Identification and prediction of immune checkpoint inhibitors-related pneumonitis by machine learning

作者全名："Gong, Li; Gong, Jun; Sun, Xin; Yu, Lin; Liao, Bin; Chen, Xia; Li, Yong-sheng"

作者地址："[Gong, Li; Liao, Bin; Li, Yong-sheng] Chongqing Univ, Canc Hosp, Dept PhaseClin Trial Ward 1, Chongqing Key Lab Translat Res Canc Metastasis & I, Chongqing, Peoples R China; [Gong, Jun] Chongqing Med Univ, Univ Town Hosp, Dept Informat Ctr, Chongqing, Peoples R China; [Sun, Xin; Yu, Lin] NanPeng Artificial Intelligence Res Inst Ltd, Dept Artificial Intelligence, Chongqing, Peoples R China; [Chen, Xia] Chongqing Univ, Clin Res Ctr, Canc Hosp, Chongqing Key Lab Translat Res Canc Metastasis & I, Chongqing, Peoples R China"

通信作者："Li, YS (通讯作者)，Chongqing Univ, Canc Hosp, Dept PhaseClin Trial Ward 1, Chongqing Key Lab Translat Res Canc Metastasis & I, Chongqing, Peoples R China.; Chen, X (通讯作者)，Chongqing Univ, Clin Res Ctr, Canc Hosp, Chongqing Key Lab Translat Res Canc Metastasis & I, Chongqing, Peoples R China."

来源：FRONTIERS IN IMMUNOLOGY

ESI学科分类：IMMUNOLOGY

WOS号：WOS:001026782400001

JCR分区：Q1

影响因子：5.7

年份：2023

卷号：14

期号：　

开始页：　

结束页：　

文献类型：Article

关键词：immune checkpoint inhibitors; pneumonitis; risk prediction; machine learning; risk factors

摘要："BackgroundImmune checkpoint inhibitor (ICI)-related pneumonitis (IRP) is a common and potentially fatal clinical adverse event. The identification and prediction of the risk of ICI-related IRP is a major clinical issue. The objective of this study was to apply a machine learning method to explore risk factors and establish a prediction model. MethodsWe retrospectively analyzed 48 patients with IRP (IRP group) and 142 patients without IRP (control group) who were treated with ICIs. An Elastic Net model was constructed using a repeated k-fold cross-validation framework (repeat = 10; k = 3). The prediction models were validated internally and the final prediction model was built on the entire training set using hyperparameters with the best interval validation performance. The generalizability of the final prediction model was assessed by applying it to an independent test set. The overall performance, discrimination, and calibration of the prediction model were evaluated. ResultsEleven predictors were included in the final predictive model: sindillizumab, number of & GE;2 underlying diseases, history of lung diseases, tirelizumab, non-small cell lung cancer (NSCLC), percentage of CD4(+) lymphocytes, body temperature, KPS score & LE;70, hemoglobin, cancer stage IV, and history of antitumor therapy. The external validation of the risk prediction model on an independent test set of 37 patients and showed good discrimination and acceptable calibration ability: with AUC of 0.81 (95% CI 0.58-0.90), AP of 0.76, scaled Brier score of 0.31, and Spiegelhalter-z of -0.29 (P-value:0.77). We also designed an online IRP risk calculator for use in clinical practice. ConclusionThe prediction model of ICI-related IRP provides a tool for accurately predicting the occurrence of IRP in patients with cancer who received ICIs."

基金机构：Chongqing Scientific Research Institutions Performance Incentive and Guidance Project (2020); Major International (Regional) Joint Research Program of the National Natural Science Foundation of China [81920108027]; Chongqing Outstanding Youth Foundation [cstc2020jcyj-jqX0030]; Founding of Chongqing University Innovation Group; Chongqing Youth Talents Program [CQYC20200301111]

基金资助正文："& nbsp;This study was funded by Chongqing Scientific Research Institutions Performance Incentive and Guidance Project (2020), the Major International (Regional) Joint Research Program of the National Natural Science Foundation of China (No. 81920108027), Chongqing Outstanding Youth Foundation (No. cstc2020jcyj-jqX0030), the Founding of Chongqing University Innovation Group, and the Chongqing Youth Talents Program (No. CQYC20200301111)."