Multi-parametric MRI Determination of Prostate Cancer Aggressiveness and Extent

描述(由申请人提供)：目前的诊断测试不能可靠地确定前列腺癌的范围(体积和位置)或生物侵袭性。我们的长期目标是开发一种非侵入性成像技术，准确评估前列腺癌的临床意义，并可用于诊断、治疗计划和治疗监测。这一特殊应用的主要目标是通过将解剖学和功能研究产生的多参数数据结合在一个新的统计模型中，实现3Tesla MRI生成癌症概率图的全部潜力。因此，基于我们小组和其他人以前的结果，中心假设是多参数解剖、血管和代谢数据可以确定前列腺癌的范围和侵袭性，这与术后组织病理学确定的范围和肿瘤分级以及侵袭性的分子评估的相关性得到了验证。在之前的发展支持下，这一假设将得到四个具体目标的检验：1)从用新技术获得和处理的MRI数据生成参数图；2)开发和验证三维(3D)策略，以将MRI图像与前列腺癌切除的组织病理学切片进行空间联合配准；3)开发基于3T MRI数据的分类器，以生成癌症的3D概率图；以及4)确定预测组织学和分子侵袭性标记的MRI特征。在前两个目标下，MRI数据将被获取，并使用开发的方法进行处理，以生成改进的参数图，然后将其配准以重建组织病理学体积。在第三个目标下，使用MRI参数图和组织病理学结果来训练统计分类器，以便于生成患者特定的癌症概率图。最后，在第四个目标中，已证实的侵袭性分子标记将与MRI、组织病理学和标准临床因素相关。这项拟议的工作在几个方面具有创新性：1)它将在3特斯拉系统上实现新的DCE-MRI和3DSI的获取和量化方法；2)它将使用新的和稳健的统计建模来同时结合解剖和MRI数据来确定癌症的范围；以及3)它将把MRI特征与空间注册的组织病理学和已证实的侵袭性生物标记物相关联。我们期望的结果是开发一种新的基于MRI的成像方法，以非侵入性和可靠的方式确定前列腺癌的范围和侵袭性。我们希望，这项研究中开发的方法可以让医生和他们的患者做出更好的治疗决定，并降低前列腺癌的发病率和死亡率。

DESCRIPTION (provided by applicant): Current diagnostic tests cannot reliably determine prostate cancer extent (volume and location) or biological aggressiveness. Our long term goal is to develop a non-invasive imaging technique that accurately assesses the clinical significance of prostate cancer and that can be used for diagnosis, treatment planning, and therapeutic monitoring. The main objective of this particular application is to realize the full potential of 3 Tesla MRI to generate cancer probability maps by combining the multi-parametric data generated from anatomic and functional studies within a new statistical model. Therefore, based on previous results from our group and others, the central hypothesis is that multi-parametric anatomic, vascular and metabolic data can determine the extent and aggressiveness of prostate cancer as validated by correlation with postoperative histopathologic determination of extent and tumor grade, and molecular assessment of aggressiveness. Supported by previous developments, this hypothesis will be tested with four specific aims: 1) generate parametric maps from MRI data acquired and processed with novel techniques; 2) develop and validate a 3-dimensional (3D) strategy to spatially co-register MRI images to histopathology sections from prostatectomy; 3) develop a classifier based on 3T MRI data to produce a 3D probability map of cancer; and 4) identify MRI features that predict histological and molecular markers of aggressiveness. Under the first two aims MRI data will be acquired and processed with developed methods to generate improved parametric maps which are then registered to reconstructed histopathology volumes. Under the third aim, the MRI parametric maps and histopathology results are used to train a statistical classifier to facilitate the generation of patient specific cancer probability maps. Finally, in the fourth aim, proven molecular markers of aggressiveness will be correlated with MRI, histopathology and standard clinical factors. The proposed work is innovative in several ways: 1) it will implement new acquisition and quantitation methods for DCE-MRI and 3DSI on a 3 Tesla system; 2) it will use novel and robust statistical modeling to simultaneously combine anatomic and MRI data to determine the extent of cancer, and 3) it will correlate MRI features with spatially registered histopathology and proven aggressiveness biomarkers. Our expected outcome is the development of a novel MRI-based imaging method to non-invasively and reliably determine both the extent and aggressiveness of prostate cancer. It is our hope that the methods developed in this study may permit doctors and their patients to make better treatment decisions and reduce morbidity and mortality due to prostate cancer.