Long-Term Clinical Safety of High Proton Radiation Doses Delivered with Pencil Beam Scanning Paradigm to the Spinal Cord for Extracranial Malignancies

Standardly employed radiotherapy dose limits for the spinal cord tend to be very conservative. Historical data from re-irradiation and combined photon/proton beam series suggest that the tolerance might be much higher. Nevertheless, uncertainties persist due to short follow-up, mixed techniques, and dose summation in these efforts. To add further clarity, we analyzed our institutional experience regarding the incidence of radiation-induced spinal cord toxicity following treatment exclusively with high dose, pencil beam scanning proton therapy (PBSPT). Seventy two patients (median age: 51, range 23-79; male/female ratio 1:1), treated for spinal chordoma (n=51) or chondrosarcoma (n=21), met the following criteria and were retrospectively analyzed: maximum dose to the spinal cord of ≥ 45 Gy (RBE), proton therapy only, ≥ 18 years of age, and clinical follow-up of >12 months. Patients received 64.0 – 74.4 Gy (RBE) (median: 74.0 Gy (RBE)), generally with 1.8-2.0 Gy (RBE) per fraction, between 2000 and 2014. The Dmax, D2%, V40, V45, V50, V55, and V60 of the spinal cord (SC) and center of spinal cord (CSC; 2-3 mm diameter ROI at geometric center of the cord) were calculated for all patients. Median follow-up was 66 months (range: 13-160 months). Toxicity was scored according to the Common Toxicity Criteria of Adverse Events (CTCAE v4.03). Descriptive and linear regression analyses were performed. Patients received a mean Dmax / D2% to the spinal cord of 59 Gy (RBE) / 55 Gy (RBE) and to the CSC of 52 Gy (RBE) / 51 Gy (RBE), respectively. Four patients (6%) developed acute radiation-induced neurologic toxicity (Grade 1: n=1, Grade 2: n=3). Seventeen percent of patients experienced minor late neurologic toxicities (Grade 1: n=7, Grade 2: n=4). Of these patients, 4 developed Lhermitte’s syndrome. One patient experienced Grade 4 toxicity (tetraplegia) after receiving 58 Gy (RBE) to the surface and 54 Gy (RBE) to the CSC (SC: V45 11.8 cc, V50 9.35 cc; CSC: V40 3.17 cc, V45 3.1 cc). Of note, the patient was considered high-risk due to postoperative, transient tetraparesis prior to PBSPT and potential spinal instability with a narrow canal. On final analysis, no significant correlation was found between spinal cord Dmax, CSC Dmax, or length of CTV and toxicity. However, higher CSC Dmax and longer CTV tended to correlate with ≥ Grade 2 neurologic toxicity, but this was not statistically significant. PBSPT with spinal cord doses of up to 60 Gy (RBE) to the surface and 50 Gy (RBE) to the center is associated with minimal spinal toxicity. In our experience, doses up to 64 Gy (RBE) to the surface of the spinal cord are deemed acceptable if unavoidable in achieving adequate tumor coverage.