Department of Environmental Engineering, National Chung Hsing University, South Dist., Taichung, Taiwan, R.O.C.
Department of Dermatology, Changhua Christian Hospital, Changhua, Taiwan, R.O.C.
PLoS One. 2018 Sep 17;13(9):e0201241. doi: 10.1371/journal.pone.0201241. eCollection 2018.
Cumulative estrogen concentration is an important determinant of the risk of developing breast cancer. Estrogen carcinogenesis is attributed to the combination of receptor-driven mitogenesis and DNA damage induced by quinonoid metabolites of estrogen. The present study was focused on developing an improved breast cancer prediction model using estrogen quinone-protein adduct concentrations. Blood samples from 152 breast cancer patients and 71 healthy women were collected, and albumin (Alb) and hemoglobin (Hb) adducts of estrogen-3,4-quinone and estrogen-2,3-quinone were extracted and evaluated as potential biomarkers of breast cancer. A multilayer perceptron (MLP) was used as the predictor model and the resultant prediction of breast cancer was more accurate than other existing detection methods. A MLP using the logarithm of the concentrations of the estrogen quinone-derived adducts (four input nodes, 10 hidden nodes, and one output node) was used to predict breast cancer risk with accuracy close to 100% and area under curve (AUC) close to one. The AUC value of one showed that both data sets were separable. We conclude that Alb and Hb adducts of estrogen quinones are promising biomarkers for the early detection of breast cancer.
雌激素浓度累积是乳腺癌发病风险的重要决定因素。雌激素致癌归因于受体驱动的有丝分裂和雌激素醌类代谢物诱导的 DNA 损伤的结合。本研究旨在使用雌激素醌蛋白加合物浓度开发改良的乳腺癌预测模型。采集了 152 名乳腺癌患者和 71 名健康女性的血液样本，并提取和评估了雌激素-3,4-醌和雌激素-2,3-醌的白蛋白 (Alb) 和血红蛋白 (Hb) 加合物作为乳腺癌的潜在生物标志物。多层感知器 (MLP) 用作预测模型，乳腺癌的预测结果比其他现有检测方法更准确。使用雌激素醌衍生加合物浓度的对数的 MLP（四个输入节点、10 个隐藏节点和一个输出节点）来预测乳腺癌风险，准确率接近 100%，曲线下面积 (AUC) 接近 1。AUC 值为 1 表明两个数据集都是可分离的。我们得出结论，雌激素醌的 Alb 和 Hb 加合物是乳腺癌早期检测的有前途的生物标志物。
