Li Hongqing, Tang Zhonghao, Zhu Huili, Ge Haiyan, Cui Shilei, Jiang Weiping
Department of Respiratory Medicine, Huadong Hospital Affiliated to Fudan University, Shanghai, 200040, China.
J Cancer Res Clin Oncol. 2016 Jun;142(6):1191-200. doi: 10.1007/s00432-016-2130-7. Epub 2016 Mar 5.
Lung adenocarcinoma can easily cause malignant pleural effusion which was difficult to discriminate from benign pleural effusion. Now there was no biomarker with high sensitivity and specificity for the malignant pleural effusion.
This study used proteomics technology to acquire and analyze the protein profiles of the benign and malignant pleural effusion, to seek useful protein biomarkers with diagnostic value and to establish the diagnostic model.
We chose the weak cationic-exchanger magnetic bead (WCX-MB) to purify peptides in the pleural effusion, used matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF-MS) to obtain peptide expression profiles from the benign and malignant pleural effusion samples, established and validated the diagnostic model through a genetic algorithm (GA) and finally identified the most promising protein biomarker.
A GA diagnostic model was established with spectra of 3930.9 and 2942.8 m/z in the training set including 25 malignant pleural effusion and 26 benign pleural effusion samples, yielding both 100 % sensitivity and 100 % specificity. The accuracy of diagnostic prediction was validated in the independent testing set with 58 malignant pleural effusion and 34 benign pleural effusion samples. Blind evaluation was as follows: the sensitivity was 89.6 %, specificity 88.2 %, PPV 92.8 %, NPV 83.3 % and accuracy 89.1 % in the independent testing set. The most promising peptide biomarker was identified successfully: Isoform 1 of caspase recruitment domain-containing protein 9 (CARD9), with 3930.9 m/z, was decreased in the malignant pleural effusion.
This model is suitable to discriminate benign and malignant pleural effusion and CARD9 can be used as a new peptide biomarker.
肺腺癌易导致恶性胸腔积液，难以与良性胸腔积液相鉴别。目前尚无对恶性胸腔积液具有高敏感性和特异性的生物标志物。
本研究采用蛋白质组学技术获取并分析良性和恶性胸腔积液的蛋白质谱，寻找具有诊断价值的有用蛋白质生物标志物并建立诊断模型。
我们选择弱阳离子交换磁珠（WCX-MB）纯化胸腔积液中的肽段，使用基质辅助激光解吸/电离飞行时间质谱（MALDI-TOF-MS）获取良性和恶性胸腔积液样本的肽段表达谱，通过遗传算法（GA）建立并验证诊断模型，最终鉴定出最有前景的蛋白质生物标志物。
在包括25例恶性胸腔积液和26例良性胸腔积液样本的训练集中，建立了一个以3930.9和2942.8 m/z光谱为基础的GA诊断模型，其敏感性和特异性均为100%。在包含58例恶性胸腔积液和34例良性胸腔积液样本的独立测试集中验证了诊断预测的准确性。盲法评估结果如下：独立测试集中的敏感性为89.6%，特异性为88.2%，阳性预测值为92.8%，阴性预测值为83.3%，准确性为89.1%。成功鉴定出最有前景的肽生物标志物：含半胱天冬酶募集结构域蛋白9（CARD9）的异构体1，m/z为3930.9，在恶性胸腔积液中含量降低。
该模型适用于鉴别良性和恶性胸腔积液，CARD9可作为一种新的肽生物标志物。
