Abstract
A variety of patterns of enhancement of liver lesions and liver parenchyma is observed in the hepatobiliary phase (HBP) of gadoxetic acid-enhanced MRI. It is becoming increasingly apparent that many lesions may exhibit HBP enhancement. Much of the literature regarding the role of gadoxetic acid-enhanced MRI in characterising liver lesions is dichotomous, focusing on whether lesions are enhancing or non-enhancing in the HBP, rather than examining the patterns of enhancement. We provide a pattern-based description of HBP enhancement of liver parenchyma and of liver lesions. The role of OATP1B3 transporters, hepatocyte function and lesion composition in influencing patterns of HBP hyperintensity are discussed.
Background
Gadoxetic acid
Gadoxetic acid is a hepatocyte-specific contrast agent which is useful in differentiating focal nodular hyperplasia (FNH) from other arterial phase enhancing lesions. FNH is typically HBP hyperintense, whilst other arterial phase enhancing lesions, primarily adenoma, are predominantly HBP hypointense. It is also useful in the detection of liver metastases and, in some countries, gadoxetic-enhanced MRI is used for the detection of hepatocellular carcinoma (HCC).
Gadoxetic acid and gadoxetate disodium (Primovist or Eovist, Bayer Healthcare) are hepatocyte-specific contrast agents, characterised by initial extracellular distribution followed by selective uptake by functioning hepatocytes. Hepatocyte excretion is 50%. Gadobenate dimeglumine (MultiHance, Bracco Diagnostics) is another widely used hepatocyte-specific contrast agent which has increased extracellular distribution and 3–5% is excreted by hepatocytes.
Gadoxetic acid is actively transported into functioning hepatocytes via the organic anion transporting polypeptide 1B3 (OATP1B3, synonymous with OATP8) resulting in enhancement in the hepatobiliary phase (HBP). OATP1B3 transporters are considered to be the main uptake transporter of gadoxetic acid in hepatocellular nodules, which are present on the hepatocyte basolateral membrane1 and influence the signal intensity in the HBP.2 A positive correlation between the amount of OATP1B3 transporter expression and enhancement in the HBP has been verified in FNH, HCC, hepatocellular adenoma (HCA) and dysplastic nodules.2
Protocol and hepatic enhancement
In addition to conventional sequences, a 3D T1 weighted fat-saturated gradient echo pre-contrast sequence is obtained. Gadoxetic acid (Primovist) is administered at a dose of 0.1 ml/kg (maximum dose 10 ml) and is injected at a rate of 1 ml/s, followed by 3D T1 weighted fat-saturated gradient echo sequences with multiplanar acquisition, with or without subtraction.
Arterial phase imaging tends to be less intense than with extracellular agents due to the smaller dose of extracellular agent and higher frequency of imaging artefacts in the arterial phase when compared with extracellular agents. These artefacts are poorly understood but possibly relate to abrupt concentration changes during high spatial k-space filling that may alter intensity and anatomical detail of structures.3 The gadoxetic acid entry into the liver cells causes intense parenchymal enhancement beginning 1–2 min after it is administered. There is no true extracellular phase.4 The transitional phase at 2–5 min represents the transition from extracellular-dominant (portal venous phase) to intracellular-dominant phase (HBP).4 Liver enhancement progresses until it reaches a peak at 10–120 min after injection. It is then excreted from these cells into biliary canaliculi.4
Bilirubin is taken up competitively by OATP1B3; therefore, the degree of biliary excretion depends on liver function. Hepatocellular enhancement of the liver tends to be less intense in patients with hyperbilirubinaemia.5
Adequacy of enhancement in the HBP is determined by unequivocal hyperintensity of the liver parenchyma relative to the blood vessels (Figure 1a).6 Visible excretion into the bile duct alone does not equate to adequate liver enhancement.6
Figure 1.
Patterns of hepatic parenchymal enhancement in the 20 min delayed hepatobiliary phase of gadoxetic acid-enhanced MRI in three different patients: (a) normal enhancement, (b) heterogeneous enhancement in a patient with cirrhosis and (c) periportal enhancement in a different patient with cirrhosis. Splenomegaly related to portal hypertension is noted in cases (b) and (c, not shown).
