KEY POINTS
- This retrospective single-institution study from the University of Alabama at Birmingham included 496 patients with 2,460 previously untreated brain metastases measuring <2 cm, treated between 2015 and 2022. Of the analyzed lesions, 1,598 received single-fraction stereotactic radiosurgery (SRS) and 862 received fractionated SRS (fSRS). Treatment fractionation was typically determined by the largest or most critically located metastasis within a multi-target treatment plan, while larger lesions themselves were excluded from the analysis. Median radiographic follow-up was seven months.
- All treatments used frameless, single-isocenter, multitarget VMAT with zero CTV-to-PTV margin. Single-fraction prescriptions included 16, 18, or 20 Gy, whereas fractionated regimens included 9 Gy × 3, 6 Gy × 5, or 5 Gy × 5. Plans required 99–100% GTV coverage at prescription dose and used orthogonal kV imaging, CBCT, and intrafraction surface monitoring. Median lesion diameter was slightly larger in the fSRS group (7.4 versus 5.5 mm; p<0.001), illustrating nonrandom treatment selection.
- One-year local control was nearly identical between the two treatment approaches. Local control reached 96.1% with SRS (95% CI 94.8–97.4%) and 96.6% with fSRS (95% CI 95.0–98.2%), with no statistically significant difference (p=0.326). Across the entire cohort, one- and two-year local control were 96.2% and 94.0%, respectively. In multivariable shared-frailty analysis, fSRS was not significantly associated with local failure (HR 1.23, 95% CI 0.62–2.45; p=0.55), but this does not demonstrate treatment equivalence.
- Local control varied across individual fractionation schedules. At one year, control was 94.4% after 18 Gy × 1, 96.6% after 20 Gy × 1, 99.3% after 9 Gy × 3, and 95.7% after 6 Gy × 5. The corresponding BED10 values were 50.4, 60.0, 51.3, and 48.0 Gy, respectively. Although 9 Gy × 3 produced the highest observed control estimate among common regimens, dose selection was nonrandomized, and no direct comparative efficacy advantage can be established from these descriptive results.
- Severe CNS toxicity was uncommon and did not significantly differ by fractionation. The overall grade ≥3 toxicity estimate was 1.2% (95% CI 0.7–1.8%), with no significant difference across regimens (p=0.258). There were 21 grade 3 events, comprising 11 new-onset seizures and ten radiologically diagnosed radionecrosis events, plus three grade 4 events, all involving urgent surgical resection with pathological radionecrosis. Limited follow-up reduces confidence in the detection of delayed radiation necrosis.
- Tumor histology and diameter were stronger predictors of local failure than fractionation. Melanoma metastases had significantly greater failure risk than lung cancer metastases (adjusted HR 8.16, 95% CI 2.91–22.90; p<0.001), while each additional centimeter of tumor diameter was associated with worse control (HR 2.88, 95% CI 1.61–5.14; p<0.001). The patient-level shared-frailty model identified substantial clustering of lesion failures within individuals (Kendall's τ=0.72), suggesting that disease- and patient-level factors contribute to outcomes beyond individual lesion characteristics.
- Using a single fractionation prescription may simplify treatment planning, but efficiency was not directly measured in this cohort. A unified plan may avoid separate optimization, independent QA, and additional treatment delivery for smaller lesions. However, the study did not prospectively compare planning time, treatment-room duration, or workload between single- and multiple-prescription approaches. Its retrospective design, short median imaging follow-up, mixed histologies, and absence of randomized assignment mean that the findings support feasibility rather than formal noninferiority.
CLINICAL TAKEAWAY
Fractionated SRS achieved one-year local control comparable to single-fraction treatment for brain metastases smaller than 2 cm, supporting the feasibility of using a unified fractionation schedule when multiple lesions of different sizes are treated together. This strategy may simplify clinical workflows, but efficiency gains were not directly measured, and the study cannot establish equivalent late toxicity or long-term control. Prospective randomized comparisons remain necessary before a single-prescription approach can be considered interchangeable with lesion-specific fractionation.