Purpose: To highlight the value of IROC’s on-site proton audits and key findings.
Methods: IROC has performed 38 on-site measurement-based peer reviews of proton centers participating in NCI-funded clinical trials. The reviews covered beam calibration, lateral and depth measurements, mechanical checks, treatment planning, and quality assurance (QA) practices. The CT Number to Relative Linear Stopping Power (RLSP) conversion was tested with a phantom with known RLSP values, and by comparing the institution’s curve to other proton centers.
Results: Institutions received an average of 3 [1, 8] recommendations for practice improvements. The number of deficiencies did not decrease over time, highlighting the continued need for this type of peer review. The most common deficiencies were for QA (97% of centers), CT Number to RLSP conversion (50%), and treatment planning (45%). 32% of institutions failed at least one lateral beam profile measurement (<90% of pixels passing 3%/3mm, 10% threshold), despite passing internal QA measurements. These failures occurred for several different plan configurations (large, small, shallow, and deep targets), and at different depths in the beam path (proximal to target, central, and distal). CT Number to RLSP conversion curves showed deviations at low, mid, and high CT Numbers. This test highlights areas of inconsistency between proton centers, with many centers falling outside 2σ of the mean curve of their peers. All deficiencies from the peer review were discussed with the institutions, and many implemented dosimetric, treatment planning, and practice changes to improve the accuracy of their system and consistency with other institutions.
Conclusion: This peer review program has been integral in confirming comparability and consistency across proton centers for clinical trials, minimizing deviations for outcomes data.
Funding Support, Disclosures, and Conflict of Interest: NCI grant CA180803