鼻咽癌乏氧微环境影像学评估体系的构建及抑制HIF-1α改善乏氧放疗增敏分子机制研究

The radiosensitivity of nasopharyngeal carcinoma cells is closely related to hypoxic microenvironment which consists of diffusion hypoxia and perfusion hypoxia, and there are currently no effective methods for assessing hypoxic response of tumor cells. BOLD MRI can effectively evaluate tissue oxygen partial pressure by R2* value, and IVIM DWI can get overall diffusion and perfusion information of tumor tissue by measuring D value, D* value and f value, which may be related with hypoxic type of tumor. In the present study, BOLD MRI and IVIM DWI play an important role in depicting longitudinal changes of tumor hypoxic microenvironment, including oxygen partial pressure, diffusion hypoxia and perfusion hypoxia. We attempt to investigate the feasibility of IVIM DWI as non-invasive methods in evaluating the relationship between tumor hypoxia and angiogenesis, as well as the correlation between IVIM DWI parameters and HIF-1α, VEGF, which is generally acknowledged to be correlated with tumor hypoxia. Meanwhile, we endeavor to analyze how celecoxib inhibit HIF-1α expression to improve oxygenation and radiosensitization of nasopharyngeal carcinoma cells in vitro and in vivo. All the results will help to clarify the change process of diffusion hypoxia and perfusion hypoxia. Furthermore, the project will test the possibility and feasibility of quantitative IVIM DWI parameters as independent factor to monitor the hypoxic microenvironment and treatment response, which may serve as a novel research tool in the future.

乏氧可导致鼻咽癌放疗抵抗，肿瘤乏氧微环境由灌注缺氧和弥散缺氧构成，目前尚缺乏合适的评估手段。本项目将IVIM DWI应用于鼻咽癌乏氧微环境的评估，与BOLD MRI结合，将定量影像参数D值、D*值、f值及R2*值与HIF-1α、VEGF表达水平关联，针对性评价弥散缺氧及灌注缺氧，构建肿瘤乏氧微环境全面立体的评估体系。并基于前期基础，进一步将功能影像参数与实测pO2对照，深入分析离体低氧及活体状态下塞来昔布对鼻咽癌放疗增敏效果，探讨抑制HIF-1α改善细胞氧合状态的放疗增敏分子机制。本项目有助于揭示鼻咽癌弥散缺氧及灌注缺氧的微观体系，阐明以HIF-1α为靶点的放疗增敏调控路径，为定量影像生物指标评价乏氧微环境及预测鼻咽癌放疗敏感性提供理论依据；同时，本项目也为肿瘤乏氧微环境的分类和细化研究提供了良好的思路和研究手段，有望将IVIM DWI作为无创性评估乏氧微环境和监测缺氧靶点治疗的新工具。

乏氧可导致鼻咽癌放疗抵抗，肿瘤乏氧微环境由灌注缺氧和弥散缺氧构成，目前尚缺乏合适的评估手段。本研究针对鼻咽癌放疗抵抗的关键问题，首次提出了乏氧内源性标记物碳酸酐酶IX作为鼻咽癌乏氧的分子靶点，构建了鼻咽癌原位荷瘤及颈部自发转移淋巴结动物模型，实现了鼻咽癌乏氧微环境的活体可视化。其次，基于定量的影像组学分析方法，创新性的首次构建了一系列鼻咽癌影像组学模型。例如，鼻咽癌预后的精准预测，通过挖掘多参数磁共振图像中定量的纹理特征，构建鼻咽癌无进展生存预后预测模型，该模型预测精准度比临床模型提高10%。同时，在此过程中发现由于不同纹理特征筛选方法与不同的模型构建方法间的不同组合会产生不一样的预测效果。基于此设想，将多种特征筛选算法及分类方法进行两两组合，观察到串联效应，得到预测最优模型。机器学习算法的预测效能依具体任务而定，既往影像组学模型基本采用单种算法进行建模，该研究发现不同算法之间的预测效能存在较大差异，随机森林不论是特征筛选还是分类均表现最优。同时，由于鼻咽癌失败的原因主要归咎于远处转移，因为我们通过影像组学的方法构建了鼻咽癌远处转移预测模型，该模型为高远处转移风险的鼻咽癌患者提供预警，以便得到更进一步的治疗方案。首次运用体素内不相干运动(intravoxel incoherent motion, IVIM)定量参数分别评价鼻咽部组织的扩散系数及组织微血管灌注。IVIM技术在鼻咽癌应用能将肿瘤的扩散和灌注用不同参数独立评估，且无需引入对比剂，易操作，可实现鼻咽癌精准诊断。以上研究从基础研究到临床研究，致力于实现精准影像，准确评估鼻咽癌患者的预后预测，以达到更加更个体化治疗方案。