INTRODUCTION: The lung radiosensitivity of the most sensitive patients limits doses that can be given to the majority of lung cancer patients. The purpose of the current study was to illustrate the concept of personalizing prescription dose by performing a retrospective study in which the prescription is determined using an individualized dose-volume constraint that is calculated from a toxicity prediction model. We test whether using a model-generated personalized lung-dose limit results in a clinically significant change to the prescription. METHODS: A model consisting of a dose-volume component and a genetic component (single-nucleotide polymorphism information) was used to determine iso-risk mean lung-dose (MLD) limits for each patient. The prescription dose for each patient was scaled according to the individualized MLD constraint and population-based constraints for the cord, esophagus, and heart. The difference between the model-determined prescription dose and the prescription the patient was originally treated with was evaluated. RESULTS: For 59% of the patients the change in prescription using the model-determined limit was greater than 5 Gy (either dose escalation or de-escalation). For 96% of the patients who developed radiation pneumonitis the model predicted that the prescription should have been lowered. CONCLUSIONS: Our results indicate that using a model-generated personalized MLD results in a clinically different (≥ 5 Gy) prescription. A model used in the manner described by the study can help physicians further personalize radiation therapy and aid them in determining how much dose can safely be delivered to the tumor and normal tissues.
INTRODUCTION: The lung radiosensitivity of the most sensitive patients limits doses that can be given to the majority of lung cancerpatients. The purpose of the current study was to illustrate the concept of personalizing prescription dose by performing a retrospective study in which the prescription is determined using an individualized dose-volume constraint that is calculated from a toxicity prediction model. We test whether using a model-generated personalized lung-dose limit results in a clinically significant change to the prescription. METHODS: A model consisting of a dose-volume component and a genetic component (single-nucleotide polymorphism information) was used to determine iso-risk mean lung-dose (MLD) limits for each patient. The prescription dose for each patient was scaled according to the individualized MLD constraint and population-based constraints for the cord, esophagus, and heart. The difference between the model-determined prescription dose and the prescription the patient was originally treated with was evaluated. RESULTS: For 59% of the patients the change in prescription using the model-determined limit was greater than 5 Gy (either dose escalation or de-escalation). For 96% of the patients who developed radiation pneumonitis the model predicted that the prescription should have been lowered. CONCLUSIONS: Our results indicate that using a model-generated personalized MLD results in a clinically different (≥ 5 Gy) prescription. A model used in the manner described by the study can help physicians further personalize radiation therapy and aid them in determining how much dose can safely be delivered to the tumor and normal tissues.
Authors: Yijun Ding; Harrison H Barrett; Matthew A Kupinski; Yevgeniy Vinogradskiy; Moyed Miften; Bernard L Jones Journal: Med Phys Date: 2019-06-07 Impact factor: 4.071
Authors: Dirk De Ruysscher; Jianyue Jin; Tim Lautenschlaeger; Jin-Xiong She; Zhongxing Liao; Feng-Ming Spring Kong Journal: Transl Lung Cancer Res Date: 2017-12