2635 - Objective Evaluation Using Analytical metrics for Adaptive Prostate Radiotherapy and Synthetic CT Dose-of-the-Day

Purpose/Objective(s):

Adaptive radiotherapy (ART) decisions in prostate cancer are often based on subjective visual evaluation of daily CBCTs, with no validated, objective metrics to determine which patients & fractions require adaptation. This study addresses gap by using Elekta Analytics to compute quantitative indices,Robustness Index (RI), Dose-Filling Elasticity (DFE), Sparing-Coverage Efficiency (SCE) from CBCT datasets, relating them to synthetic CT (sCT)–based dose recalculation. As CBCT Hounsfield Unit inaccuracies & scatter can compromise direct dose computation, sCT conversion was employed to enable reliable per-fraction dose estimation.

Materials/Methods:

We retrospectively analyzed 150 CBCT datasets from six prostate cancer patients (25fx/ patient). Each fraction, reference CT underwent deformable registration to the corresponding CBCT for contour propagation. Each CBCT was converted to sCT using bulk density overrides, original treatment plan was recalculated on sCT to estimate delivered dose & quantify interfraction dosimetric variation.

Analytical tool ProKnow computed patient indices-

RI: coefficient of variation CV = SD/mean across fractions.

DFE: per patient regression slope of change in PTV coverage per 1% change in OAR high-dose volume

SCE: per fraction metric quantifies how efficiently PTV coverage is achieved relative to high dose delivered to rectum & bladder.

Results:

Results of 150 fractions per six patients

RI: The cohort median RI-0.04 (bootstrap 95% CI 0.015–0.149; range 0.01–0.197), implies heterogeneous interfraction stability across patients. Low RI -consistent anatomy & preparation; high RI -high variability

DFE:

Bladder DFE: median 0.0075 (IQR 0.0035–0.055); sign test vs 0: p = 0.219 (5+ve, 1-ve). Slopes were small & near zero, suggesting limited sensitivity of PTV coverage to changes in bladder volume

Rectum DFE: median 0.0415 (IQR -0.0138–0.3975); sign test vs 0: p = 1.00 (3+ve, 3-ve). Rectum DFE exhibited greater dispersion & mixed directionality, indicating patient-specific trade-offs. Large-magnitude negative slopes (e.g., -0.212) indicated PTV coverage decreased as rectal high-dose volume increased. (Negative DFE -detriment.)

SCE: Per-patient summaries & per-fraction scores (higher = efficient PTV coverage per unit OAR high-dose) median 58.65 (IQR 38.2–67.63; range 27.07–71.3). Low SCE -recurrent suboptimal coverage–sparing balance, benefit from plan adaptation; high SCE - efficient sparing & maintained coverage.

Conclusion:

The study presents objective ART decision-making framework by integrating sCT-based dose recalculation with ProKnow-derived indices. sCT conversion mitigated CBCT dose limitations & enabled reliable dose estimation. Based on these, we propose patient-specific adaptive thresholds initialized from the first 3–5 fractions for plan adaptation. This aligns with existing evidence on prostate interfraction variability & will be prospectively validated to establish objective ART triggers in future.