Copyright
©The Author(s) 2023. Published by Baishideng Publishing Group Inc. All rights reserved.
Comparison and development of machine learning for thalidomide-induced peripheral neuropathy prediction of refractory Crohn’s disease in Chinese population
Jing Mao, Kang Chao, Fu-Lin Jiang, Xiao-Ping Ye, Ting Yang, Pan Li, Xia Zhu, Pin-Jin Hu, Bai-Jun Zhou, Min Huang, Xiang Gao, Xue-Ding Wang
Jing Mao, Fu-Lin Jiang, Ting Yang, Pan Li, Bai-Jun Zhou, Min Huang, Xue-Ding Wang, Institute of Clinical Pharmacology, School of Pharmaceutical Sciences, Sun Yat-sen University, Guangzhou 510006, Guangdong Province, China
Jing Mao, Ting Yang, Pan Li, Min Huang, Xue-Ding Wang, Guangdong Provincial Key Laboratory of New Drug Design and Evaluation, Sun Yat-sen University, Guangzhou 510006, Guangdong Province, China
Kang Chao, Xia Zhu, Pin-Jin Hu, Xiang Gao, Department of Gastroenterology, The Sixth Affiliated Hospital, Sun Yat-sen University, Guangzhou 510006, Guangdong Province, China
Xiao-Ping Ye, Department of Pharmacy, Guangdong Women and Children Hospital, Guangzhou 510000, Guangdong Province, China
Author contributions: Mao J, Chao K, Zhu X, and Wang XD designed the study; Mao J and Zhou BJ performed the experiments; Chao K, Gao X, Yang T and Li P enrolled the patients and collected the clinical data; Mao J and Ye XP performed the machine learning analyses; Huang M and Hu PJ supervised the study; Mao J and Wang XD wrote the manuscript.
Supported by National Natural Science Foundation of China, No. 81973398, No. 81730103, No. 81573507 and No. 82020108031; The National Key Research and Development Program, No. 2017YFC0909300 and No. 2016YFC0905001; Guangdong Provincial Key Laboratory of Construction Foundation, No. 2017B030314030 and No. 2020B1212060034; Science and Technology Program of Guangzhou, No. 201607020031; National Engineering and Technology Research Center for New Drug Druggability Evaluation (Seed Program of Guangdong Province), No. 2017B090903004; The 111 Project, No. B16047; China Postdoctoral Science Foundation, No. 2019M66324, No. 2020M683140 and No. 2020M683139; and Natural Science Foundation of Guangdong Province, No. 2022A1515012549 and No. 2023A1515012667.
Institutional review board statement: This study was reviewed and approved by the Ethics Committee of the Sixth Affiliated Hospital (No. E2016022), Sun Yat-Sen University, Guangzhou, China.
Informed consent statement: All study participants or their legal guardian provided informed written consent about personal and medical data collection prior to study enrolment.
Conflict-of-interest statement: The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
Data sharing statement: Technical appendix, statistical code, and dataset available from the corresponding author at wangxd@mail.sysu.edu.cn consent was not obtained but the presented data are anonymized and risk of identification is low.
Open-Access: This article is an open-access article that was selected by an in-house editor and fully peer-reviewed by external reviewers. It is distributed in accordance with the Creative Commons Attribution NonCommercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See:
https://creativecommons.org/Licenses/by-nc/4.0/ Corresponding author: Xue-Ding Wang, PharmD, Professor, Teacher, Institute of Clinical Pharmacology, School of Pharmaceutical Sciences, Sun Yat-sen University, No. 132 Waihuan Dong Road, Guangzhou 510006, Guangdong Province, China. wangxd@mail.sysu.edu.cn
Received: April 5, 2023
Peer-review started: April 5, 2023
First decision: April 21, 2023
Revised: May 7, 2023
Accepted: May 23, 2023
Article in press: May 23, 2023
Published online: June 28, 2023
Processing time: 84 Days and 1.2 Hours
ARTICLE HIGHLIGHTS
Research background
Thalidomide-induced peripheral neuropathy (TiPN), a life-threatening condition in Crohn's disease (CD), has a high incidence in Asia. However, there are no effective medical interventions for TiPN.
Research motivation
Can we develop a predictive model of TiPN combining genetic and clinical variables? Which variable affects TiPN more?
Research objectives
To establish an optimal model using clinical variables and genotypes to predict TiPN and improve the safety for the thalidomide treatment.
Research methods
A total of 164 patients diagnosed with CD at the Sixth Affiliated Hospital of the Sun Yat-Sen University were included in this study. Peripheral blood was collected from the patients to detect the genotypes at School of Pharmaceutical Sciences, Sun Yat-Sen University. The X2 method or Single-sample Kolmogorov–Smirnov test was used to determine the association of TiPN with 18 clinical features and 150 genetic variables. Five predictive models were established and evaluated by the confusion matrix receiver operating characteristic curve (AUROC), area under the precision-recall curve (AUPRC), specificity, sensitivity (recall rate), precision, accuracy, and F1 score.
Research results
TiPN was observed in 59 individuals. Among the five models, extreme gradient boosting (XGBoost) algorithm exhibited the best predictive performance with the highest specificity (0.857), accuracy (0.818), AUPRC (0.86) and AUROC (0.89) after evaluation. The top-ranking five risk variables associated with TiPN were interleukin-12 rs1353248 [P = 0.0004, odds ratio (OR): 8.983, 95% confidence interval (CI): 2.497-30.90), dose (mg/d, P = 0.002), brain-derived neurotrophic factor (BDNF) rs2030324 (P = 0.001, OR: 3.164, 95%CI: 1.561-6.434), BDNF rs6265 (P = 0.001, OR: 3.150, 95%CI: 1.546-6.073) and BDNF rs11030104 (P = 0.001, OR: 3.091, 95%CI: 1.525-5.960).
Research conclusions
The XGBoost algorithm accurately predicts TiPN using 18 clinical features and 14 genetic variables. It is able to identify high-risk patients using single nucleotide polymorphisms.
Research perspectives
Applying the machine learning to adjust thalidomide therapies based on these specific genotypes is recommended before the thalidomide treatment.
