Gene expression analysis of pretreatment biopsies predicts the pathological response of esophageal squamous cell carcinomas to neo-chemoradiotherapy
Gene expression analysis of pretreatment biopsies predicts the pathological response of esophageal squamous cell carcinomas to neo-chemoradiotherapy
复制标题
治疗前活检的基因表达分析可预测食管鳞状细胞癌对新放化疗的病理反应。
DOI：
10.1093/annonc/mdu201
复制
发表时间：
2014-09-01
影响因子：
50.5
通讯作者：
Fu, J. H.
中科院分区：
文献类型：
--
作者：
Wen, J.;Yang, H.;Fu, J. H.
背景
新辅助放化疗（neo-CRT）后手术已被证明可以提高食管鳞状细胞癌（ESCC）患者的生存率相比，单纯手术。然而，CRT的结果是异质性的，目前没有临床或病理学方法可以预测CRT的反应。在这项研究中，我们的目的是确定mRNA标记物用于ESCC CRT反应预测。
患者和方法
对28例接受新CRT和手术的ESCC的预处理癌活检进行基因表达分析。评估手术标本对CRT的病理反应。通过实时定量聚合酶链反应（qPCR）对表达谱鉴定的差异表达基因进行验证，并利用Fisher线性判别分析对qPCR数据建立分类模型。在第二组32例ESCC中进一步评估了该模型的预测能力。
结果
对28例ESCC的分析鉴定了10个差异表达基因，在病理完全缓解（pCR）和小于pCR（<pCR）的患者之间变化超过两倍。生成基于三个基因的qPCR值的预测模型，其在留一法交叉验证时提供86%的预测准确度。此外，该模型的预测能力在另一组32例ESCC中进行了验证，其中预测准确率为81%。重要的是，分别在训练集（P = 0.015）和验证集（P = 0.017）中，发现判别分数是影响新CRT反应的唯一独立因素。
结论
通过qPCR确定的三个基因的表达水平为ESCC CRT预测提供了可能的模型，这将促进ESCC治疗的个体化。需要在更大的独立队列中进行进一步的前瞻性验证，以充分评估其预测能力。
BACKGROUND
Neoadjuvant chemoradiotherapy (neo-CRT) followed by surgery has been shown to improve esophageal squamous cell carcinoma (ESCC) patients' survival compared with surgery alone. However, the outcomes of CRT are heterogeneous, and no clinical or pathological method can currently predict CRT response. In this study, we aim to identify mRNA markers useful for ESCC CRT-response prediction.
PATIENTS AND METHODS
Gene expression analyses were carried out on pretreated cancer biopsies from 28 ESCCs who received neo-CRT and surgery. Surgical specimens were assessed for pathological response to CRT. The differentially expressed genes identified by expression profiling were validated by real-time quantitative polymerase chain reaction (qPCR), and a classifying model was built from qPCR data using Fisher's linear discriminant analysis. The predictive power of this model was further assessed in a second set of 32 ESCCs.
RESULTS
The profiling of the 28 ESCCs identified 10 differentially expressed genes with more than a twofold change between patients with pathological complete response (pCR) and less than pCR (<pCR). A prediction model based on the qPCR values of three genes was generated, which provided a predictive accuracy of 86% upon leave-one-out cross-validation. Furthermore, the predictive power of this model was validated in another cohort of 32 ESCCs, among which a predictive accuracy of 81% was achieved. Importantly, the discriminant score was found to be the only independent factor that affected neo-CRT response in both the training (P = 0.015) and validation (P = 0.017) sets, respectively.
CONCLUSION
The expression levels of three genes determined by qPCR provide a possible model for ESCC CRT prediction, which will facilitate the individualization of ESCC treatment. Further prospective validation in larger independent cohorts is necessary to fully assess its predictive power.