Page 75 - Read Online

Page 4 of 15 Osho et al. Hepatoma Res 2020;6:55 I http://dx.doi.org/10.20517/2394-5079.2020.42

detection and found GALAD to be superior to ultrasound, with an area under the curve (AUC) of 0.95
(95%CI: 0.93-0.97) vs. 0.82 (95%CI: 0.77-0.87), respectively. When the GALAD model was combined with
ultrasound, the GALADUS score had significantly better performance compared to ultrasound alone with
a sensitivity of 95%, specificity of 91% and an AUC of 0.98 (95%CI: 0.96-0.99). In addition, there has also
been increased interest in methylated DNA markers and circulating tumor cells (CTCs) for early detection
of HCC [32,33] . However, these biomarkers still require evaluation in large phase II and phase III studies
[34]
before adoption in clinical practice .
CT/MRI for HCC surveillance
Given the limitations of ultrasound-based surveillance, there has been increasing interest in alternative
imaging modalities, such as computed tomography (CT) or magnetic resonance imaging (MRI), but
neither is recommended by current practice guidelines [2,3,6,7] . Although both CT and MRI have been
shown to be superior in sensitivity and specificity for HCC diagnosis and staging compared to ultrasound
(discussed below), there are limited data evaluating these tests in a surveillance manner. A small
randomized trial comparing semi-annual ultrasound to annual multiphase CT found that ultrasound was
[35]
similar in sensitivity but less costly than CT . Further, CT is associated with screening harms including
radiation exposure and potential contrast injury [36,37] . Therefore, there has been increasing interest in
MRI surveillance, which obviates some of these concerns. A prospective cohort study from South Korea
(PRIUS study) comparing ultrasound and MRI surveillance in patients with cirrhosis found that MRI had
significantly higher sensitivity for early stage HCC (86% vs. 27.9%) as well as higher specificity (97% vs.
[38]
94.4%) . The authors of this study subsequently performed a cost-effectiveness analysis, suggesting that
MRI may be cost-effective; however, these data still require validation in non-HBV Western populations .
[39]
Furthermore, there would be concerns about radiologic capacity and patient acceptability if an MRI-based
strategy were adopted in larger populations. There have been increasing data on alternative MRI strategies,
including abbreviated MRI and non-contrast MRI. Abbreviated MRI protocols use selected sequences from
a full diagnostic protocol and can shorten the examination from ~45 min to ~15 min, which may address
[40]
some concerns about radiologic capacity and improve cost-effectiveness . Abbreviated MRI protocols
have been studied for HCC diagnosis and characterization of lesions [40,41] , but no trials or studies have been
done specifically for surveillance. There is an ongoing clinical trial at Seoul National University Hospital
comparing annual abbreviated MRI to ultrasound for early HCC detection (NCT03731923). Two recent
studies have also evaluated non-contrast MRI as a possible surveillance strategy. A post-hoc analysis of
the PRIUS study suggested that non-contrast MRI is superior to ultrasound for HCC detection, with per-
lesion and per-examination sensitivity of 77.1% and 79.1% for non-enhanced MRI compared to just 25.0%
[42]
and 27.9% for ultrasound, respectively . Specificity of non-contrast MRI was also higher than that of
ultrasound 97.9% vs. 94.5%, P < 0.001]. In addition, the estimated scan time was < 6 min with a total room
occupancy time of only 25-35 min. Two ongoing prospective trials, MIRACLE-HCC (NCT02514434) and
MAGNUS-HCC (NCT02551250), are comparing non-contrast MRI and ultrasound for surveillance of
HCC [43,44] .

Most analyses for HCC surveillance have tried to implement a “one-size-fits-all” strategy for all at-risk
patients, despite known variation in HCC risk between patients with cirrhosis. For example, a validated
tissue-based signature has been shown to accurately risk stratify patients with cirrhosis into high,
[45]
intermediate, and low risk of HCC, with annual HCC incidences of 5.8, 2.2, and 1.5%, respectively .
Similarly, other risk stratification markers can accurately distinguish patients with high risk and low risk of
[46]
developing HCC . Accurate risk stratification could allow more intensive and costly surveillance strategies
to be applied to those at highest risk, while using lower intensity and inexpensive surveillance strategies in
lower risk patients. A modeling study suggested that a risk-stratified approach was cost-effective compared
to ultrasound and AFP in all patients . Currently, the Japan Society of Hepatology (JSH) is the only
[47]
professional society that recommends a differential HCC surveillance strategy by individual patient risk, i.e.,