KEY POINTS
- The development cohort included 11 head-and-neck patients and 312 fractions treated on Halcyon-E. Daily exit-dose fluence maps were compared with the first treatment fraction, while changes in external body volume were calculated from CBCT.
- Increasing anatomical volume change correlated strongly with worsening gamma agreement. Kendall’s τ was −0.83 for 1%/1 mm and −0.75 for 2%/2 mm (both p<0.001), while looser criteria showed weaker correlations.
- Median 1%/1-mm passing rate declined from approximately 87% with 0–1% volume change to below 50% when volume change exceeded 5%.
- The proposed decision framework used 1%/1 mm below 80% as the primary alert and 2%/2 mm below 90% as confirmation, particularly when external-volume change exceeded 5%.
- Retrospective validation included 20 patients and 618 fractions. The workflow identified all six patients who underwent clinical replanning, with alerts generally preceding or coinciding with the clinical decision.
- The primary alert occurred, on average, around fraction 6, compared with approximately fraction 15 for the confirmatory threshold. In one patient, replanning restored gamma passing rates from critically low values to above 97%.
- Several of the 14 patients who were not replanned also crossed one or both thresholds, including minimum 1%/1-mm passing rates of 16.6% and 18.9%. Because these patients did not undergo systematic repeat planning CT and dose reconstruction, the alerts cannot be classified confidently as true positives or false positives.
CLINICAL TAKEAWAY
Halcyon exit-dose fluence can provide an automated early-warning signal for anatomical change and help prioritize patients for CBCT review or repeat simulation. It should not be used as an automatic replanning rule because global body-volume change and two-dimensional gamma analysis cannot determine target or organ-at-risk dose.