基于磁共振功能影像的定量生物标志物研究交界性上皮性卵巢肿瘤的生物学特性

Borderline epithelial ovarian tumors (BEOT), also known as tumors of low malignant potential, account for 29% of ovarian epithelial malignancy. Previous studies and our work indicate that imageology and intraoperative frozen section could not accurately discriminate BEOT from ovarian cancer, which result in the radical surgical excision of reproductive organs， thereby permanent loss of fertility for young patients with BEOT. Our previous studies show that the solid components in BEOT have some characteristic pathologic features differenting from these in ovarian cancer. Conventional MRI can not yet demonstrate these differences. However, We infer that these differences can be elucidated by using functional MRI techniques such as diffusion weighted imaging (DWI), perfusion weighted imaging (PWI) and MR spectroscopy (MRS), from the perspectives of molecular movement, angiogenesis and tumor metabolism. We will investigate the quantitative biomarkers of DWI, PWI and MRS, and correlate them with quantitative cell density, vascular endothelial growth factor receptors (VEGFR), pericyte coverage index (PCI), microvessel density (MVD) and metabolits of ex vivo tumor tissue to validate BEOT's features of DWI，PWI and MRS and the differences between BEOT and ovarian cancer. We expect to clarify the histopathologic, molecular and metabolic features of BEOT and validate a noninvasive imaging tool to evaluate biological traits of ovarian malignancy.

交界性上皮性卵巢肿瘤（BEOT）为低度恶性肿瘤，占卵巢上皮性恶性肿瘤的29%，其治疗方案与卵巢良性肿瘤和卵巢癌不同。文献和我们的研究均显示：影像学和术中冰冻切片尚无法准确鉴别BEOT与卵巢癌，导致许多年轻患者接受了激进的分期手术，永久失去了生育能力。我们前期的常规磁共振成像（MRI）与病理对照发现：BEOT实性成分的结构特点与卵巢癌不同，虽然这些特点还无法在常规MRI上显示，但我们推测应用MR功能影像技术如扩散加权(DWI）、灌注加权(PWI)和磁共振波谱（MRS）可从分子运动、血管生成和肿瘤代谢的角度用定量生物标志物阐明；通过与离体肿瘤组织的细胞密度、血管内皮生长因子受体、毛细血管外膜细胞覆盖指数、微血管密度及代谢化合物进行定量比较，更可靠地评价BEOT的DWI、PWI和MRS特点及与卵巢癌的差异，阐明BEOT的组织病理、分子和代谢学特征，探索活体无创地评价卵巢癌的生物学特性的方法。

第一部分（DWI技术的探索）：探讨DWI鉴别卵巢交界性与恶性上皮性肿瘤的最佳 b 值。四个b 值（150 s/mm2、500 s/mm2、800 s/mm2及1000 s/mm2）DWI图像的SNR值与CNR值均有统计学差异（P＜0.05），随b 值增大DWI图像SNR值及CNR值均下降。.第二部分（MRS技术的探索）：探讨1H-MRS对卵巢良性、交界性和恶性上皮性囊性肿瘤的鉴别诊断价值。Cho峰随着肿瘤恶性程度增加而增高，明显升高NAA峰提示肿瘤为交界性。.第三部分（DCE技术的探索）：探讨全体积定量DCE-MRI直方图参数鉴别交界性与恶性上皮性卵巢肿瘤的价值，并评价DCE-MRI参数与病理学标记物间的相关性。Ktrans de 75th百分位数、kep 的90th百分位数和 ve的25th百分位数分别具有最高的诊断准确性和敏感度，值分别为77.17%和91.94%，78.26%和87.1%，77.17%和91.94%。.第四部分（动物模型）：动态观察二甲基苯蒽（DMBA）诱导SD大鼠上皮性卵巢肿瘤的发生与发展过程，建立时间、计量依赖的卵巢肿瘤模型并探究其发病机制。按肿瘤分化程度分为三组：囊肿或囊腺瘤96个（54.55%）、交界性肿瘤60个（34.09%）和腺癌20个（11.36%），计量和培养时间与肿瘤的分化程度存在线性正相关。