Patterns of HBP enhancement of the hepatic parenchyma (non-lesional)
Diffuse heterogeneous enhancement
Cirrhotic livers may exhibit heterogeneous enhancement in the HBP (Figure 1b). The heterogeneity directly correlates with Child-Pugh class. In a study by Tamada et al,7 homogeneous liver enhancement was observed more frequently in patients with Child-Pugh class A cirrhosis, while the heterogeneous pattern was seen more frequently in patients with Child-Pugh Class C.7 Gadoxetic acid uptake ratios in hepatocytes negatively correlates with liver function8 and hepatocyte HBP enhancement decreases as liver fibrosis progresses.9 The mechanism for this is not well understood but is likely to reflect the reduced number of functional hepatocytes or dysfunctional cellular transport mechanisms.6
Periportal enhancement pattern
HBP enhancement unrelated to a focal lesion may be seen along the portal tracts in approximately 3% of patients with an underlying hepatic disorder (Figure 1c). This particularly occurs in primary biliary cirrhosis and in cases of idiopathic portal hypertension. This pattern of enhancement is thought to represent regenerative changes in periportal hepatocytes, reflecting relatively increased enhancement compared to the damaged background liver, rather than an absolute increase in the degree of enhancement in the HBP.10 The mechanisms for the preservation of function of the periportal hepatocytes is not known. It possibly relates to differences in gene expression, which is influenced by the hepatocyte location along the porto-centrovenular axis of the hepatic lobule.11
Patterns of HBP enhancement of liver lesions
Diffuse homogeneous HBP enhancement
FNH
FNH is the most common cause of an isointense or hyperintense lesion in the HBP in patients without chronic liver disease.12 It is not a true tumour, but a polyclonal tumour-like lesion.13 It is typically asymptomatic and found incidentally, most frequently in females of childbearing and middle age.14 These lesions have no known malignant potential and result in few, if any, complications.14 Some cases are thought to undergo self-regression.15 In the majority of cases, FNH is a solitary lesion, usually smaller than 5 cm and occurs near the liver surface.16 In approximately 20% of cases, multiple lesions are present.17
The development of FNH is thought to be related to a hyperplastic response of the liver to focal increase in blood flow, secondary to a pre-existing arterial malformation.18 It was previously theorised that the development of FNH was hormonally dependent and related to oral contraceptive use, however recent studies suggest that oestrogens are not implicated in their development nor growth.19
On conventional MRI sequences, the typical appearances of FNH include: iso-intensity or mild hypointensity on T1 weighted sequences and iso-intensity or mild T2 hyperintensity. On post-contrast sequences with extracellular contrast, there is marked arterial phase enhancement and iso-intensity or mild hyperintensity in the portal venous phase (Figure 2).20
Figure 2.
Gadoxetic acid-enhanced MRI in a patient with focal nodular hyperplasia in segment five which demonstrates typical MRI characteristics. The lesion is isointense on the T1 (a) and T2 weighted (b) sequences, with marked arterial (c) and portal venous phase (d) enhancement. In the hepatobiliary phase (e), the lesion is homogeneously hyperintense.
A central T2 hyperintense scar is present in approximately 85% of cases on MRI (Figure 3).21 Atypical appearances of FNH on conventional MRI sequences include an absent central scar. T1 hyperintensity may be attributed to copper accumulation, high protein concentration, blood degradation products or sinusoidal dilatation.13T1 hyperintensity may also be due to intralesional steatosis, which is a rare finding in FNH.13
Figure 3.
Gadoxetic acid-enhanced MRI of a large, biopsy proven FNH in the right lobe of the liver. The initial study shows heterogeneous T2 hyperintense signal (a) in the right lobe of the liver with a very small T2 hyperintense central scar (arrow). The lesion shows moderate to marked arterial enhancement (b) and heterogeneous signal with a hyperintense rim in the 20 min hepatobiliary phase (HBP) (c). On a repeat MRI 5 years later (d), the T2-hyperintense scar (arrow) is larger. The lesion appears more isointense in the 20 min HBP (e) when compared to the prior study.
