HyperSight CBCT approached simulation-CT dosimetry for offline adaptive breast radiotherapy

HyperSight reduced PTV dose-recalculation deviations versus Halcyon 3.1 while using a lower imaging dose in 12 breast radiotherapy patients.

KEY POINTS

  • The retrospective analysis included 12 women after breast-conserving surgery—seven right- and five left-sided cases—who had clinically acquired images on both Halcyon 3.1 and Halcyon-HyperSight within ±1 day. All received whole-breast RT 40 Gy/16 fractions plus 10 Gy/4-fraction boost, without regional nodal irradiation.
  • HyperSight combines rapid acquisition, an expanded field of view and improved reconstruction/HU calibration. Its breast protocol used an average CTDIvol of 3.39 mGy, compared with 6.04 mGy for Halcyon 3.1 and 7.53 mGy for the reference simulation CT.
  • Phantom HU performance was substantially closer to simulation CT. For solid water, mean HU was 6.9±3.1 on simulation CT, 6.4±4.1 on HyperSight and −42.5±47.5 on Halcyon 3.1, illustrating the bias and variability of the older system.
  • CBCT-based plan recalculation likewise approached simulation CT more closely. Paddick conformity index was 0.903±0.034 on CT, 0.883±0.053 on HyperSight and 0.840±0.066 on Halcyon 3.1; all pairwise comparisons were statistically significant.
  • Relative to simulation CT, HyperSight changed PTV D95 by only −21 cGy, D98 by −51 cGy and V95%Rx by −0.5%. Corresponding Halcyon 3.1 deviations were considerably larger at −157 cGy, −353 cGy and −3.0%, respectively.
  • No statistically significant recalculation differences were identified for ipsilateral lung, heart, contralateral breast or maximum dose. Anatomically, a volume present in the original PTV but absent from the adaptive PTV—a potential unnecessary high-dose region if the original plan were maintained—was found in 11/12 patients, averaging 41.2 cm³.
  • The analysis remains exploratory: only 12 patients were included, scans were acquired on different days, treatment was free-breathing, and the CBCT-based dose calculation lacked independent verification against repeat CT or measurement. No prospective endpoint demonstrated that HyperSight-guided adaptation improves toxicity or tumor control.

CLINICAL TAKEAWAY

HyperSight appears substantially more suitable than older Halcyon CBCT for quantitative offline breast adaptation, particularly when anatomy changes during a multiweek course. It is promising as a way to identify patients who may need replanning, but this study does not yet define when adaptation should be triggered or prove that repeat CT can routinely be omitted.

SOURCE

Frontiers in Oncology