paraplegia at presentation do not improve and have a very short survival. In this series no difference in outcome was seen between patients treated with one or two radiation doses compared to multifraction treatment; a randomised trial comparing fractionation schedules would be justified.