Approximately 90% of FNH show hyperintensity in the HBP.17This correlates with equal or stronger OATP1B3 transporter expression compared with background liver. The β-catenin pathway (Wnt/CTNNB1 pathway) is activated in FNH, which results in polyclonal over proliferation of hepatocytes. The β-catenin pathway is a major molecular pathway of oncogenesis22 and is possibly activated by the perivascular hypoxic conditions. It is involved in regulation of OATP1B3 transporter expression. Whilst the OATP1B3 transporter expression correlates with the degree and pattern of HBP enhancement with hepatocyte-specific contrast on MRI, other factors such as the molecular background and tissue components may also influence the imaging findings.23 The presence of abnormal bile ductules that fail to communicate with the normal biliary system possibly results in defective or delayed contrast excretion, which may contribute to persistent contrast retention.13
Many different patterns of enhancement of FNH in the HBP of gadoxetic acid-enhanced MRI have been described. The most frequently reported pattern of enhancement varies between studies. Mohajer et al14 reported the most common pattern of enhancement as homogeneous isointensity or hyperintensity (Figure 4).
Figure 4.

Gadoxetic acid-enhanced MRI in a patient with focal nodular hyperplasia which exhibits diffuse hyperintensity in the hepatobiliary phase.
FNH-like nodules
FNH-like lesions are observed in 3.4% of cirrhotic livers.24 They are microscopically and immunohistochemically identical to classic FNH in non-cirrhotic livers3 and are thought to result from an acquired hyperplastic response to vascular alterations associated with cirrhosis.3 The radiological features are the same as those described for FNH.
Regenerative nodules in cirrhosis
The vast majority of cirrhosis-related nodules exhibit regenerative changes without cellular atypia and are known as regenerative nodules. A minority have dysplastic features. They are typically isointense or hyperintense on T1 weighted images, isointense to hypointense on T2 weighted images and isointense on diffusion-weighted sequences. If iron containing, they are markedly T2 hypointense. If lipid containing, loss of signal on oppose-phase imaging compared to in-phase gradient echo imaging is seen.3 With extracellular contrast, they show similar enhancement to adjacent liver in arterial and venous phases. Occasional enhancement in the arterial phase is seen with fading to isointensity in the portal venous phase. Regenerative nodules tend to be iso or mildly hyperintense to background liver in the HBP (Figure 5) and usually enhance homogeneously. Occasionally, small HCCs may have a similar appearance;3 however, regenerative nodules are usually differentiated by their iso-intensity in the arterial and portal venous phases.5
Figure 5.
Gadoxetic acid-enhanced MRI in a patient with cirrhosis and a segment six regenerative nodule that was stable over 5 years of follow-up. On pre-contrast T1 fat-saturated sequence (a), the nodule is mildly hyperintense and exhibits diffuse homogeneous hyperintensity in the hepatobiliary phase (b).
Diffuse heterogeneous enhancement
FNH and FNH-like nodules
As mentioned, the frequency of reported patterns of HBP enhancement of FNH varies between studies. An et al25 describe heterogeneous hyper or iso-intensity as the one of the most common patterns of HBP enhancement (Figure 6a). The degree of heterogeneity is variable. There may be only a mild degree of heterogeneity with a reticulated or lace-like appearance (Figure 6b).3 More marked heterogeneity in HBP enhancement may be related to the central scar as well as the presence of fibrous septa that divide the lesion into multiple smaller nodules, resulting in a micro-lobulated margin and the lesion may resemble the heterogeneous HBP enhancement seen in a cirrhotic liver. This accounts for FNH previously being described as ‘focal nodular cirrhosis’ in the pathology literature.26 With gadoxetate disodium, FNH is more often heterogeneously enhancing than with gadobenate dimeglumine. The central scar of FNH is consistently hypointense with both contrast agents.27
Figure 6.
Hepatobiliary phase of gadoxetic acid-enhanced MRI in two different patients with focal nodular hyperplasia showing marked heterogeneous hyperintensity (a) and mild heterogeneous hyperintensity with a lace-like appearance (b). A central scar is seen in both lesions.
Hepatocellular adenoma
HCAs are a heterogeneous group of tumours characterised by specific genetic, pathological abnormalities and tumour biology.28 HCAs are the main differential diagnosis of FNH. It is important to distinguish between these conditions due to the risk of complications of HCA, such as haemorrhage and malignant transformation.
