Precise prediction of the radiation pneumonitis in lung cancer: an explorative preliminary mathematical model using genotype information

J Cancer. 2020 Feb 10;11(8):2329-2338. doi: 10.7150/jca.37708. eCollection 2020.

Abstract

Purpose: Radiation pneumonitis (RP) is the most significant dose-limiting toxicity and is one major obstacle for lung cancer radiotherapy. Grade ≥2 RP usually needs clinical interventions and serve RP could be life threatening. Clinically, tissue response could be strikingly different even two similar patients after identical radiotherapy. Previous methods for the RP prediction can hardly distinguish substantial variations among individuals. Reliable predictive factors or methods emphasizing the individual differences are strongly desired by clinical radiation oncologists. The purpose of this study is to develop an approach for the personalized RP risk prediction. Experimental Design: One hundred eighteen lung cancer patients who received radiotherapy were enrolled. Seven hundred thousand single-nucleotide polymorphism (SNP) sites were assessed via Generalized Linear Models via Lasso and Elastic-Net Regularization (GLMNET) to determine their synergistic effects on the RP risk prediction. Non-genetic factors including patient's phenotypes and clinical interventional parameters were separately assessed by statistic test. Based on the results of the aforementioned analysis, a multiple linear regression model named Radiation Pneumonitis Index (RPI) was built, for the assessment of Grade ≥2RP risk. Results: Only previous surgery and fractional dose were discovered statistical significantly associated with grade ≥2RP. Thirty-nine effective SNPs for predicting the Grade ≥2RP risk were discovered and their coefficients of the synergistic effect were determined. The RPI score can successfully distinguish the RP≥2 population with 92.0% sensitivity and 100% specificity. Conclusions: Individual radiation sensitivity can be determined with genotype information and personalized radiotherapy could be achieved based on mathematical model result.