Prospective Study of Non-Contrast, Abbreviated MRI for Hepatocellular Carcinoma Surveillance in Patients with Suboptimal Hepatic Visualisation on Ultrasound

Simple Summary Patients with chronic liver disease and cirrhosis are at risk of developing liver cancer (HCC). Regular ultrasound screening for HCC is recommended for these patients so that HCC can be found early. However, ultrasound is not always effective at picking up small cancers, especially in patients who are overweight or obese. Other tests including CT and MRI are expensive. Our study looked at a shorter version of an MRI scan (abbreviated MRI, or aMRI) in thirty patients who had recently had an ultrasound with poor views of the liver. All thirty patients tolerated the aMRI scan well. In these patients, the aMRI scan found one HCC and five other liver abnormalities which had not been picked up on ultrasound. Experts evaluated the aMRI scans and felt they were of good quality. Our study shows that aMRI is possible and useful in patients undergoing screening for HCC, especially those who have had poor views on an ultrasound.Abstract Background: Biannual ultrasound (US) is recommended for hepatocellular carcinoma (HCC) surveillance in patients with cirrhosis. However, US has limited sensitivity for early-stage HCC, particularly in overweight cohorts, where hepatic visualisation is often inadequate. Currently there are no robust imaging surveillance strategies in patients with inadequate US visualisation. We investigated the ability of non-contrast, abbreviated magnetic resonance imaging (aMRI) to adequately visualise the liver for HCC surveillance in patients with previously inadequate US. Methods: Patients undergoing US surveillance, where liver visualisation was inadequate (LI-RADS VIS-B and VIS-C), were prospectively recruited. Patients underwent non-contrast T2-weighted and diffusion-weighted aMRI. The images were reviewed and reported by an expert liver radiologist. Three independent, blinded radiologists assessed the aMRI visualisation quality using a binary score assessing five parameters (parenchymal definition, vascular definition, coverage of the liver, uniformity of liver appearance and signal-to-noise ratio). Results: Thirty patients completed the aMRI protocol. The majority (90%) had underlying cirrhosis and were overweight (93.3%), with 50% obese and 20% severely obese. A total of 93.3% of the aMRI scans were of satisfactory quality. Six patients (20%) had hepatic abnormalities detected with aMRI that were not seen on their US: one HCC, one haemangioma and three clinically insignificant lesions. For the aMRI visualisation quality assessment, the coverage of the liver, vascular definition and parenchymal definition were consistently rated to be of sufficient quality by all three radiologists. Conclusions: Non-contrast aMRI provided good visualisation of the liver and detection of abnormalities in patients with inadequate US. aMRI should be further explored in a larger, prospective study as an alternative surveillance strategy in patients with inadequate US.