Evaluating a Panel of Autoantibodies Against Tumor-Associated Antigens in Human Osteosarcoma

• Published in: Frontiers in genetics Vol. 13; p. 872253
• Main Authors: Luo, Manli, Wu, Songmei, Ma, Yan, Liang, Hong, Luo, Yage, Gu, Wentao, Fan, Lijuan, Hao, Yang, Li, Haiting, Xing, Linbo
• Format: Journal Article
• Language: English
• Published: Switzerland Frontiers Media S.A 25.04.2022
• Subjects: autoantibody; detection; early diagnosis; Genetics; osteosarcoma; panel; tumor-associated antigen; early diagnosis; detection; autoantibody; tumor-associated antigen; panel; osteosarcoma
• ISSN: 1664-8021; 1664-8021
• DOI: 10.3389/fgene.2022.872253

The aim of this study was to identify a panel of candidate autoantibodies against tumor-associated antigens in the detection of osteosarcoma (OS) so as to provide a theoretical basis for constructing a non-invasive serological diagnosis method in early immunodiagnosis of OS. The serological proteome analysis (SERPA) approach was used to select candidate anti-TAA autoantibodies. Then, indirect enzyme-linked immunosorbent assay (ELISA) was used to verify the expression levels of eight candidate autoantibodies in the serum of 51 OS cases, 28 osteochondroma (OC), and 51 normal human sera (NHS). The rank-sum test was used to compare the content of eight autoantibodies in the sera of three groups. The diagnostic value of each indicator for OS was analyzed by an ROC curve. Differential autoantibodies between OS and NHS were screened. Then, a binary logistic regression model was used to establish a prediction logistical regression model. Through ELISA, the expression levels of seven autoantibodies (ENO1, GAPDH, HSP27, HSP60, PDLIM1, STMN1, and TPI1) in OS patients were identified higher than those in healthy patients ( < 0.05). By establishing a binary logistic regression predictive model, the optimal panel including three anti-TAAs (ENO1, GAPDH, and TPI1) autoantibodies was screened out. The sensitivity, specificity, Youden index, accuracy, and AUC of diagnosis of OS were 70.59%, 86.27%, 0.5686, 78.43%, and 0.798, respectively. The results proved that through establishing a predictive model, an optimal panel of autoantibodies could help detect OS from OC or NHS at an early stage, which could be used as a promising and powerful tool in clinical practice.