Five subtypes of HCAs have been described based on their molecular signatures, including the recently described sonic hedgehog-activated HCA (4–5%).29 The other subtypes are hepatocyte nuclear factor 1-α-mutated HCA (35%), β-catenin-mutated HCA (10–15%), inflammatory HCA (35–45%) and unclassified HCA (7%). The HNF 1α-mutated HCA and inflammatory HCA subtypes are seen in the context of oral contraceptive use. The inflammatory subtype is also associated with obesity and alcohol intake. The β-catenin-mutated HCA subtype is typically found in males taking anabolic steroids or in the context of glycogen storage disease.30 The risk of malignant transformation ranges from 4.2 to 13% cases and is highest in the β-catenin-mutated subtype.29
On conventional MRI sequences, the presence of haemorrhage and intralesional fat suggests adenoma rather than FNH, although, as mentioned, FNH can rarely contain fat. Steatosis within the lesion is seen particularly in the hepatocyte nuclear factor 1-α-mutated subtype.31 A rim of T2 hyperintensity, termed the atoll sign, which may show enhancement in the delayed portal venous phase is thought to be characteristic of inflammatory adenomas.32 They demonstrate strong arterial phase enhancement and persistent portal venous/transitional phase enhancement.2 β-HCAs are often poorly defined, are heterogeneous on T2 weighted sequences and demonstrate mild-to-moderate hypervascularity.33
MRI with hepatocellular-specific contrast agents has previously been thought to confer a sensitivity of greater than 90% for distinguishing FNH from HCA.34Purysko et al35 demonstrated a lower average enhancement ratio for adenoma compared to FNH. HCA enhanced to a lesser degree than adjacent liver in the HBP with a 97% accuracy for differentiating from FNH when the enhancement ratio was 0.7. However, in a systematic review of the literature by McInnes et al.,32 the diagnostic accuracy of HBP gadoxetic acid-enhanced MRI in discriminating HCA from FNH is likely to be overestimated, as the studies are heterogeneous and few in number.
It has been shown that parenchymal expression of OATP1B3 transporters is minimal or absent in hepatic adenoma,14 which accounts for 79% of HCAs appearing hypointense in the HBP of gadoxetic acid MRI.1 β-catenin-mutated HCAs, however, may exhibit iso- or hyperintensity in the HBP (Figure 7), which reflects the relationship between the activation of the β-catenin pathway and OATP1B3 transporter expression.1 10% of inflammatory HCAs have β-catenin activation and may demonstrate enhancement in the HBP.2
Figure 7.
Gadoxetic acid-enhanced MRI in a male patient on anabolic steroids with a biopsy proven β-catenin positive hepatocellular carcinoma (HCC) arising in an adenoma. HCCs are typically hepatobiliary phase (HBP) hypointense, however 8% may exhibit HBP enhancement, in particular, HCC with the β-catenin mutation. The lesion is T1 hypointense (a) and shows mixed areas of hypointensity and hyperintensity in the HBP (b) with a nodule-in-nodule appearance (short arrow).
HBP-enhancing adenomas may mimic FNH, particularly in β-catenin-mutated HCAs which contain a central scar in 75% cases.2 Several patterns of HBP enhancement in adenomas have been described; however, there is a paucity of literature regarding the frequency of these patterns. In our experience, diffuse heterogenous enhancement is most commonly seen (Figure 8). In a minority of cases, the differentiation between FNH and adenoma remains challenging and definitive diagnosis may require biopsy or resection.5
Figure 8.
Gadoxetic acid-enhanced MRI in a young female with biopsy confirmed hepatocellular adenoma. Most adenomas are hepatobiliary phase (HBP) hypointense. A minority of adenomas may occasionally demonstrate HBP hyperintensity. T1FS pre-contrast (a) and HBP (b) sequences demonstrate a large, predominantly right lobe mass which is T1 heterogeneously iso/hypointense and demonstrates heterogeneous HBP internal hyperintensity with an irregular, lobulated hyperintense rim.
