Room: Exhibit Hall | Forum 7
Purpose: To assess Autoplan quality in lung SBRT and correlate the organ dose from a prediction model.
Methods: Twenty lung SBRT cases planned manually with 50 Gy in 5 fractions were selected to be replanned using Pinnacle Autoplan module according to the constraint tables from the RTOG 0813 protocol. Organ at risk (OAR) sparing were compared between the manual and Autoplans. The dosimetric endpoints were compared with predicted doses from a commercial product, PlanIQ, which is based on the geometric relationships among the tumor volumes and OARs and the characteristic of radiation beams. Spearman's rank-order correlation test was used.
Results: All manual and Autoplans achieved clinically required target coverageâ€”at least 95% of the PTV received 100% of prescription dose. The Autoplans achieved equal or better OAR sparing: 14 of 20 in the maximum dose (Dmax) of the spinal cord, 18 of 20 in Dmax of the ipsilateral brachial plexus, 12 of 20 in V20Gy of whole lung, 13 of 20 in Dmax of the heart, 16 of 20 in Dmax of the esophagus, and 14 of 20 in Dmax of the trachea. PlanIQ predictions were obtained for Dmax of the spinal cord, heart, esophagus, and trachea, with feasibility factor set to 0, 0.1, and 0.5. The highest correlation for each OAR is: 0.5489 for the spinal cord, 0.9106 for the heart, 0.8229 for the esophagus, and 0.8027 for the trachea.
Conclusion: Autoplan in lung SBRT improved OAR sparing while keeping the same dose coverage to the tumor. The OAR dose prediction with the feasibility factor set to 0.1 correlates strongly with Dmax of the heart, esophagus, and trachea from the Autoplans, but moderately with Dmax of the spinal cord.