{"title":"Interpretable machine learning model predicting immune checkpoint inhibitor-induced hypothyroidism: A retrospective cohort study","authors":"Su-Yan Zhu, Tong-Tong Yang, Yi-Zhuo Zhao, Yu Sun, Xiao-Meng Zheng, Hong-Bin Xu","doi":"10.1111/cas.16352","DOIUrl":null,"url":null,"abstract":"<p>Hypothyroidism is a known adverse event associated with the use of immune checkpoint inhibitors (ICIs) in cancer treatment. This study aimed to develop an interpretable machine learning (ML) model for individualized prediction of hypothyroidism in patients treated with ICIs. The retrospective cohort of patients treated with ICIs was from the First Affiliated Hospital of Ningbo University. ML methods applied include logistic regression (LR), random forest classifier (RFC), support vector machine (SVM), and extreme gradient boosting (XGBoost). The area under the receiver-operating characteristic curve (AUC) was the main evaluation metric used. Furthermore, the Shapley additive explanation (SHAP) was utilized to interpret the outcomes of the prediction model. A total of 458 patients were included in the study, with 59 patients (12.88%) observed to have developed hypothyroidism. Among the models utilized, XGBoost exhibited the highest predictive capability (AUC = 0.833). The Delong test and calibration curve indicated that XGBoost significantly outperformed the other models in prediction. The SHAP method revealed that thyroid-stimulating hormone (TSH) was the most influential predictor variable. The developed interpretable ML model holds potential for predicting the likelihood of hypothyroidism following ICI treatment in patients. ML technology offers new possibilities for predicting ICI-induced hypothyroidism, potentially providing more precise support for personalized treatment and risk management.</p>","PeriodicalId":9580,"journal":{"name":"Cancer Science","volume":null,"pages":null},"PeriodicalIF":4.5000,"publicationDate":"2024-09-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11531944/pdf/","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Cancer Science","FirstCategoryId":"3","ListUrlMain":"https://onlinelibrary.wiley.com/doi/10.1111/cas.16352","RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"ONCOLOGY","Score":null,"Total":0}
引用次数: 0
Abstract
Hypothyroidism is a known adverse event associated with the use of immune checkpoint inhibitors (ICIs) in cancer treatment. This study aimed to develop an interpretable machine learning (ML) model for individualized prediction of hypothyroidism in patients treated with ICIs. The retrospective cohort of patients treated with ICIs was from the First Affiliated Hospital of Ningbo University. ML methods applied include logistic regression (LR), random forest classifier (RFC), support vector machine (SVM), and extreme gradient boosting (XGBoost). The area under the receiver-operating characteristic curve (AUC) was the main evaluation metric used. Furthermore, the Shapley additive explanation (SHAP) was utilized to interpret the outcomes of the prediction model. A total of 458 patients were included in the study, with 59 patients (12.88%) observed to have developed hypothyroidism. Among the models utilized, XGBoost exhibited the highest predictive capability (AUC = 0.833). The Delong test and calibration curve indicated that XGBoost significantly outperformed the other models in prediction. The SHAP method revealed that thyroid-stimulating hormone (TSH) was the most influential predictor variable. The developed interpretable ML model holds potential for predicting the likelihood of hypothyroidism following ICI treatment in patients. ML technology offers new possibilities for predicting ICI-induced hypothyroidism, potentially providing more precise support for personalized treatment and risk management.
期刊介绍:
Cancer Science (formerly Japanese Journal of Cancer Research) is a monthly publication of the Japanese Cancer Association. First published in 1907, the Journal continues to publish original articles, editorials, and letters to the editor, describing original research in the fields of basic, translational and clinical cancer research. The Journal also accepts reports and case reports.
Cancer Science aims to present highly significant and timely findings that have a significant clinical impact on oncologists or that may alter the disease concept of a tumor. The Journal will not publish case reports that describe a rare tumor or condition without new findings to be added to previous reports; combination of different tumors without new suggestive findings for oncological research; remarkable effect of already known treatments without suggestive data to explain the exceptional result. Review articles may also be published.