Purpose: With the increased commercialnavailability of intensity modulated arc therapy (IMAT)
n comes the need for comprehensive QA programs, covering thendifferent aspects of this newly available
n technology. This manuscript proposes such a program for thenRapidArc (RA) (Varian Medical
n Systems, Palo Alto) IMAT solution.
Methods: The program was developed and testednout for a Millennium120 MLC on iX Clinacs and
n a HighDefinition MLC on a Novalis TX, using a variety ofnmeasurement equipment including Gafchromic
n film, 2D ion chamber arrays (Seven29 and StarCheck, PTW, Freiburg,nGermany) with inclinometer
n and Octavius phantom, the Delta4 systam (ScandiDos, Uppsala,nSweden) and the portal
n imager (EPID). First, a number of complementary machine QA testsnwere developed to monitor the
n correct interplay between the accelerating/decelerating gantry,nthe variable dose rate and the MLC
n position, straining the delivery to the maximum allowed limits.nSecond, a systematic approach to
n the validation of the dose calculation for RA was adopted,nstarting with static gantry and RA specific
n static MLC shapes and gradually moving to dynamic gantry, dynamicnMLC shapes. RA plans
n were then optimized on a series of artificial structures creatednwithin the homogeneous Octavius
n phantom and within a heterogeneous lung phantom. These served thendouble purpose of testing the
n behavior of the optimization algorithm (PRO) as well as thenprecision of the forward dose calculation.
n Finally, patient QA on a series of clinical cases was performednwith different methods. In
n addition to the well established in-phantom QA, we evaluated thenportal dosimetry solution within
n the Varian approach.
Results: For routine machine QA, the “SnookernCue” test on the EPID proved to be the most sensitive
n to overall problem detection. It is also the most practical one.nThe “Twinkle” and “Sunrise”
n tests were useful to obtain well differentiated information on thenindividual treatment delivery components.
n The AAA8.9 dose calculations showed excellent agreement with allncorresponding measurements,
n except in areas where the 2.5 mm fixed fluence resolution wasninsufficient to accurately
n model the tongue and groove effect or the dose through nearlynclosed opposing leafs. Such cases
n benefited from the increased fluence resolution in AAA10.0. In thenclinical RA fields, these effects
n were smeared out spatially and the impact of the fluencenresolution was considerably less pronounced.
n The RA plans on the artificial structure sets demonstrated someninteresting characteristics
n of the PRO8.9 optimizer, such as a sometimes unexpected dependencenon the collimator rotation
n and a suboptimal coverage of targets within lung tissue. Althoughnthe portal dosimetry was successfully
n validated, we are reluctant to use it as a sole means of patientnQA as long as no gantry angle
n information is embedded.
Conclusions: The all-in validation programnallows a systematic approach in monitoring the different
n levels of RA treatments. With the systematic approach comes anbetter understanding of both the
n capabilities and the limits of the used solution. The program cannbe useful for implementation, but
n also for the validation of major upgrades. VC 2011 AmericannAssociation of Physicists in Medicine.
n [DOI: 10.1118/1.3622672]