The study, led by Brock J. Gjesdal, Samuel Adida, and Suchet Taori under the senior authorship of neurosurgeon Peter C. Gerszten, drew from a prospectively maintained database of consecutive patients treated between 2002 and 2024. From 600 patients with 936 spinal metastases in the database, the team identified 182 patients aged 70 or older; after excluding those without baseline imaging, 161 patients with 208 metastases formed the final cohort. The median age was 75 years, with 30 patients over 80 and six over 85. Seventy-two percent of the metastases had already been irradiated with conventional external beam radiotherapy, making this predominantly a reirradiation population, and 76 percent received their entire SBRT dose in a single fraction.

The technical sophistication of SBRT is what sets it apart from conventional palliative radiotherapy. While conventional external beam radiotherapy delivers relatively uniform doses across a broad field, SBRT exploits steep dose gradients to concentrate ablative radiation doses within the tumor while respecting the tolerance of the spinal cord. In this cohort, the median single-fraction margin dose was 16 Gy, with a range of 12 to 18 Gy, and multi-fraction regimens delivered a median total margin dose of 24 Gy in two to five fractions. Dose was most commonly prescribed to the 80 percent isodose line, and a planning target volume margin of zero to two millimeters was applied while excluding the spinal cord or cauda equina, which were contoured as the spinal canal or thecal sac plus one millimeter. Treatment was delivered using CyberKnife, Elekta Synergy, and Varian TrueBeam platforms, with image guidance achieved through six-dimensional tracking, daily cone-beam computed tomography, or ExacTrac. Magnetic resonance imaging was registered to the simulation CT for target delineation, and CT myelography was used when metal artifact limited visualization of the neural elements.

The outcomes challenge the assumption that elderly patients are poor candidates for aggressive local therapy. After a median follow-up of nine months, the cumulative incidence rates of local failure were 6 percent at six months, 9 percent at twelve months, 9 percent at twenty-four months, and 10 percent at thirty-six months, with death treated as a competing risk. Vertebral compression fracture rates were similarly low: 4 percent at six months, 7 percent at twelve months, 8 percent at twenty-four months, and 9 percent at thirty-six months. Complete or partial pain response followed 61 of 148 evaluable treatments, representing 41 percent of evaluable cases among 186 symptomatic metastases. Grade one or two radiation toxicity occurred in 18 percent of patients, most commonly pain flare, which was managed with corticosteroids in most cases. Critically, no radiation myelopathy or radiculopathy occurred, and apart from vertebral compression fracture, no grade three or higher toxicity was observed.

But the most scientifically compelling finding is what the researchers did not find. Using Fine-Gray competing-risk models with cluster-robust variance to account for multiple lesions per patient, the team analyzed age as a continuous variable, at clinically relevant thresholds of 75 and 80 years, and with exploratory maximally selected rank statistics that searched for the age cutoff most strongly associated with each outcome. No statistically significant association between age and local failure or vertebral compression fracture was detected under any parameterization. Restricted cubic spline analysis with knots at the fifth, thirty-fifth, sixty-fifth, and ninety-fifth age percentiles confirmed the absence of any overall association or nonlinearity, with confidence intervals obtained from 300 bootstrap replicates resampling patients.

What did predict failure? Tumor biology and spinal mechanics, not the patient’s birth year. On multivariable analysis, gross tumor volume greater than 37 cubic centimeters was independently associated with local failure, with a subdistribution hazard ratio of 6.27, and the Spinal Instability Neoplastic Score, analyzed per point, carried a subdistribution hazard ratio of 1.23. For vertebral compression fracture, only SINS remained independently associated, with a per-point subdistribution hazard ratio of 1.19. These findings suggest that the size of the tumor and the mechanical integrity of the spine, rather than the age of the patient, are the true determinants of treatment success and fracture risk.

Age did matter for one outcome: overall survival. Each additional year of age carried a hazard ratio of 1.05 for death, and Karnofsky Performance Status greater than 70 was independently associated with superior survival, with a hazard ratio of 0.56. Prior conventional radiotherapy was associated with shorter survival, likely reflecting a longer and more advanced disease course rather than harm from reirradiation. Interestingly, the effect of age was not proportional over time: age was not significantly associated with survival during the first twelve months after SBRT but became strongly associated thereafter, with a hazard ratio of 1.14 per year beyond the first year.

The study also revealed a temporal trend in fracture risk that raises important questions about treatment evolution. Vertebral compression fracture risk increased with later treatment year, with a subdistribution hazard ratio of 1.56 per five years, and was higher after linear accelerator treatment compared with robotic CyberKnife delivery. All treatments after 2016 were delivered with gantry-based linear accelerators, and most were multi-fraction, making it impossible to reliably separate the effects of treatment year, platform, and fractionation. Prior or adjuvant immunotherapy also became more common over time, from 7 of 133 treatments in the early period to 16 of 46 in the recent era, potentially reflecting changing patient selection as systemic therapies improved.

Eleven of the 17 vertebral compression fractures required intervention, either percutaneous cement augmentation or open surgical decompression with or without instrumented fusion. This is not a trivial complication, and the researchers acknowledge that most fractures were clinically significant. However, they also note that because vertebral body collapse contributes to the SINS classification and progression of a baseline fracture was counted as a fracture event, the association between spinal instability and fracture may partly reflect pre-existing structural compromise rather than pure treatment effect. The researchers suggest that antiresorptive therapy with bisphosphonates or denosumab prior to SBRT may represent an additional strategy for mitigating fracture risk, and prophylactic cement augmentation with kyphoplasty prior to SBRT has been shown to be safe without compromising treatment planning or delivery.

The findings arrive at a pivotal moment. The incidence of spinal metastases is increasing, particularly among the elderly, and advances in targeted and immunotherapies are prolonging survival, making durable local control increasingly important for maintaining quality of life and functional independence. Yet elderly patients remain underrepresented in randomized controlled trials of spine radiation therapy and may be preferentially treated with conventional radiotherapy based on concerns about bone quality and fracture risk. This study, one of the largest to specifically characterize outcomes in patients aged 70 and older, suggests those concerns may be overstated for appropriately selected patients. The researchers caution that their single-institution retrospective design is subject to selection bias, that the median follow-up of nine months reflects the limited survival of this population, and that no comorbidity index, frailty measure, or geriatric assessment was available. Nevertheless, the consistency of the multivariable associations across all sensitivity analyses, including cause-specific Cox models, Firth-penalized models, and multiple imputation of missing performance status, strengthens the central conclusion: treatment selection for spinal metastases should be guided by tumor characteristics, spinal stability, and overall prognosis rather than chronological age alone.

Subject of Research: Stereotactic body radiation therapy in elderly patients with spinal metastases

Article Title: Stereotactic body radiation therapy in elderly patients with spinal metastases

Article References: Gjesdal, B. J., Adida, S., Taori, S., Zureick, A. H., Choi, S., Flickinger, J. C., Zinn, P. O., Bayley, J. C., Gerszten, P. C., & Sefcik, R. K. (2026). Stereotactic body radiation therapy in elderly patients with spinal metastases. Journal of Neuro-Oncology, 180(2), Article 22. https://doi.org/10.1007/s11060-026-05838-3

Image Credits: AI Generated

DOI: 10.1007/s11060-026-05838-3

Keywords: SBRT, spinal metastases, elderly patients, radiation oncology, vertebral compression fracture, spine oncology, Stereotactic, body, radiation, therapy, elderly, patients