Stereotactic body radiation therapy for benign spine tumors: is dose de-escalation appropriate?

OBJECTIVE Akin to the nonoperative management of benign intracranial tumors, stereotactic body radiation therapy (SBRT) has emerged as a nonoperative treatment option for noninfiltrative primary spine tumors such as meningioma and schwannoma. The majority of initial series used higher doses of 16-24 Gy in 1-3 fractions. The authors hypothesized that lower doses (such as 12-13 Gy in 1 fraction) might provide an efficacy similar to that found with the dose de-escalation commonly used for intracranial radiosurgery to treat acoustic neuroma or meningioma and with a lower risk of toxicity. METHODS The authors identified 38 patients in a prospectively maintained institutional radiosurgery database who were treated with definitive SBRT for a total of 47 benign primary spine tumors between 2004 and 2016. SBRT consisted of 9-21 Gy in 1-3 fractions using the CyberKnife (n = 11 [23%]), Synergy S (n = 21 [45%]), or TrueBeam (n = 15 [32%]) radiosurgery platform. For a comparison of SBRT doses, patients were dichotomized into 1 of 2 groups (low-dose or high-dose SBRT) using a cutoff biologically effective dose (BED10Gy) of 30 Gy. Tumor control was calculated from the date of SBRT to the last follow-up using Kaplan-Meier survival analysis, with comparisons between groups completed using a log-rank method. To account for potential indication bias, a propensity score analysis was completed based on the conditional probabilities of SBRT dose selection. Toxicity was graded using Common Terminology Criteria for Adverse Events version 4.0 with a focus on grade 3+ toxicity and the incidence of pain flare. RESULTS For the 38 patients, the most common histological findings were meningioma (15 patients), schwannoma (13 patients), and hemangioblastoma (7 patients). The median age at SBRT was 58 years (range 25-91 years). The 47 treated lesions were located in the cervical (n = 18), thoracic (n = 19), or lumbosacral (n = 10) spine. Five (11%) lesions were lost to follow-up after SBRT. The median follow-up duration for the remaining 42 lesions was 54 months (range 1.2-133 months). Six (16%) patients (with a total of 8 lesions) experienced pain flare after SBRT; no significant predictor of pain flare was identified. No grade 3+ acute- or late-onset complication was noted. The 5-year local control rate was 76% (95% CI 61%-91%). No significant difference in local control according to dose, fractionation, previous radiation, surgery, tumor histology, age, treatment platform, planning target volume, or spine level treated was found. The 5-year local control rates for low- and high-dose treatments were 73% (95% CI 53%-93%) and 83% (95% CI 61%-100%) (p = 0.52). In propensity score-adjusted multivariable analysis, no difference in local control was identified (HR 0.30, 95% CI 0.02-5.40; p = 0.41). CONCLUSIONS Long-term follow-up of patients treated with SBRT for benign spinal lesions revealed no significant difference between low-dose (BED10Gy <= 30) and high-dose SBRT in local control, pain-flare rate, or long-term toxicity.