Knowledge-Based Planning and Multi-criteria Optimization Combination with Volumetric Modulated Arc Therapy for Patients with Laryngeal Cancer

Purpose: This study aimed to evaluate the effectiveness of knowledge-based planning (KBP) with additional multi-criteria optimization (MCO) for the volumetric modulated arc therapy (VMAT) of the larynx. Methods: Twenty previously treated VMAT was used to train the KBP model. A manual plan (MP) was created using 10 patients. The KBP was generated by a single optimization with no planner intervention during the optimization. An additional MCO was applied to KBP (KBP + MCO) to further improve the plan’s quality. All plans were normalized to the planning target volume (PTV) D 50% (percentage of the volume receiving 50% of the prescribed dose). The PTV D95%, conformity index (CI), doses to the carotid artery (V 10-50 Gy : absolute volume receiving 10–50 Gy), and maximum dose to the spinal cord were calculated for the three planning methods. Results: The average PTV D 95% for MP, KBP, and KBP + MCO were 90.2%, 92.8%, and 94.1%, respectively. The average CI95% for KBP + MCO was 0.06 and 0.02 higher than that for MP and KBP, respectively. PTV and CI 95% were not significantly different between the three planning methods ( P > 0.05). KBP + MCO showed that the carotid artery for V10-30 Gy with MP was smaller than that for KBP, and there were no significant differences between V 40 Gy and V 50 Gy . The maximum dose to the spinal cord did not differ significantly among the three planning methods. Conclusions: KBP + MCO could improve dose sparing of the carotid artery while maintaining PTV coverage in laryngeal VMAT.