MALDI蛋白指纹图谱筛选NSCLC患者化疗耐药相关血清标志物

Identification of serum biomarkers for predicting chemotherapy resistance in NSCLC by MALDI-TOF-MS

目的通过检测吉西他滨联合顺铂方案化疗敏感及耐药的晚期非小细胞肺癌患者治疗前血清蛋白质谱的差异,寻找可能的化疗耐药血清预测因子。方法收集非小细胞肺癌(non-small cell lung cancer,NSCLC)Ⅳ期患者治疗前血清,予以吉西他滨联合顺铂化疗,每2疗程后评估疗效。基质辅助激光解析电离化飞行时间质谱(matrix-assisted laser desorption/ionization-time of flight-mass spectrometry,MALDI-TOF-MS)检测血清蛋白,生物信息学分析找出差异蛋白峰,对筛选的差异蛋白峰进一步用遗传算法结合支持向量机模型的方法筛选最佳模型。结果收集25例晚期NSCLC患者,14例部分缓解(partial response,PR)进入化疗敏感组,2例疾病稳定(stable disease,SD)和9例出现疾病进展(progression disease,PD)进入化疗耐药组。化疗敏感组与耐药组出现10个明显差异蛋白质峰(P<0.05)。其中筛选得到4 978 Da、2 999 Da、4 996 Da、2 627 Da、4 188 Da、1 458 Da及1 984 Da 7个蛋白峰组成的最佳模型。结论 MALDI-TOF-MS技术检测得到血清差异蛋白模型可用于预测吉西他滨联合铂类二药化疗疗效。但需进一步扩大样本量以完善及验证疗效预测模型。

Objective To identify serum protein markers for predicting chemotherapy resistance in patients with advanced non-small cell lung cancer（NSCLC）.Methods Twenty-five patients with stage Ⅳ NSCLC receiving cisplatin plus gemcitabine chemotherapy were recruited in the study.The therapeutic response of patients was evaluated after 2 cycles of chemotherapy.Serum protein profiles were detected by matrix-assisted laser desorption/ionization-time of flight-mass spectrometry（MALDI-TOF-MS） before chemotherapy,and the spectra were analyzed with ZJU-Protein Chip Data Analyze System and a support vector machine（SVM）.Results Patients with partial response were classified as chemotherapy sensitive group（n = 14）,and those with stable disease（n = 2） and progressive disease（n = 9） as chemotherapy resistant group.Ten mass peaks were detected in chemotherapy sensitive group and chemotherapy resistant group.A modal was built,which consisted of 7 protein peaks,4 978 Da,2 999 Da,4 996 Da,2 627 Da,4 188 Da,1 458 Da and 1 984 Da,with significantly differential expression between chemotherapy sensitive group and resistant group.Conclusion A model consisting of differentially expressed serum proteins has been built to predict chemotherapy resistance in advanced NSCLC patients treated with cisplatin plus gemcitabine.The validity of the model needs to be tested by further studies on large scale.