Dosimetric validation of two different radiobiological models for parotid gland functionality of tongue cancer

Abstract

Objective The aim of this study was to investigate the normal tissue complication probability (NTCP) for parotid gland of ca tongue patients from two different radiobiological models and to establish a dosimetric validity. Methods Lyman–Kutcher–Burman (LKB) model and linear quadratic (LQ) model were considered for determination of NTCP and tolerance dose (TD5/5 and TD50/5) for the parotid gland of 67 number of ca tongue patients who were treated with the volumetric arc therapy (VMAT) technique. An in‐house developed software on Excel (VBA) was used for this study. Results The equivalent uniform dose (EUD) was linearly proportional to the effective volume (veff) for parotid glands and there was a strong correlation between EUD and veff. At EUD = 46 to 47 Gy, the NTCP of parotid was 0.5 for both the models. The tolerance doses, TD5/5 (veff) and TD50/5 (veff) were exponentially reduced with increase of veff for LKB model; whereas these parameters were volume‐independent in the LQ model. TD5/5 (veff) and TD50/5 (veff) were 31.98 Gy and 45.98 Gy respectively for all 67 patients in the LQ model. Below TD50/5, NTCP of LKB model was less than the NTCP , calculated from LQ model. Conclusion One may consider radiobiological LQ model for estimation of clinical tolerance dose for OARs. Due to lack of clinical data, there are inaccuracy in determination of NTCP from LQ model. If sufficient number of tolerance data for partial volumes are available, the prediction of NTCP would be more confident.

Impact of Modern Low Dose Involved Site Radiation Therapy on Normal Tissue Toxicity in Cervicothoracic Non-Hodgkin Lymphomas: A Biophysical Study

“…Calculation of normal tissue complication probabilities (NTCP): A variety of different mathematical functions exist to describe the relationship between dose and tissue response. The LKB model is often mentioned as the classical dose-response model [16,19,24,30,31]. Its suitability for predicting probabilities of acute side effects was confirmed in previous projects [6].…”