Session Item

To evaluate and quantify the planning performance of a commercially available automated planning module with a lexicographic optimization-based algorithm in order to increase treatment planning efficiency, achieve high quality radiation therapy treatment plans and reduce inter-planner variability for a random cohort of prostate cancer patients.

An automated planning routine was applied to a sample of 12 previously treated patients. For the first two patients 540 automated plans were generated to refine the optimization goals and determine a planning strategy. For the remaining 10 patients, a total amount of 460 automated plans (AP) for different radiation therapy techniques were generated by the use of the “Plan Explorer” auto planning module (RaySearch Laboratories, Stockholm, Sweden). Clinical objectives were checked for the selected AP candidates and evaluated against the respective clinical manual plan (MP) created in the “Raystation” TPS that were taken as base line. Plan evaluations considered selected dosimetric plan parameters. Target coverage, homogeneity index (HI), conformity index (CI), organs at risk sparing, efficiency of design and planned delivered times were compared between MP and APs. Quality assurance of MP and AP was performed in order to evaluate the dosimetric accuracy of automatically generated VMAT plans.

The lexicographic optimization-based algorithm implemented in Plan Explorer was able to produce plans of clinical quality, meeting target coverage objectives and organs at risk constraints for a random cohort of prostate cancer patients. With equivalent Target Planning Volume (PTV) V95%, D98% , Dmax% , HI and CI, significant improvements in organ at risk (OARs) sparing of the rectum, bladder and intestine were obtained, with mean dose reductions for AP of 4.5% (Range: 7.1% - 2.7%); 3.5% (Range: 5.1% - 1.2%) and 2.7 % (Range: 5.3 – 0.2), respectively, for the mean-organ-dose parameter. Significant differences in the average number of monitor units (MU) were found between MP and AP, depending on the radiation therapy technique. Therefore delivery times can be reduced considering clinical criteria. Not only overall active plan design time for AP was reduced by 55% with respect to MP, but, in addition, resources for AP optimizations could be used outside of normal working hours, allowing to increase production and optimize resource use.

Plan Explorer can automatically generate acceptable clinical treatment plans for prostate cancer with either improved or similar results compared to manually created plans. Time reduction in plan design as well as human & computational resources optimization improves clinical efficiency.