Development of a novel prediction model for lymphopenia after concurrent chemo-radiotherapy

Abstract

Radiotherapy plays a pivotal role in cancer treatment, treating approximately 50% of cancer patients worldwide.Recently, immunotherapy has been introduced to clinics and the synergetic effect of combining radiotherapy andimmunotherapy is becoming a new standard of care for some cancer types. On the other hand, radiation-inducedlymphopenia, which is a common complication of radiotherapy, has been reported as an important factor, with itsstrong correlation not only to the treatment outcome but also to the survival. In this thesis study, we statistically analyzed the lymphocyte dynamics of locally-advanced (stage III) non-small cell lung cancer patients treated with chemo-radiotherapy and immunotherapy. Then, the correlation between radiotherapy plan and lymphocyte changes was investigated with three-dimensional dose distribution data using voxel-based analysis (VBA). Furthermore, based on the analyzed relationship in this study, we developed a novel neural network to predict the lymphopenia after chemo-radiotherapy and discussed the clinical meaning of the results.