Estimating the effects of x-ray radiations on lymphocytes by minimally invasive holographic imaging flow cytometry

Nowadays, radiotherapy (RT) is a consolidated treatment for the cancer care. In fact, ionizing radiations are employed in over 50% of cancer patients. However, some side effects are correlated with RT, such as the radiation-induced lymphopenia (RIL). RIL is due to circulating lymphocytes (LCs) that pass through the irradiation field. LCs are the most radiosensitive blood cells, therefore RT can affect the count and activity of LCs. LCs are an important component of the human immune system, thus RIL has been linked with worse outcomes in multiple solid tumors and poor survival. As the occurrence rate of RIL is ∼40%-70% of patients treated with RT, an effective tool for predicting and minimizing RIL is highly demanded. Here we propose the proof-of-concept of a minimally invasive approach to monitor alterations induced by the radiation exposure inside LCs. To this aim, we combine label-free Holographic Imaging Flow Cytometry and Machine Learning to study Jurkat cells as model of T-LCs irradiated with two x-ray doses (i.e. 2 and 10 Gy of 6 MeV photons). The proposed method allows correlating the morphological features extracted by the quantitative phase-contrast maps of irradiated LCs flowing in a microfluidic chip to their radiation response. Furthermore, we train several machine learning models at different time points after RT for assessing the best strategy to reveal its effect on irradiated LCs. The attained results pave the way to future and deeper investigations for the development of a label-free, minimally invasive, and high-throughput tool for predicting and minimizing the RIL side effects.