Hepatocellular carcinoma (HCC) is the most common primary hepatic malignancy (more than 80%) and the third most common cause of death from cancer worldwide. Contrast-enhanced multi-phase multi-detector CT (MDCT) is a routine imaging modality for patients diagnosed with hepatic malignancy. The change of tumor size measured on single-phase (mostly portal-venous phase) MDCT images is used routinely for assessment of tumor response to treatment. With the advent of targeted cancer therapies (such as antiangiogenic treatment) over the past decade, the clinical outcome in the treatment of advanced HCC has been significantly improved. However, because of the difference in the mechanisms of therapies, measurement of tumors treated with targeted therapy does not necessarily accurately represent the change in viable tumor size due to the presence of necrosis. Therefore, the assessment of tumor response using tumor size alone may inadequately evaluate the treatment response for tumors like HCC when treated with targeted therapies. Instead of using tumor size alone to assess targeted therapies, we propose to develop an innovative quantitative imaging biomarker, denoted as hepatic tumor viability (HTV), using multi-phase hepatic MDCT images for quantification of viable and necrotic tumor tissues in addition to the size of liver and tumors for HCC patients treated with targeted therapies. This project will be built upon existing technologies for quantitative imaging analysis developed at the 3D Imaging Lab at the Massachusetts General Hospital (MGH). It will make use of the classification and segmentation algorithms developed by the co-PI for liver and liver tumors, which classifies viable/necrotic tumor regions by using pattern analysis of time-intensity curve (TIC) in multi-phase MDCT images. Project collaborators include oncologists specialized in targeted therapy for HCC (particular on antiangiogenic treatment) at the MGH Cancer Center, and imaging scientists specializing in quantitative imaging analysis from the 3D Imaging Lab at the MGH.
The specific aims of the project are: (1) Development of HTV biomarker: We will develop the prototype HTV biomarker on a cloud-computing platform, including HTV-Server that is an automated HTV post-processing pipeline for the registration, classification and segmentation of liver, tumors and viable/necrotic tumor regions in multi-phase MDCT images, and HTV-Client that is a web-based user interface for interactive visualization, quantitative analysis, and point-of-care data access for the HTV quantification. (2) Evaluation of HTV biomarker: We will conduct a clinical study to evaluate the clinical performance of the proposed HTV biomarker in the prognosis of tumor progress and treatment response by using 50 advanced HCC cases treated with antiangiogenic therapies at the MGH Cancer Center. (3) Plan of project Phase II: We will establish the company processes to meet FDA regulations for translation and validation of the product in Phase II of the project. Our industrial collaborators (TeraRecon Inc, Intrasense) have shown their high level of interests in licensing or purchasing the proposed technology to integrate it into their medical imaging workstations.
Tumor quantification is crucial for diagnosis, staging, and treatment evaluation, and multi-detector CT (MDCT) is often used because of its relatively low cost and wide availability. However, size-based MDCT tumor quantification using either the traditional linear measurement or the more advanced volumetric measurement has demonstrated significant limitations due to the lack of tumor viability information. The MDCT imaging biomarker proposed for this project, hepatic tumor viability (HTV), provides automated quantification of tumor viability in addition to the size of liver and liver tumors with using multple-phase hepatic MDCT images for reliable assessment of tumor progression and evaluation of treatment response for patients with liver cancers.
|Lu, Difei; Wu, Yin; Harris, Gordon et al. (2015) Iterative mesh transformation for 3D segmentation of livers with cancers in CT images. Comput Med Imaging Graph 43:1-14|