RSNA 2017

Abstract Archives of the RSNA, 2017


SST04-06

Diagnostic Performance of an Abbreviated Gadoxetic Acid-Enhanced MRI Protocol for Colorectal Cancer Liver Metastases Surveillance

Friday, Dec. 1 11:20AM - 11:30AM Room: E353B



Participants
Rodrigo Canellas, MD, Boston, MA (Presenter) Nothing to Disclose
Aoife Kilcoyne, MBBCh, Boston, MA (Abstract Co-Author) Nothing to Disclose
Dushyant V. Sahani, MD, Boston, MA (Abstract Co-Author) Research support, General Electric Company; Medical Advisory Board, Allena Pharmaceuticals, Inc

For information about this presentation, contact:

rcsouza@partners.org

PURPOSE

To evaluate the diagnostic performance of an abbreviated gadoxetic acid-enhanced MRI protocol in colorectal cancer (CRC) liver metastases surveillance.

METHOD AND MATERIALS

In this IRB approved retrospective study, gadoxetic acid-enhanced (Eovist, Bayer healthcare) MRI scans of 43 patients with histologically proven liver metastases from colorectal cancer (mCRC) were assessed. Two readers independently assessed two sets of images of each patient. The first set consisted of T2WI + 15-20 min delayed T1-WFS followed by DWI (b values ranging from 50 to 800 s/mm2) and ADC maps. Readers evaluated liver lesions and characterized them as being or malignant. A maximum number of 10 lesions per patient were recorded. Malignant classification was based on low signal on the 15-20 min T1-FS, high signal on the DWI and low signal on ADC. Benign diagnosis was based on T2 bright, hypo-intense 15-20 min T1-FS and no restricted diffusion. The full MRI protocol (set 1 + set 2 + dynamic study + other sequences) was used as our reference standard. Cohen kappa analysis was used to assess the agreement between the readers.

RESULTS

Reader 1 characterized 184 lesions as malignant and reader 2 characterized 200 lesions as malignant with a good inter-reader agreement. Although useful in lesion characterization, DWI did not add much in terms of lesion detection compared to hepatobiliary phase T1-WFS. Full MRI exam with dynamic study was found useful in characterizing previously treated 3 small metastatic lesions, but did not improve the detection of metastatic liver lesions nor the characterization of few small lesions deemed indeterminate by the abbreviated protocol. The mean time for an abbreviated protocol takes around 15 minutes to be completed compared to 35 minutes of the the full MRI protocol.

CONCLUSION

An abbreviated gadoxetic acid-enhanced MRI protocol (T2WI, DWI and 15-20 min delayed T1FS) is faster, has robust quality and provides high diagnostic performance in detection and characterization of liver lesions in mCRC surveillance.

CLINICAL RELEVANCE/APPLICATION

MRI is a scarce resource and has high demand due to the clinical benefits in variety of clinical settings. Workflow strategies need to be engineered to meet the clinical demands without compromising the diagnostic performance of MRI. An abbreviated liver MRI protocol can serve as potentially faster and lower-cost alternative to conventional MRI protocol.