Direct Radiologic-Pathologic Correlation of Colorectal Liver Metastases with MR-Compatible Sectioning and Localization Device

The PI of this project was:

This project was funded by: Radiology RD

The term of this project was: February 2018 to May 2020

The number of subjects scanned during this project was: 30

Patients with colorectal cancer (CRC) liver metastases derive a significant survival benefit from surgical resection of liver metastases. However, accurate pre-operative detection of all metastatic lesions is essential to achieve a survival benefit. Lesions described as “indeterminate” are problematic for pre-operative planning. Advances in imaging technologies, particularly with gadoxetic acid-enhanced magnetic resonance (MR) imaging, have improved sensitivity for small liver lesion detection. However, significant limitations exist for characterizing these lesions. Accurate radiologic-pathologic correlation of lesions, the gold standard, remains challenging. In order to improve characterization of small liver lesions with accurate histopathologic diagnoses, our group developed an MR-compatible sectioning device for ex vivo imaging of liver specimens. The specimens are stabilized and imaged in the device and then sectioned according to the device’s radio-opaque grid, performing precise radiologic-pathologic correlation of liver lesions. Our group previously validated our design using swine liver and demonstrated 100% accuracy of lesion identification in this feasibility study.

Our long-term goal is to establish definitive histologic characterization of radiologically indeterminate liver lesions in CRC patients. In the current project, our objective is to validate our novel method for radiologic-pathologic correlation in the characterization of liver lesions in CRC patients. Specifically, we aim to establish rigorous one-to-one correlation of radiologic lesions with lesions identified on pathology in resected human liver specimens. Secondly, we aim to determine the accuracy of ex vivo MR imaging liver lesion diagnosis using direct pathologic correlation as the reference standard.
We expect our device will provide precise correlation of liver lesions and identify radiologic inaccuracies in lesion characterization. While this technique has obvious application to patients with colorectal cancer and liver metastases, the potential impact extends to other organ systems where technological advances in radiology have outpaced the resolution of standard gross anatomic sectioning techniques.