KEY POINTS
- The replanning study included 22 patients with 154 intact brain metastases, with 5–10 lesions per patient, maximum lesion diameter ≤2 cm and cumulative planning target volume of 4–20 cm³. Lesions within 5 mm of major central organs at risk were excluded.
- Three single-isocenter, three-arc strategies were compared: a geometric-centre isocenter with composite-target optimization, a target-anchored isocenter with composite optimization, and the same target-anchored isocenter with each metastasis optimized individually. All plans prescribed 18 Gy in one fraction.
- Target coverage remained essentially identical: planning target volume V95 was 99.50%, 99.52% and 99.37%, respectively (p=0.202). Conformity and Radiation Therapy Oncology Group coverage indices also did not significantly differ.
- Individual-target optimization deliberately increased intratumoral dose: maximum target dose rose from 20.26 to 21.14 to 22.41 Gy (p<0.001) across the three strategies, while the intermediate-dose gradient improved, with R50 falling from 13.43 to 12.07 to 10.31 and Paddick gradient index from 12.91 to 12.10 to 9.41.
- Normal-brain V12 fell from 42.27 cm³ with the geometric-centre approach to 35.53 cm³ with individual optimization. Normal-brain median dose similarly decreased from 370 to 322 cGy.
- Maximum hippocampal dose fell from 507 to 305 cGy, brainstem maximum dose from 854 to 662 cGy, and spinal-canal maximum dose from 640 to 508 cGy with the individually optimized approach; several cochlear metrics also improved.
- The trade-off was greater modulation: monitor units increased from 6,014 to 6,495 to 7,269 (p<0.001). Importantly, placing the isocenter on one target increased mean target-to-isocenter distance from 5.74 to 6.33 cm, yet the study did not test the resulting sensitivity to rotational setup error, a central concern in single-isocenter treatment.
CLINICAL TAKEAWAY
Individual optimization of each metastasis may reduce dose bridging and normal-brain exposure in multi-target radiosurgery without sacrificing nominal target coverage. Before clinical adoption, the strategy needs delivery, rotational-error and outcome validation because improved static dosimetry does not guarantee more robust treatment.