Development of machine learning-based malignant pericardial effusion-related model in breast cancer: Implications for clinical significance, tumor immune and drug-therapy
Wendi Zhan,
Haihong Hu,
Bo Hao,
Hongxia Zhu,
Ting Yan,
Jingdi Zhang,
Siyu Wang,
Saiyang Liu,
Taolan Zhang
Affiliations
Wendi Zhan
School of Pharmacy, Hengyang Medical College, University of South China, 28 Western Changsheng Road, Hengyang, Hunan, 421001, China; Department of Pharmacy, The First Affiliated Hospital, Hengyang Medical School, University of South China, Hengyang, Hunan, 421001, China
Haihong Hu
School of Pharmacy, Hengyang Medical College, University of South China, 28 Western Changsheng Road, Hengyang, Hunan, 421001, China; Department of Pharmacy, The First Affiliated Hospital, Hengyang Medical School, University of South China, Hengyang, Hunan, 421001, China
Bo Hao
Department of Pharmacy, The First Affiliated Hospital, Hengyang Medical School, University of South China, Hengyang, Hunan, 421001, China
Hongxia Zhu
School of Pharmacy, Hengyang Medical College, University of South China, 28 Western Changsheng Road, Hengyang, Hunan, 421001, China; Department of Pharmacy, The First Affiliated Hospital, Hengyang Medical School, University of South China, Hengyang, Hunan, 421001, China
Ting Yan
Department of Breast and Thyroid Surgery, The First Affiliated Hospital, Hengyang Medical School, University of South China, Hengyang, Hunan, 421001, China
Jingdi Zhang
School of Pharmacy, Hengyang Medical College, University of South China, 28 Western Changsheng Road, Hengyang, Hunan, 421001, China; Department of Pharmacy, The First Affiliated Hospital, Hengyang Medical School, University of South China, Hengyang, Hunan, 421001, China
Siyu Wang
Department of Medical Oncology, The First Affiliated Hospital, Hengyang Medical School, University of South China, Hengyang, Hunan, 421001, China
Saiyang Liu
Shandong University of Traditional Chinese Medicine, Jinan, Shandong, 250355, China
Taolan Zhang
Department of Pharmacy, The First Affiliated Hospital, Hengyang Medical School, University of South China, Hengyang, Hunan, 421001, China; Phase I Clinical Trial Center, The First Affiliated Hospital, Hengyang Medical School, University of South China, Hengyang, Hunan, 421001, China; Corresponding author. The First Affiliated Hospital, Department of Pharmacy, Hengyang Medical School, 69 Chuanshan Road, Hengyang, Hunan, 421000, China.
Background: Malignant pericardial effusion (MPE) is a common complication of advanced breast cancer (BRCA) and plays an important role in BRCA. This study is aims to construct a prognostic model based on MPE-related genes for predicting the prognosis of breast cancer. Methods: The BRCA samples are analyzed based on the expression of MPE-related genes by using an unsupervised cluster analysis method. This study processes the data by least absolute shrinkage and selection operator and multivariate Cox analysis, and uses machine learning algorithms to construct BRCA prognostic model and develop web tool. Results: BRCA patients are classified into three clusters and a BRCA prognostic model is constructed containing 9 MPE-related genes. There are significant differences in signature pathways, immune infiltration, immunotherapy response and drug sensitivity testing between the high and low-risk groups. Of note, a web-based tool (http://wys.helyly.top/cox.html) is developed to predict overall survival as well as drug-therapy response of BRCA patients quickly and conveniently, which can provide a basis for clinicians to formulate individualized treatment plans. Conclusion: The MPE-related prognostic model developed in this study can be used as an effective tool for predicting the prognosis of BRCA and provides new insights for the diagnosis and treatment of BRCA patients.
