Combing NLR, V20 and mean lung dose to predict radiation induced lung injury in patients with lung cancer treated with intensity modulated radiation therapy and chemotherapy

The purpose was to evaluate the predictive value of baseline neutrophil to lymphocyte ratio (NLR) level in the incidence of grade 3 or higher radiation induced lung injury (RILI) for lung cancer patients. A retrospectively analysis with 166 lung cancer patients was performed. All of the enrolled patients received chemoradiotherapy at our hospital between April 2014 and May 2016. The Cox proportional hazard model was used to identify the potential risk factors for RILI. In this cohort, the incidence of grade 3 or higher RILI was 23.8%. Univariate analysis showed that radiation dose, volume at least received 20Gy (V20), mean lung dose and NLR were significantly associated with the incidence of grade 3 or higher RILI (P = 0.012, 0.008, 0.012, and 0.039, respectively). Multivariate analysis revealed that total dose ≥ 60 Gy, V20 ≥ 20%, mean lung dose ≥ 12 Gy, and NLR ≥ 2.2 were still independent predictive factors for RILI (P = 0.010, 0.043, 0.028, and 0.015, respectively). A predictive model of RILI based on the identified risk factors was established using receiver operator characteristic curves. The results demonstrated that the combination analysis of V20, mean lung dose and NLR was superior to either of the variables alone. Additionally, we found that the constraint of V20 and mean lung dose were meaningful for patients with higher baseline NLR level. If the value of V20 and mean lung dose lower than the threshold value, the incidence of grade 3 or higher RILI for the high NLR level patients could be decreased from 63.3% to 8.7%. Our study showed that radiation dose, V20, mean lung dose and NLR were independent predictors for RILI. Combination analysis of V20, mean lung dose and NLR may provide a more accurate model for RILI prediction.