Hepatocellular carcinoma (HCC)
HCC is the most common primary malignant hepatic tumour, predominantly arising on a background of cirrhosis in the context of viral hepatitis B, C, alcoholism and non-alcoholic steatohepatitis (NASH).3 On conventional MRI sequences, HCCs are T1 hypointense although they may be T1 hyperintense, particularly if they are haemorrhagic. HCCs are typically T2 hyperintense. A mosaic architecture may be present, which is described as randomly distributed internal nodules that have different imaging features.6 Similarly, a nodule-in-nodule pattern represents a well-circumscribed inner nodule with different imaging features to the larger, outer nodule.6 Steatotic lesions demonstrate signal loss on out-of-phase gradient echo sequences, relative to in-phase gradient echo sequences.3 With extracellular contrast, 80% HCC exhibit intense arterial enhancement and washout in the portal venous and delayed phases.3 Restricted diffusion and an enhancing capsule may be observed.6
On gadoxetic acid MRI, the washout is only assessed in the portal venous phase and not in the transitional phase as hepatocyte enhancement may result in a pseudowashout appearance of the lesion.4 However, it has been suggested that extending the washout appearance to the transitional or HBP phase allows higher sensitivity without a significant reduction in specificity in diagnosing HCC.36 According to the Consensus Report from the Eighth International Forum for Liver Magnetic Resonance Imaging,37 further research is needed to define the physiological beginning and end of the transitional phase based on imaging features rather than fixed time points.
Gadoxetic acid-enhanced MRI is variably used in the investigation of HCC. Geographical differences exist between guidelines which are driven largely by differences in treatment practices. In North America and Europe, the greatest concern is for high specificity, since patients with a diagnosis of HCC may undergo liver transplantation based on imaging criteria alone. Strict diagnostic criteria are used to avoid false-positive HCC diagnoses. In contrast to this approach, in Asia, the primary aim is to maximise the sensitivity of HCC diagnosis. This is justified by the greater use of locoregional ablative therapies in Asia. In the Asian Pacific Association for the Study of the Liver (APASL) guidelines, gadoxetic acid–enhanced MRI is preferred over extracellular contrast media-enhanced MRI as a first-line diagnostic test.37
The Liver Imaging Reporting and Data System (LIRADS) 2018,6 endorsed by the American College of Radiology, standardises interpretation of the imaging features of hepatic lesions in patients who are at risk for HCC.4 It provides guidance for use of gadoxetic acid-enhanced MRI; however, it does not recommend hepatocyte-specific contrast over conventional extracellular contrast agents. LIRADS suggests that patient factors, institutional and radiologist expertise require consideration when determining the choice of contrast agent.6 According to LIRADS, HBP hypointensity of a lesion is an ancillary feature favouring malignancy and HBP isointensity is an ancillary feature that favours benignity (Table 1).6
Table 1.
Hepatobiliary phase enhancement patterns of the liver parenchyma and liver lesions on gadoxetic acid MRI
| Liver parenchyma | |
| HBP enhancement pattern | Conditions |
| Heterogeneous enhancement | Hepatic fibrosis |
| Periportal | Hepatic disorders: Primary biliary cirrhosis, idiopathic portal hypertension, liver cirrhosis, chronic hepatitis |
| Liver lesions | |
| HBP enhancement pattern | Conditions |
| Diffuse homogeneous enhancement | FNH and FNH-like lesions Regenerative nodules |
| Diffuse heterogeneous enhancement | FNH and FNH-like lesions HCA (although majority are HBP hypointense) HCC: nodule-in-nodule, mosaic patterns (although majority are HBP hypointense) |
| Rim enhancement | Multiacinar cirrhotic nodules: doughnut pattern HCA: irregular (although majority are hypointense) Metastases: perilesional enhancement - neuroendocrine tumour, gastrointestinal stromal tumours (although majority are HBP hypointense) |
| Central enhancement | Adenocarcinoma metastases: targetoid appearance Intrahepatic cholangiocarcinoma: targetoid appearance |
Intrahepatic cholangiocarcinoma: targetoid appearance
8% of HCCs are iso- or hyperintense in the HBP of gadoxetic acid MRI (Figure 9), which has been shown to be related to OAT1B3 expression.38 HCCs with β-catenin mutation show greater enhancement in the HBP.38β-catenin activation due to the β-catenin mutation is observed in 30–40% of patients with HCC39 and is thought to confer a relatively favourable prognosis compared to HCCs without the β-catenin mutation.39,40 Most HBP hyperintense HCCs are well or moderately differentiated.3
Figure 9.
