Harima Yoko, Ikeda Koshi, Utsunomiya Keita, Shiga Toshiko, Komemushi Atsushi, Kojima Hiroyuki, Nomura Motoo, Kamata Minoru, Sawada Satoshi
Department of Radiology, Takii Hospital, Kansai Medical University, Moriguchi, Osaka, Japan.
Int J Radiat Oncol Biol Phys. 2009 Nov 15;75(4):1232-9. doi: 10.1016/j.ijrobp.2009.07.002.
To identify a set of genes related to the progression and metastasis of advanced cervical cancer after radiotherapy and to establish a predictive method.
A total of 28 patients with cervical cancer (15 stage IIIB, 13 stage IVA patients) who underwent definitive radiotherapy between May 1995 and April 2001 were included in this study. All patients were positive for human papillomavirus infection and harbored the wild-type p53 gene. The expression profiles of 14 tumors with local failure and multiple distant metastasis and 14 tumors without metastasis (cancer free) obtained by punch biopsy were compared before treatment, using a cDNA microarray consisting of 23,040 human genes.
Sixty-three genes were selected on the basis of a clustering analysis, and the validity of these genes was confirmed using a cross-validation test. The most accurate prediction was achieved for 63 genes (sensitivity, 78.8%; specificity, 38.1%). Some of these genes were already known to be associated with metastasis via chromosomal instability (TTK, BUB1B), extracellular matrix components (matrix metalloproteinase 1 [MMP-1]), and carcinogenesis (protein phosphatase 1 regulatory subunit 7 [PPP1R7]). A "predictive score" system was developed that could predict the probability for development of metastases using leave-one-out cross-validation methods.
The present results may provide valuable information for identified predictive markers and novel therapeutic target molecules for progression and metastasis of advanced cervical cancer.
鉴定一组与晚期宫颈癌放疗后进展和转移相关的基因，并建立一种预测方法。
本研究纳入了1995年5月至2001年4月期间接受根治性放疗的28例宫颈癌患者（15例IIIB期，13例IVA期）。所有患者均为人乳头瘤病毒感染阳性且携带野生型p53基因。在治疗前，使用包含23,040个人类基因的cDNA微阵列，比较通过穿刺活检获得的14例局部复发和多发远处转移肿瘤以及14例无转移（无癌）肿瘤的表达谱。
基于聚类分析选择了63个基因，并使用交叉验证测试确认了这些基因的有效性。对63个基因的预测最为准确（敏感性，78.8%；特异性，38.1%）。其中一些基因已知与通过染色体不稳定发生转移（TTK、BUB1B）、细胞外基质成分（基质金属蛋白酶1 [MMP-1]）和致癌作用（蛋白磷酸酶1调节亚基7 [PPP1R7]）有关。开发了一种“预测评分”系统，该系统可使用留一法交叉验证方法预测转移发生的概率。
本研究结果可能为鉴定晚期宫颈癌进展和转移的预测标志物及新型治疗靶点分子提供有价值的信息。
