Autoplanning in the setting of stereotactic body radiation therapy for lung cancer

Purpose: Automatic planning (AP) has been compared to manual planning (MP) in lung stereotactic body radiation therapy (SBRT) to validate the former and to implement it in clinical practice. Methods: A new developing Guided Planning System (GPS) engine was used to reoptimize 20 lung SBRT plans with the RayStation (TM) treatment planning system (TPS). The original manual plans were optimized to deliver 60 Gy in 5 or 8 fractions to the target with constraints on organs at risk (OARs) based on an internal protocol. AP plans were compared to MP based on (i) planning target volume (PTV) and OARs dosimetric evaluation, (ii) clinician's blind plan comparison, (iii) Plan QA results, and (iv) plan quality metrics. Differences between continuous variables were explored through the Mann-Whitney test (p < 0.05). Results: Target and OARs dosimetry showed no significant difference, with the only exception of the spinal cord maximum dose that was significantly lower for AP in the 5 fractions scheme (MP: 8.93 Gy +/- 3.94 Gy vs AP: 6.45 Gy +/- 2.72 Gy, p = 0.034). In the blind comparison, AP was preferred in 45 % of cases while MP in 35 % of cases (no preference was expressed in 20 % of cases). A trend towards lower monitor units (MUs) was found for AP in the 5 fractions scheme (MP: 3383 +/- 943 vs AP: 2662 +/- 588, p = 0.059). No significant difference was found in any of the plan quality metrics. Conclusions: AP plans were not inferior to MP plans; therefore, GPS is ready for clinical use in a pulmonary SBRT setting, reducing the planning workload and harmonizing the planning procedure.