(P044) Feasibility of Preoperative SBRT for Chondrosarcomas and Chordomas

Julian Johnson, MD, Steve Braunstein, MD, PhD, Igor J. Barani, MD; University of California, San Francisco

Purpose: To determine the outcomes of chordomas and chondrosarcomas treated with preoperative, postoperative, and definitive CyberKnife radiosurgery. The goals of preoperative radiosurgery are to improve disease-free survival (DFS) by achieving clean margins and to decrease morbidity by achieving a smaller resection volume.

Materials and Methods: We identified 14 patients with chordomas and chondrosarcomas treated at our institution with radiosurgery using CyberKnife robotic radiosurgery and a stereotactic body radiation therapy (SBRT) system. Using original imaging at the time of diagnosis, preferably fat-saturated, T2-weighted sequences, clinical target volume (CTV) was defined as gross disease and, in the spine, the entire involved vertebral body or sacral segment. Any tissue felt to be at risk was included. For patients treated preoperatively, the entire tumor was targeted along with a 3–5-mm margin. In all cases, a planning target volume (PTV) was generated by expanding the target by 2–3 mm. All patients received 25–45 Gy in five fractions. All patients had follow-up magnetic resonance imaging (MRI) with T2-weighted sequences. Some patients with spine or sacral tumors also had follow-up CT scans. Toxicity data were retrospectively collected by reviewing clinic notes.

Results: The median age for these patients was 46 years (range: 15–76 yr). Eastern Cooperative Oncology Group (ECOG) performance status ranged from 0–1. Eight tumors were intracranial, three were in the spine, and three were in the sacrum. Four patients received prior external beam (three with photons, one with protons; range: 50.4–73.8 Gy) (all postoperative). Of the patients who had undergone surgery, six were gross total, three were near-total, and three were subtotal resections. The median gross tumor volume was 20.9 cc (mean 54.72; range: 3–309.4 cc). The corresponding original PTVs were 27.34, 97.63 (range: 11.16–418 cc). Median time between imaging diagnosis and CyberKnife was 7.78 months. Mean dose was 35 Gy (median 35 Gy; range: 25–40 Gy). Mean maximum dose to PTV was 55.23 Gy (median 57.21 Gy; range: 37.63–67.80 Gy). Median imaging follow-up after CyberKnife was 6.27 months. Median imaging follow-up after diagnosis was 17.57 months (range: 3.9–173.8 mo). The median follow-up after diagnosis was 17.67 months (range: 6.77–173.87 mo). Only one recurrence was noted after CyberKnife. The recurrence was 3.47 months after CyberKnife in a patient who had previously undergone surgery. No patients developed metastatic disease. One patient is deceased. There were no recurrences in the definitive or preoperative cohorts. There were five toxicities > grade 2. One patient treated with preoperative radiosurgery for a large sacral chondrosarcoma experienced a grade 3 postoperative wound infection and poor wound healing. There were too few patients and too few recurrences to determine whether a dose-response relationship was present. For the entire cohort, overall survival (OS) was 93% and locoregional control (LRC) was 86%. There was no difference in LRC, OS, or toxicity based on pre-, post-, or definitive CyberKnife radiosurgery (CKRS).

Conclusion: SBRT for definitive, preoperative, postoperative, or salvage treatment of chordomas and chondrosarcomas demonstrates a local control of 86%. In the future, we would seek to hypofractionate spine or sacral chordomas preoperatively and collect more data on the incidence of toxicities, especially wound healing and infection.