Gadoxetic acid-enhanced MRI in a patient with pathologically confirmed hepatocellular carcinoma in segment 4B. HCCs are typically hepatobiliary phase (HBP) hypointense, however a minority may show HBP hyperintensity. The lesion is hypointense on the fat saturated T1 weighted sequence (a). It shows heterogeneous enhancement in the arterial phase (b), with heterogeneous washout in the portal venous phase (c). In the 20 min delayed HBP (d), the lesion (long arrow) is predominantly hypointense with small areas of internal hyperintensity (arrowhead).The hyperintense rim along the right lateral margin (short arrows) represents normal enhancement of the adjacent liver parenchyma.
In a retrospective study by Suh et al.41which included 16 HBP hyperintense HCC, 69% showed a focal defect in enhancement in the HBP and 75% demonstrated a nodule-in-nodule appearance. 75% exhibited an HBP hypointense rim which corresponded to peritumoral capsules on pathology. Internal septations may be seen. There may also be a mosaic pattern with areas of mixed hyper and hypointensity in the HBP (Figure 10).
Figure 10.
Gadoxetic acid-enhanced MRI in a patient with a pathologically confirmed hepatocellular carcinoma in segment 4B. The lesion is hypointense on the fat-saturated T1 weighted sequence (a) and heterogeneously hyperintense on the T2 weighted sequence (b). In the hepatobiliary phase (HBP), the lesion is predominantly hypointense with areas of heterogeneous hyperintensity, especially along its lateral aspect, with a mosaic pattern. The HBP hypointense rim likely reflects the presence of a capsule.
Central HBP enhancement
Adenocarcinoma metastases and intrahepatic cholangiocarcinoma
On conventional MRI sequences, adenocarcinoma metastases and intrahepatic cholangiocarcinoma tend to be T1 hypointense, T2 hyperintense and have restricted diffusion.42 With extracellular contrast, rim enhancement may be seen in the arterial phase. Washout in the outer parts of the lesion in the portal venous and delayed phases occurs with progressive enhancement towards the centre of the lesion.3 Areas of prolonged enhancement of the tumour with extracellular contrast correspond to contrast retention in fibrotic stroma.3 Similar findings occur with gadoxetic acid due to its extracellular properties.4
Gadoxetic acid-enhanced MRI is widely used in the investigation of liver metastases as most are hypointense in the HBP. When compared with conventional MRI, the sensitivity in the detection of colorectal liver metastases is 95 vs 87%.42
Tumours with fibrotic stroma such as metastatic adenocarcinoma and cholangiocarcinoma may enhance in the HBP due to gadoxetic acid retention in the extracellular space.4 The central areas of the tumours may demonstrate ill-defined, cloud-like enhancement due to gradual accumulation of contrast in the central fibrotic stroma (Figure 11).43 This central hyperintensity in the HBP has been referred to as a targetoid pattern of enhancement.4 It has been suggested that this finding may be seen more commonly on MRI using gadobenate dimeglumine due to its more prominent extracellular properties.12
Figure 11.
Gadoxetic acid-enhanced MRI in two patients with lesions displaying targetoid pattern of hyperintensity in the hepatobiliary phase (HBP). The first patient has a history of breast cancer and developed FDG PET-avid lung (not shown) and liver nodules, which were presumed to represent metastases. The pre-contrast T1 fat-saturated sequence (a) demonstrates a hypointense lesion in segment 2/4a. Mild central hyperintensity is seen in the HBP sequence with a targetoid pattern (b), likely to be related to contrast pooling in fibrous tissue. The second patient has an intrahepatic cholangiocarcinoma. On the pre-contrast T1 fat-saturated sequence (c), the lesion is hypointense. Targetoid hyperintensity is seen in the HBP (d).
The hyperintensity caused by gadoxetic acid retention in the extracellular space tends to be lower compared to the hyperintensity caused by transported uptake of gadoxetic acid into hepatocytes.3
Rim HBP enhancement
FNH and FNH-like nodules
In addition to Mohajer et al.14 describing diffuse enhancement of FNH as mentioned, a rim-enhancing pattern was also reported – this consisted of either a hyperintense rim and a central core that is hypointense to liver (Figure 12a) or a hyperintense rim with a central core that was isointense (Figure 12b) or hyperintense to liver.14 On gadoxetic acid-enhanced MRI, the central scar of FNH is hypointense in the HBP27 and may display a stellate appearance.12 When central HBP hypointensity is observed in FNH, it may be attributed to both reduced enhancement of the scar as well as of the hepatocytes surrounding the scar.44 The latter finding is thought to be related to the relatively lower expression of OATP1B3 transporters in the hyperplastic hepatocytes surrounding the central scar.2 The ring enhancement in some FNH lesions, which has been described as a doughnut-like pattern on HBP imaging, has been shown to correspond to higher expression of OATP1B3 in the periphery of the lesion.38
Figure 12.
