Salvage Reirradiation for Spinal Metastases Following Local Failure after Stereotactic Radiosurgery

Abstract

The management of recurrences following spine stereotactic radiosurgery (SSRS) poses a challenge due to concerns for toxicity associated with reirradiation (re-RT) and a paucity of data on treating SSRS failures. We report our institutional experience on salvage re-RT for spinal tumors using conventional radiotherapy (RT) and SSRS techniques following an ablative SSRS regimen. We retrospectively identified 33 metastatic spinal segments in 30 patients who received salvage conventional re-RT or SSRS following imaging-confirmed local progression at a site which previously received SSRS between 2003 and 2013 at our institution. Clinicopathologic features were abstracted from the medical record. Overall survival (OS) for each patient and local control (LC) for each treated segment were calculated from the date of initial SSRS and estimated using the Kaplan-Meier method. Median follow-up was 26 months (range, 2-168). The 1- and 2-year OS was 77% and 50%, respectively, with a median OS of 27 months (range, 3-168). Male patients and renal clear cell carcinoma tumors comprised 57% (17/30) and 33% (11/33) of the cohort, respectively. Eight spinal segments among 7 patients were irradiated with conventional RT prior to SSRS and then a second course of conventional RT: 5 spine directed-treatments, 2 from thoracic-directed RT, and 1 from pelvic-directed RT. The median initial SSRS dose and number of fractions was 27 Gy in 3 fractions (range, 16-30 Gy in 1-5 fractions). Treatment sites included the C-spine (21%), T-spine (45%), L-spine (30%), and sacrum (3%). The median duration to local failure after the initial SSRS course was 9 months (range, 1-67) with a median time from initial SSRS to salvage RT of 13 months (interquartile range, 7-34). For salvage, a total of 25 spinal segments were treated with RT without surgery: 16 segments in 14 patients were treated with conventional RT (20-36 Gy /5-14 fractions) whereas 9 segments in separate patients were salvaged with a second course of SSRS (18-27 Gy/1-3 fractions). Additionally, 8 segments received surgery followed by postoperative RT (n=5, conventional RT – 30 Gy /10-20 fractions; n=3 SSRS 20-27 Gy/1-3 fractions). Following salvage re-RT, the crude LC rate was 76% (25/33 lesions), with a median LC of 19 months (range, 2-23). No local failures were observed in the 8 spinal segments treated salvage surgery and postoperative RT. No grade ≥3 neurological toxicities or vertebral compression fractures were observed after re-RT. We observed favorable LC rates following salvage re-RT with low toxicity in patients who locally recurred following previous SSRS. We recommend ideally at least a 6 month window between initial SSRS and salvage re-RT. Prospective clinical studies are needed to determine the optimal salvage dose and fractionation.