Five year results of an IMRT QA programme

Full text: Introduction. Intensity modulated radiation therapy (IMRT) is a treatment technique that delivers nonuniform fluences for each radiation portal with the aim of shaping the dose distribution in the patient. These IMRT fluences are usually created by using a multi-leaf collimator (MLC) that moves during the radiation delivery either continuously or in a 'step and shoot' fashion. This complex beam delivery relies heavily upon the linac's ability to move the MLC accurately and on the treatment planning system's ability to calculate the required number of monitor units to deliver a given patient dose correctly. At the McGill University Health Centre we have implemented a quality assurance (QA) programme for IMRT to address both of these issues. As of December 2005, approximately 300 patients have undergone IMRT treatments at our centre. This work summarizes our five year QA results and discusses some of the difficulties encountered during that period and presents our current IMRT QA guidelines. Methods and materials MLC mechanical QA is carried out at quarterly and annual intervals. Quarterly tests involve delivering a series of static and dynamic MLC test patterns to Kodak X-Omat V films placed at the isocentre under full buildup conditions. Specific test patterns are used to assess leaf positioning accuracy, leaf acceleration and deceleration, and in-plane and cross-plane skew. Annual tests include MLC centering with collimator rotation, head sag, measuring inter-leaf, intra-leaf and abutting transmissions. The patient specific IMRT QA programme aims to verify that the treatment plan can be delivered accurately and that the calculated number of monitor units will result in the calculated dose. Due to the difficulties of in vivo dosimetry methods, measurements are carried out in solid water phantoms using the actual beam fluences destined for the patient treatment. A single point measurement is carried out using cylindrical Farmer type ionization chambers. A planar dose map is obtained using Kodak EDR2 film. The ionization chamber measurement point is selected to be in a low gradient region of the dose distribution. The measurements are compared with the results of a treatment plan generated by applying the patient beam fluences to a solid water phantom. The equipment used (Farmer chambers, rectangular solid water phantoms, and radiographic film) is the same equipment that is used routinely for other dosimetric uses in our department. Results. The results of mechanical QA tests are assessed qualitatively. Additionally, log files that record MLC motions during beam delivery are verified to determine leaf position accuracy in any situation requiring further investigation. To date, no significant deviations have been observed on any of these tests on four linacs used for IMRT. Our patient-specific ionization chamber QA results suggest that 95% of our measurements are within ±4% of the pre-calculated values. Additionally, the standard deviation of the difference between ionization chamber measurements and calculated values has decreased with experience. Several factors have contributed to this improvement. Specifically, we now account for daily linac output variations (less than ±2%) and have on several occasions adjusted the treatment planning calibration factor, a user determined parameter accounting for differences between planning and treatment delivery. Our current ionization chamber QA guidelines are: 1) Absolute point dose measurements must be within 5% of the planned values otherwise the QA procedure is repeated or redone selecting a more suitable point of measurement. If the subsequent QA measurement also reveals a difference greater than 5%, the patient is re-planned in order to achieve a more deliverable set of fluences. 2) Ionization chamber QA data are reviewed once a year. 3) The treatment planning calibration factor is adjusted if the mean discrepancy changes by more than 1%. The results of our film dosimetry QA are also evaluated qualitatively (isodose shape) and quantitatively (distance to agreement [DTA]). Our film QA policy is to accept a 3 mm DTA or ±5% dose difference in the high dose region and 5 mm DTA or ±10% dose difference in the low dose region. Conclusions. A QA programme for IMRT delivery can be implemented using fairly common tools available to the medical physicist (radiographic film, Farmer type ionization chambers, solid phantoms). Together with appropriate methodology, this can lead to acceptable results. However, a continued review of QA results with time is essential. (author)