Dosimetric comparison of rapidplan and manually optimized plans for hippocampus avoidance whole brain radiation therapy

Abstract

Nowadays, whole-brain radiation therapy (WBRT) is a good method to cure brain metastases. However, it has been noted that whole-brain radiation therapy could be a cause to neurological toxicity which could damage the hippocampus that is an important part of forming long-term memories. The HA-WBRT is a complicated plan depending on planner skills, the researcher would like to simplify this problem using knowledge-based planning called RapidPlan. The objective of this study was to compare the dosimetric difference of RapidPlan and manually optimized plans for the treatment of HA-WBRT. The RapidPlan model was created for HA-WBRT. The PTV and OARs constraints are followed by RTOG 0933 criteria. Twenty cases were used for creating the RapidPlan model and ten cases were used for comparing RapidPlan and manual plans. The treatment prescription to the whole brain PTV was prescribed to 30 Gy. The results showed the PTV optimization using RapidPlan was more effective than manual plans. With the same D90% of PTV between RapidPlan and manual plans, the D98% of RapidPlan and manual plans were 27.69±0.98 Gy and 25.83±1.86 Gy, respectively. Moreover, RapidPlan showed slightly better in CI and HI of PTV than manual plans. The maximum dose to the hippocampus of RapidPlan and manual plans were 15.66±1.29 Gy and 14.42±1.23 Gy, respectively. The doses between the two optimized types were not significantly different (p>0.05). The D100% of hippocampus of RapidPlan and manual plans were 9.25±0.35 Gy and 8.72±0.55 Gy, respectively. On the other hand, RapidPlan can spare more optic nerves than manual plans while manual plans can spare more lenses than RapidPlan.

In conclusion, HA-WBRT in VMAT technique using RapidPlan and a manual plan was able to maintain the radiation dose within the RTOG 0933 constraint. RapidPlan shows better PTV coverage and presents significantly conforms the dose to the target than manually optimized plans. Although the manual plans performed better than RapidPlan in the hippocampus, it was not a significant different. It can be concluded that the RapidPlan can be used in clinical terms with a slight adjustment for well-optimized plans.