Laboratory of Human Carcinogenesis, Center for Cancer Research, National Cancer Institute, National Institutes of Health, Bethesda, Maryland 20892, USA.
Clin Cancer Res. 2010 Dec 1;16(23):5824-34. doi: 10.1158/1078-0432.CCR-10-1110. Epub 2010 Oct 14.
Esophageal cancer is one of the most aggressive and deadly forms of cancer; highlighting the need to identify biomarkers for early detection and prognostic classification. Our recent studies have identified inflammatory gene and microRNA signatures derived from tumor and nontumor tissues as prognostic biomarkers of hepatocellular, lung, and colorectal adenocarcinoma. Here, we examine the relationship between expression of these inflammatory genes and micro RNA (miRNA) expression in esophageal adenocarcinoma and patient survival.
We measured the expression of 23 inflammation-associated genes in tumors and adjacent normal tissues from 93 patients (58 Barrett's and 35 Sporadic adenocarcinomas) by quantitative reverse transcription-polymerase chain reaction. These data were used to build an inflammatory risk model, based on multivariate Cox regression, to predict survival in a training cohort (n = 47). We then determined whether this model could predict survival in a cohort of 46 patients. Expression data for miRNA-375 were available for these patients and was combined with inflammatory gene expression.
IFN-γ, IL-1α, IL-8, IL-21, IL-23, and proteoglycan expression in tumor and nontumor samples were each associated with poor prognosis based on Cox regression [(Z-score)>1.5] and therefore were used to generate an inflammatory risk score (IRS). Patients with a high IRS had poor prognosis compared with those with a low IRS in the training (P = 0.002) and test (P = 0.012) cohorts. This association was stronger in the group with Barrett's history. When combining with miRNA-375, the combined IRS/miR signature was an improved prognostic classifier than either one alone.
Transcriptional profiling of inflammation-associated genes and miRNA expression in resected esophageal Barrett's-associated adenocarcinoma tissues may have clinical utility as predictors of prognosis.
食管癌是最具侵袭性和致命性的癌症之一；这凸显了需要识别生物标志物以进行早期检测和预后分类的必要性。我们最近的研究已经确定了源自肿瘤和非肿瘤组织的炎症基因和 microRNA 特征，它们是肝癌、肺癌和结直肠腺癌的预后生物标志物。在这里，我们研究了这些炎症基因和 microRNA（miRNA）在食管腺癌和患者生存中的表达之间的关系。
我们通过定量逆转录聚合酶链反应测量了 93 名患者（58 名 Barrett's 和 35 名散发性腺癌）肿瘤和相邻正常组织中 23 种炎症相关基因的表达。这些数据用于基于多变量 Cox 回归构建炎症风险模型，以预测训练队列（n = 47）中的生存情况。然后，我们确定该模型是否可以预测 46 名患者的生存情况。这些患者的 miRNA-375 表达数据可用，并与炎症基因表达相结合。
基于 Cox 回归（Z 分数>1.5），肿瘤和非肿瘤样本中的 IFN-γ、IL-1α、IL-8、IL-21、IL-23 和蛋白聚糖表达均与预后不良相关，因此用于生成炎症风险评分（IRS）。在训练（P = 0.002）和测试（P = 0.012）队列中，IRS 较高的患者与 IRS 较低的患者相比，预后较差。在有 Barrett's 病史的患者中，这种相关性更强。当与 miRNA-375 结合时，联合 IRS/miR 特征比单独使用任何一个特征作为预后分类器的效果都要好。
对切除的食管 Barrett's 相关腺癌组织中炎症相关基因和 miRNA 表达的转录谱分析可能具有作为预后预测因子的临床应用价值。
