Reyes Adalgiza, Marti Josep, Marfà Santiago, Jiménez Wladimiro, Reichenbach Vedrana, Pelegrina Amalia, Fondevila Constantino, Garcia Valdecasas Juan Carlos, Fuster Josep
Liver Surgery & Transplantation Unit, Department of Surgery, ICMDM, Hospital Clinic, IDIBAPS, CIBERehd, Villarroel, 170, 08036, Barcelona, Spain.
Biochemistry & Molecular Genetics Service, Hospital Clinic, IDIBAPS, CIBERehd, Villarroel, 170, 08036, Barcelona, Spain.
Future Oncol. 2017 Apr;13(10):875-882. doi: 10.2217/fon-2016-0461. Epub 2017 Jan 16.
To obtain proteomic profiles in patients with colorectal liver metastases (CRLM) and identify the relationship between profiles and the prognosis of CRLM patients.
MATERIALS & METHODS: Prognosis prediction (favorable or unfavorable according to Fong's score) by a classification and regression tree algorithm of surface-enhanced laser desorption/ionization TOF-MS proteomic profiles from cryopreserved CRLM (patients) and normal liver tissue (controls).
The protein peak 7371 m/z showed the clearest differences between CRLM and control groups (94.1% sensitivity, 100% specificity, p < 0.001). The algorithm that best differentiated favorable and unfavorable groups combined 2970 and 2871 m/z protein peaks (100% sensitivity, 90% specificity).
Proteomic profiling in liver samples using classification and regression tree algorithms is a promising technique to differentiate healthy subjects from CRLM patients and to classify the severity of CRLM patients.
获取结直肠癌肝转移（CRLM）患者的蛋白质组图谱，并确定图谱与CRLM患者预后之间的关系。
通过分类回归树算法，利用表面增强激光解吸/电离飞行时间质谱（SELDI-TOF-MS）技术，对来自冷冻保存的CRLM患者和正常肝组织（对照）的蛋白质组图谱进行预后预测（根据方氏评分分为良好或不良）。
蛋白质峰7371 m/z在CRLM组和对照组之间显示出最明显的差异（敏感性94.1%，特异性100%，p<0.001）。最佳区分良好和不良组的算法结合了2970和2871 m/z的蛋白质峰（敏感性100%，特异性90%）。
使用分类回归树算法对肝脏样本进行蛋白质组分析是一种很有前景的技术，可用于区分健康受试者和CRLM患者，并对CRLM患者的严重程度进行分类。
