Final results of a study comparing short-course and long-course radiotherapy (RT) for local control of metastatic spinal cord compression (MSCC)

Abstract

9522 Background: Many MSCC patients live long enough to develop a recurrence in the irradiated spinal area. This is the first prospective study comparing different RT schedules for local control (LC) of MSCC. Methods: 265 patients treated with RT alone (1/06–12/07) were included in this prospective non-randomized study. The primary goal was to compare short-course (1×8 Gy/5×4 Gy, N=131) and long-course RT (10×3 Gy/15×2.5 Gy/20×2 Gy, N=134) for 1-year LC. Secondary endpoints were motor function and 1-year survival (OS). Dutch patients received short-course, and German patients long-course RT. The analysis of LC (no MSCC recurrence in the irradiated spinal area) included the 224 patients with improvement or no change of motor deficits during RT. The difference in 1-year LC was previously reported to be 14% between short- and long-course RT. For a statistical power of 90 % (significance level 5%), ≥218 patients were required to detect this difference. Univariate analyses (UVA) for LC and OS were performed with Kaplan-Meier-method and log-rank test, multivariate analyses (MVA) with the Cox proportional hazards model. UVA and MVA for motor function were performed with the ordered-logit-model. Eleven additional factors were evaluated. Results: 1-year LC was 61% after short-course RT and 81% after long-course RT (P=0.005). On MVA, improved LC was only associated with long-course RT (P=0.018). Motor function improved in 37% after short- and 39% after long-course RT (P=0.95). Improved motor function was associated with better performance status (P=0.015), favorable tumor (P=0.034), and slower development of motor deficits (P<0.001). 1-year OS was 23% after short- and 30% after long-course RT (P=0.28). On MVA, improved OS was associated with better performance status (P<0.001), no visceral metastases (P<0.001), involvement of 1–3 vertebrae (P=0.040), ambulatory status (P=0.038), and bisphosphonates (P<0.001). Conclusions: Long-course RT was associated with better LC, similar functional outcome, and similar OS compared to short-course RT. Patients with a favorable OS prognosis should receive long-course RT, and those with a poor OS prognosis should receive short-course RT. [Table: see text]