KEY POINTS
- This single-institution positioning and dosimetric study analyzed 120 frameless single-isocenter multi-target stereotactic radiosurgery treatments for brain metastases and 294 couch kicks. Prescription doses ranged from 18 to 24 Gy.
- Deviations exceeding 0.5 mm or 0.5° triggered six-degree-of-freedom repositioning. Intrafractional motion exceeded these tolerances after 66% of couch kicks.
- Correction reduced mean translational vector error from 0.71 ± 0.38 mm to 0.33 ± 0.19 mm and rotational vector error from 0.58 ± 0.34° to 0.38 ± 0.18°.
- In 15 selected plans containing 90 metastases, simulated errors of 0.5 mm/0.5° reduced D98% by 7.6% ± 2.5% and conformity by 9.5% ± 4.9%. At 1 mm/1°, the reductions reached 17.3% ± 4.8% and 26.6% ± 9.2%, respectively.
- Lesions measuring 0.4 cm³ or less were most sensitive. Target volume was the principal predictor of coverage loss, while distance from the isocenter was not significant in scenarios involving rotational error.
CLINICAL TAKEAWAY
For frameless noncoplanar single-isocenter multi-target stereotactic radiosurgery, initial setup verification alone may be insufficient; the findings support repeat imaging and six-degree-of-freedom correction after couch rotations, particularly for very small lesions. However, motion was sampled only after couch kicks, setup errors were simulated uniformly, and clinical control and toxicity were not assessed, making this technical implementation evidence rather than clinical outcome evidence.