A Cost-Effective End-to-End Credentialing Test for Simulation-Omitted Direct-to-Unit Adaptive Radiotherapy
Abstract
Purpose
Direct-to-Unit (simulation-omitted) radiotherapy enables treatment planning directly from diagnostic imaging, eliminating CT simulation and supporting same-day treatment for time-sensitive cases. These workflows rely on adaptive radiotherapy (ART) platforms to perform online plan re-optimization to account for CT-number discrepancies, anatomical variation, and setup uncertainty. However, no standardized end-to-end credentialing framework currently exists to verify the accuracy and safety of Direct-to-Unit ART. To address this critical gap, we developed a cost-effective, end-to-end credentialing test using a commercially available phantom, enabling institutions to independently validate Direct-to-Unit ART workflow without requiring expensive, custom-built phantoms, thereby enhancing clinical safety and supporting broader adoption.
Methods
The credentialing test was performed on a Varian Ethos ART system using the CIRS ZEUS phantom (Model 008Z), which includes interchangeable inserts for A16 ionization chamber point-dose measurements and planar film dosimetry. The A16 chamber and EBT4 film were independently cross-calibrated on an Elekta Versa Linac with an ADCL-traceable chamber. A pseudo diagnostic-CT digital phantom was created and intentionally deformed to imitate anatomical variability expected in simulation-omitted workflows. Target structures were expanded by 5mm circumferentially and 1cm superior–inferior to generate a PTV. A pre-plan was created using an institutional lung prescription of 2Gy × 30fractions with an 18-field IMRT. During the credentialing, standard ART workflows were followed. Six fractions were delivered: three scheduled plans and three adaptive plans generated online from CBCTs. Point-dose measurements were compared with Ethos-reported values, and film results were assessed with global gamma analysis (3%/3 mm).
Results
Point-dose differences at the PTV, spinal cord, and liver were within 3% for both plan types. All planar film measurements exceeded a 94% gamma pass rate using 3%/3 mm.
Conclusion
This work establishes a practical, low-cost credentialing methodology for Direct-to-Unit ART workflows. The approach verifies accurate online re-optimization and dose delivery, enabling safe clinical implementation and potential cross-institutional standardization.