KEY POINTS
- This single-center dosimetric study evaluated 276 single-target, single-isocenter SRS plans for brain metastases, comprising 178 idealized model plans and 98 clinical plans. All used linac-based treatment with a 20 Gy prescription, allowing the volume receiving 10 Gy to represent the 50% prescription isodose volume (V50%).
- The planning approach used a TrueBeam STx with 6-MV flattening-filter-free beams, a high-definition multileaf collimator, and one full plus three noncoplanar half arcs. Model plans were generated for PTVs of 0.3–4.3 cm³, excluding cases with nearby critical organs or prior irradiation to establish an achievable low-spillage reference.
- In the model cohort, V50% correlated strongly with PTV (r=0.994; p<0.001), following a power-law relationship. The derived mean reference was V50% = 3.71 × PTV^0.83, while the conservative ideal-planning 95th-percentile reference was V50% = 4.10 × PTV^0.83, with volumes expressed in cm³.
- Clinical quantile regression produced a simplified upper threshold of V50% = 5.00 × PTV^0.75. Unlike fixed radionecrosis-related volume limits, this threshold accounts for the expected increase in intermediate-dose exposure with target volume and was intended to flag unusually high spillage for review rather than serve as an absolute toxicity constraint.
- Ten of 98 clinical plans (10%) exceeded the volume-dependent threshold. Three exceedances were considered clinically justified because of target asymmetry or proximity to previously irradiated tissue, while seven plans were identified as potentially improvable.
- Replanning the seven unexplained outliers reduced V50% by a median 14% (range 7–29%) while preserving target coverage. All seven revised plans fell below the proposed clinical upper threshold, demonstrating that the benchmark could identify avoidable intermediate-dose exposure not necessarily apparent from conventional acceptance criteria.
- The model applies to single-target linac SRS with comparable hardware, beam geometry, normalization and prescription, and should not be transferred directly to multiple-target treatments, other prescription doses, or different platforms. No patient toxicity or radionecrosis outcomes were analyzed, so the threshold is a plan-quality benchmark rather than a validated clinical risk limit.
CLINICAL TAKEAWAY
A PTV-dependent V50% benchmark can complement conformity and gradient indices by identifying single-target brain metastasis SRS plans with potentially avoidable intermediate-dose spillage. The proposed threshold is useful as a trigger for structured plan review, but it does not replace established normal-brain dose constraints or demonstrate that the observed dosimetric reductions improve clinical outcomes.