Dose-Escalated 2-Fraction Spine Stereotactic Body Radiation Therapy: 28 Gy Versus 24 Gy in 2 Daily Fractions

Abstract

Stereotactic body radiation therapy (SBRT) for spine metastases improves pain response rates compared with conventional external beam radiation therapy; however, the optimal fractionation schedule is unclear. We report local control and toxicity outcomes after dose-escalated 2-fraction spine SBRT. A prospectively maintained institutional database of over 600 patients and 1400 vertebral segments treated with spine SBRT was reviewed to identify those prescribed 28 or 24 Gy in 2 daily fractions. The primary endpoint was magnetic resonance imaging based local failure (LF), and secondary endpoints included overall survival and vertebral compression fracture (VCF). A total of947 treated vertebral segments in 482 patients were identified, of which 301 segments in 159 patients received 28 Gy, and 646 segments in 323 patients received 24 Gy in 2 fractions. Median follow-up per patient was 23.5 months, and median overall survival was 49.1 months. In the 28 Gy cohort, the 6-, 12-, and 24-month cumulative incidences of LF were 3.5%, 5.4%, and 11.1%, respectively, versus 6.0%, 12.5%, and 17.6% in the 24 Gy cohort, respectively (P=.008). On multivariable analysis, 24 Gy (hazard ratio [HR], 1.525; 95% confidence interval, 1.039-2.238; P=.031), paraspinal disease extension (HR, 1.422; 95% confidence interval, 1.010-2.002; P=.044), and epidural extension in either radioresistant or radiosensitive histologies (HR, 2.117 and 1.227, respectively; P=.003) were prognostic for higher rates of LF. Risk of VCF was 5.5%, 7.6%, and 10.7% at 6, 12, and 24 months, respectively, and was similar between cohorts (P=.573). Spinal malalignment (P < .001), baseline VCF (P=.003), junctional spine location (P=.030), and greater minimum dose to 90% of planning target volume were prognostic for higher rates of VCF. Dose escalation to 28 Gy in 2 daily fractions was associated with improved local control without increasing the risk of VCF. The 2-year local control rates are consistent with those predicted by the HypofractionatedTreatment Effects in the Clinic spine tumor control probability model, and these data will inform a proposed dose escalation randomized trial.