**6. Future directions**

The technology of proton therapy is evolving very quickly, and many progresses are being made toward more accurate and efficient adaptive planning. Currently, only offline adap‐ tion has been reported in proton therapy due to the lack of accurate in‐room imaging system and long turnaround of manufacturing patient specific hardware for both PSPT and USPT. However, PBS has been increasingly used for lung cancer treatment, and CBCT and other in‐ room CT have become available. The advancement of both PBS and in‐room CT makes online adaptive planning possible in proton therapy in the future. Before online proton adaptive planning becomes a reality, many challenges need to be addressed. Better tools are needed for automatic image registration and dose accumulation, the dose calculation accuracy of in‐ room CT such as CBCT needs to be improved, and automatic and fast robust re‐planning and QA with IMPT should be developed. In addition, criteria on plan adaptation based on both dosimetric parameters and clinic outcome should be developed for quick and accurate decision‐making.

While adaptive planning is needed for proton therapy of lung cancer, it is time and effort consuming, and not every patient can benefit from this process. It would be helpful to be able to predict when adaptive planning is needed and for which patients. This would allow personalized adaptive planning process for patients, improve treatment efficiency, save costs, reduce risks of treatment errors from the plan adaptation process, and eliminate unnecessary imaging dose to patients with the repeated CT scanning. Berkovic *et al.* used volume and dosimetric data to construct lookup tables in attempt to predict whether and when ART could be useful based on the timing of the radiation treatment, the tumor volume, and whether it was a concurrent or sequential chemo‐radiotherapy [36]. Based on our experience with USPT, it is found that patients with noticeable weight change (e.g., 3% or more), pleural effusion, and pneumonitis in addition to the tumor volume change are indicatives of plan adaptation.
