Five years of clinical application of independent Monte Carlo-based patient-specific quality assurance at Maastro Proton Therapy Center

Introduction At the Maastro Proton Therapy Center in Maastricht, patient-specific quality assurance (PSQA) based on independent GPU-accelerated Monte Carlo (MC) dose calculations fully replaces conventional measurements. Measurement-based PSQA is time- and resource-intensive and has limited sensitivity to clinically relevant errors. We developed a fully automated MC-based pipeline integrating two clinical workflows using the fast MC code FRED. The system is fully operational, and automated verification reports are used in daily clinical practice. Methods The first workflow performs pre-treatment dose recalculation in FRED using the original clinical treatment plan. The second workflow combines FRED with machine log files to verify the delivered dose. Both workflows generate automated verification reports for clinical review. Results The workflow has been routinely used for over five years, providing robust 3D dosimetric verification in heterogeneous patient anatomies. More than 6000 pre-treatment plans and 3272 log file-based PSQA cases were analyzed, corresponding to a reduction of 4090 h of routine QA. The pipeline identified all true negatives and detected two pre-treatment failures related to planning issues that would have been missed by conventional measurement-based PSQA. No false positives or false negatives were observed, demonstrating high sensitivity and reliability. Conclusions The MC recalculation pipeline provides an efficient PSQA solution for proton therapy. Five years of experience confirm that measurement-free MC-based PSQA is a viable and superior alternative, offering comprehensive 3D verification and early error detection while reducing the clinical workload. This work may serve as a practical reference for implementation at other proton therapy centers.