AUTHOR=Zhang Xiao-zhen , Tao Su-ping , Liang Shi-xiong , Chen Shu-bin , Liu Fu-shuang , Jiang Wei , Chen Mao-jian 

TITLE=Nomogram based on circulating lymphocyte subsets for predicting radiation pneumonia in esophageal squamous cell carcinoma

JOURNAL=Frontiers in Immunology

VOLUME=Volume 13 - 2022

YEAR=2022

URL=https://www.frontiersin.org/journals/immunology/articles/10.3389/fimmu.2022.938795

DOI=10.3389/fimmu.2022.938795

ISSN=1664-3224

ABSTRACT=Purpose:  We aim to explore the relationship between circulating lymphocyte subsets of ESCC and RP, and further to establish and verify a nomogram model to predict RP (≥1) and RP (≥2).Patients and Methods: This study retrospectively collected 121 ESCC patients who were admitted to our treatment center from 2013 to 2021. Independent factors associated with RP (≥1) and RP (≥2) were identified by univariate and multivariate logistic regression and incorporated into nomograms. The prediction ability of the model is verified through Concordance Index (C-index), the Receiver Operating characteristic (ROC) curve, calibration curve, decision curve analysis (DCA), and internal verification. Nomograms and independent influencing factors compared the advantages of the models by comparing the area under the ROC curve (AUC). These models were further used for risk stratification to identify populations at high risk of RP (≥1) and RP (≥2).Results: Multivariate analysis suggested that TNM, pre-RT CD8+T cells, post-RT CD3+T cells, and V15 were independent predictive variables of RP (≥1), while pre-RT CD8+T cells, post-RT CD3+T cells, and V15 were independent predictors of RP (≥2). The C-indexes of RP (≥1) and RP (≥2) nomogram were 0.822(95% CI,0.73-0.914) and 0.787(95% CI,0.685-0.889) in the training cohort. The C-indexes of RP (≥1) and RP (≥2) nomogram were 0.709 (95% CI, 0.537-0.881) and 0.621(95% CI,0.404-0.837) in the validation cohort. Calibration curve suggested that the predicted values of model agreed well with the actual observations. ROC curve and DCA revealed that the advantages and the net benefit of two models were higher than that of the any independent influence variable. The incidence of RP (≥1) and RP (≥2) was higher in the high-risk group than in the low-risk group, suggesting that nomograms could help accurately identify patients at high risk of RP (≥1) and RP (≥2).Conclusion: RP is associated with an immune response. The pre-RT CD8+T cells and post-RT CD3+T cells were novel immune-related independent factors for RP (≥1) and RP (≥2). Nomograms based on circulating lymphocyte subsets significantly improved the prediction ability of RP, being helpful to guide clinicians to identify patients at high risk of RP (≥1) and RP (≥2).