ESTRO 2024 - Abstract Book

S5012

Physics - Radiomics, functional and biological imaging and outcome prediction

ESTRO 2024

1081

Poster Discussion

Comparing predictors of radiation-induced lymphopenia in various timeframes in NSCLC radiotherapy

Takahiro Kanehira 1 , Hiroshi Taguchi 2 , Jun Sakakibara-Konishi 3 , Norio Katoh 4 , Yusuke Uchinami 4 , Keiji Kobashi 5 , Takayuki Hashimoto 5 , Hidefumi Aoyama 4 1 Hokkaido University Hospital, Department of Medical Physics, Sapporo, Japan. 2 Hokkaido University Hospital, Department of Radiation Oncology, Sapporo, Japan. 3 Hokkaido University, Department of Respiratory Medicine, Sapporo, Japan. 4 Hokkaido University Faculty of Medicine and Graduate School of Medicine, Department of Radiation Oncology, Sapporo, Japan. 5 Faculty of Medicine, Hokkaido University, Global Center for Biomedical Science and Engineering, Sapporo, Japan

Purpose/Objective:

Radiation-induced lymphopenia (RIL) has been shown to affect the survival of patients with various type of cancer. While RIL is commonly investigated during radiotherapy (RT), its presence in the post-RT may influence treatment outcomes [1], potentially when followed by immunotherapy. Given dynamic destruction and recovery of lymphocytes in response to radiation, the predictors may vary depending on the timing of evaluation. This study aimed to compare predictors of RIL at different timeframes for non-small cell lung cancer (NSCLC) patients.

Material/Methods:

This retrospective study included 86 stage І- Ⅳ NSCLC patients who received RT with or without chemotherapy between 2011 and 2023 in our hospital. Two prescribed doses were employed: 60 Gy in 30 fractions or 66 Gy in 33 fractions. Absolute lymphocyte counts (ALC) were collected from blood tests performed from 1 week pre-RT to 12 weeks post-RT. Nadir ALCs were determined for four distinct periods: from RT start to 1 week post-RT (up to 1 week post-RT) and 1-4 weeks, 4-8 weeks, and 8-12 weeks post-RT. In each phase, the nadir ALC was evaluated for grade 3 or greater RIL (G3+RIL), which is defined as an ALC less than 500 [/μl], in accordance with the common terminology criteria for adverse events version 5.0. Univariate logistic regression analyses were performed to identify the association of both dose-volume histogram (DVH) and non-DVH parameters with G3+RIL. Subsequently, multivariate logistic regression analyses, using forward stepwise variable selection, were performed including factors that met a p-value <0.2 and exhibited the best Akaike information criterion (AIC) for each organ at risk (OAR) from univariate analyses. A p-value less than 0.05 was considered statistically significant, and model fit was evaluated using the AIC.

Results:

The median baseline ALC was 1614 [/μl] (IQR:1375–2024). The rates of G3+RIL were 83% (n=71) up to 1 week post-RT, and 28% (n=24), 28% (n=24), and 20% (n=17) 1-4 weeks, 4-8 weeks, and 8-12 weeks post-RT, respectively. Univariate analyses, as depicted in Figure 1, showed that baseline ALC was a significant predictor across all evaluated periods. Among the DVH parameters for lungs, body, and thoracic vertebral body, V5 (where Vx denotes the volume receiving