Stereotactic Ablative MR-Guided Radiotherapy (SMART) for Early-Stage Laryngeal Cancer: Clinical and Dosimetric Insights

Abstract

Background and Objectives Laryngeal cancer is among the most common malignancies of the upper respiratory tract, with early-stage (T1,T2) cases typically treated by surgery or radiotherapy (RT), achieving comparable overall survival rates. Nowadays intensity-modulated radiation therapy (IMRT) and volumetric modulated arc therapy (VMAT) techniques are used for laryngeal cancer treatment irradiating a whole larynx volume. New technologies, such as magnetic resonance based linear accelerator (MR-linac), led to reduce the target volume and treatment time. This study aims to explore the clinical and dosimetric feasibility of stereotactic ablative radiotherapy (SABR) guided by MR-linac technology (SMART) as an alternative for early-stage laryngeal cancer. Material (patients) and research method used Two complementary investigations were performed. The first compared standard fractionated RT (63 Gy/28 fr.) to MR-Linac-based SABR (42.5 Gy/5 fr.) in 32 patients, evaluating safety, treatment duration, toxicity, and precision. The second focused on validating dosimetric accuracy using individualized 3D-printed phantoms filled with radiation-sensitive gels and assessed treatment plans through gamma analysis. Findings/results in sufficient details to support conclusions Results demonstrated that SMART enables more precise dose delivery, reduces treatment time and irradiated volume (up to 15%), and preserves laryngeal function. Dosimetric analysis showed high concordance between planned and delivered doses, with gamma passing rates above 95%. Conclusions and recommendations In conclusion, MR-Linac-guided SMART appears to be a safe, efficient, and innovative approach for early-stage laryngeal cancer treatment, with promising implications for clinical implementation and patient quality of life. However, challenges remain in target volume definition and further accuracy enhancement, highlighting the potential role of machine learning in future personalization.