P18. Increase in patient throughput at Groote Schuur Hospital through the implementation of portal dosimetry for RapidArc patients

Abstract

Introduction RapidArc treatment was introduced at Groote Schuur Hospital (GSH) in 2013, but the programme was limited by time requirements for patient-specific Quality Assurance (QA). QA equipment consisted of a cylindrical phantom for absolute dosimetry, PTW 0.016cc pinpoint chamber, PTW Octavius and film dosimetry (X-Omat V). Portal dosimetry was implemented on the Varian Unique in May 2015 to allow faster processing of patient QA, while maintaining the required accuracy. Materials and methods Portal dosimetry was commissioned with the Varian pre-configuration package, including corrections for backscatter of the imager arm. Verification fluences were measured for each RapidArc field, and compared online to the Eclipse calculated fluence. Results were reported for composite fields. Isocentric stability for the linac/EPID combination was found to be within the 1.5 mm accuracy reported by other centres. To account for this uncertainty in position, a Gamma Agreement Index (GAI) of 3% dose-difference and 1.5 mm distance-to-agreement was selected, testing an overall accuracy of 3%/3 mm. The analysis area included 1 cm from the Complete Irradiated Area Outline (CIAO). Automatic alignment and normalization to minimize differences were applied. Absolute dose verification per field was done with a pinpoint chamber in a cylindrical water phantom, at stable dose points, with shifts applied as required. Results Pre-treatment QA of RapidArc plans resulted in a GAI passrate of 98.9 ± 1.2% on the first 80 composite arc plans (190 arcs). Absolute dosimetry passed with an average difference of 1.6 ± 1.7%. From June 2015 to May 2016 80 new RapidArc patients were started, with a total of 87 plans, compared to only 35 patients from June 2014 to May 2015. Conclusion The implementation of portal dosimetry improved patient throughput, optimized use of Physics’ and linac time, both for QA and treatment, while ensuring accurate delivery of RapidArc treatment. It did however increase planning time. RapidArc treatment was introduced at Groote Schuur Hospital (GSH) in 2013, but the programme was limited by time requirements for patient-specific Quality Assurance (QA). QA equipment consisted of a cylindrical phantom for absolute dosimetry, PTW 0.016cc pinpoint chamber, PTW Octavius and film dosimetry (X-Omat V). Portal dosimetry was implemented on the Varian Unique in May 2015 to allow faster processing of patient QA, while maintaining the required accuracy. Portal dosimetry was commissioned with the Varian pre-configuration package, including corrections for backscatter of the imager arm. Verification fluences were measured for each RapidArc field, and compared online to the Eclipse calculated fluence. Results were reported for composite fields. Isocentric stability for the linac/EPID combination was found to be within the 1.5 mm accuracy reported by other centres. To account for this uncertainty in position, a Gamma Agreement Index (GAI) of 3% dose-difference and 1.5 mm distance-to-agreement was selected, testing an overall accuracy of 3%/3 mm. The analysis area included 1 cm from the Complete Irradiated Area Outline (CIAO). Automatic alignment and normalization to minimize differences were applied. Absolute dose verification per field was done with a pinpoint chamber in a cylindrical water phantom, at stable dose points, with shifts applied as required. Pre-treatment QA of RapidArc plans resulted in a GAI passrate of 98.9 ± 1.2% on the first 80 composite arc plans (190 arcs). Absolute dosimetry passed with an average difference of 1.6 ± 1.7%. From June 2015 to May 2016 80 new RapidArc patients were started, with a total of 87 plans, compared to only 35 patients from June 2014 to May 2015. The implementation of portal dosimetry improved patient throughput, optimized use of Physics’ and linac time, both for QA and treatment, while ensuring accurate delivery of RapidArc treatment. It did however increase planning time.