影像学多维建模对早期肺腺癌脏层胸膜侵犯的术前精准分期及潜在机制研究

早期肺腺癌侵犯脏层胸膜TNM分期应从IA期提升至IB期，其精准分期对治疗方案优化及预后判断至关重要。弹力纤维（EFs）染色是诊断脏层胸膜侵犯（VPI）的主要手段，但EFs染色为术后诊断，且因采样误差低估隐匿性VPI的存在为后续研究金标准的建立产生重大影响。我们前期研究表明：术前影像肿瘤-胸膜的空间定位联合病理，可提高EFs染色的准确性。此外发现肿瘤多个CT影像组学特征及胸膜连接线的MRI动态强化方式均与VPI发生相关。据此推测：肿瘤内部异质性及邻近结构改变是导致VPI发生并进展的关键因素；但目前对于VPI的术前精准界定尚缺乏有效的评价手段。课题组拟通过CT影像组学分析肿瘤内在异质性特征、DCE-MRI测定瘤周间质液压力、三维重建并量化肿瘤与邻近结构的空间位置特征等手段，构建VPI术前精准分期的评估手段。这对于阐明VPI的发生及演进机制、提高临床精准化诊断及治疗水平提供新思路。

In the TNM Classification, stage IA tumor presented with visceral pleural invasion (VPI) should be upstaged to stage IB in lung adenocarcinoma, which leads to different clinical implications. Hematoxylin and eosin (HE) and elastic fibers (EFs) staining are the conventional tools for VPI examination, but with limitations such as post-operation and sampling bias, which might cause underestimation. Our preliminary experiment suggested that combination preoperative CT location parameters with targeted-EFs staining could improve the performance accuracy of histopathologic examination. Additionally, we prospectively found that intra-tumorous radiomics features at CT and peri-tumorous enhancement at DCE-MRI showed clearly relationships with VPI. Therefore, we assumed that study of internal heterogeneity and tumor-pleura linear tags via multimodal imaging methods might help to disclose biological mechanism of VPI. We attempt to develop a multidimensional imaging model by use of CT-based radiomics, MRI-based estimation of periphery interstitial fluid pressure (IFP) and 3D reconstruction of structural relationships between tumor and neighborhood, for precisely staging visceral pleural invasion in stage I lung adenocarcinoma. We hope this study may elucidate the evolution mechanism of VPI, clearly offer more critical roles for patient care and managements in clinics.