Hepatobiliary phase of gadoxetic acid-enhanced MRI in two different patients with lesions showing rim hyperintensity, both compatible with focal nodular hyperplasia. (a) The lesion in the first patient shows a hyperintense rim and a central core that is hypointense to liver. (b) In a different patient, the lesion has a hyperintense rim with a central core that is isointense to liver.
Doughnut-like nodules
HBP doughnut-like nodules (HBP-doughnut nodules) have been described in 6% of patients with liver cirrhosis and have been termed as ‘multiacinar cirrhotic nodules’. Unlike FNH, there is absent enhancement in the arterial phase. Enhancement in the portal venous phase is attributed to predominant portal venous supply.45 These nodules demonstrate a rim of hyperintensity in the HBP (Figure 13).
Figure 13.
Gadoxetic acid-enhanced MRI in a male patient with HBP-doughnut nodules. The pathology on the biopsy was described as nodular hyperplasia. Post-contrast sequences demonstrate multiple lesions with ring hyperintensity in the hepatobiliary phase (HBP) (a) with no arterial phase enhancement (b) and mild portal venous phase enhancement (c). Further lesions with HBP doughnut-like hyperintensity are seen more superiorly in the same patient (d).
Metastasis
A thin, irregular rim of HBP hyperintensity may infrequently be seen at the interface between a metastasis and normal liver parenchyma. The mechanism of this is not understood; however, it has been proposed that this is related to peritumoural hepatocellular reaction, biliary reaction or compressed normal hepatic parenchyma.12 This peritumoral HBP hyperintensity has occasionally been observed in neuroendocrine tumour metastases and gastrointestinal stromal tumour metastases.46 Yoneda et al46 suggests that this may indicate peritumoral hepatocyte hyperplasia with OATP1B3 expression and increased incidence of hepatic venous invasion. The hyperintense rim is observed adjacent to the periphery of the lesion rather than in the lesion itself.
Conclusion
One of the main uses of gadoxetic acid-enhanced MRI is to differentiate benign lesions which are HBP hyperintense, such as FNH, from tumours including HCA and HCC, which are commonly HBP hypointense. Although many of these liver lesions are able to be differentiated on conventional MRI sequences with extracellular contrast, occasionally these findings are indeterminate and gadoxetic acid-enhanced MRI may be used as a problem-solving tool. When lesions show hyperintensity in the HBP, the pattern of HBP hyperintensity may suggest a specific diagnosis. In particular, a doughnut-like pattern of HBP enhancement has been described in FNH, FNH-like lesions and multiacinar cirrhotic nodules but has not been found in HCAs or HCCs. The HBP pattern of hyperintensity may provide information regarding lesion architecture and composition, such as in fibrotic tumours which exhibit a central, cloud-like targetoid pattern of enhancement related to extracellular pooling of gadoxetic acid. Finally, patterns of enhancement of the liver parenchyma may provide information regarding hepatic function.
Footnotes
Acknowledgment: We thank Dr Hock Kua, pathologist, for reviewing the pathological findings in our cases.
Contributor Information
Cathryn L Hui, Email: Cathryn.Hui@monashhealth.org.
Marcela Mautone, Email: marcela.mautone@monashhealth.org.
REFERENCES
- 1.Reizine E, Amaddeo G, Pigneur F, Baranes L, Legou F, Mulé S, et al. Quantitative correlation between uptake of Gd-BOPTA on hepatobiliary phase and tumor molecular features in patients with benign hepatocellular lesions. Eur Radiol 2018; 28: 4243–53. doi: 10.1007/s00330-018-5438-7 [DOI] [PubMed] [Google Scholar]
- 2.Yoneda N, Matsui O, Kitao A, Kozaka K, Kobayashi S, Sasaki M, et al. Benign hepatocellular nodules: hepatobiliary phase of gadoxetic acid-enhanced MR imaging based on molecular background. Radiographics 2016; 36: 2010–27. Nov-Dec. doi: 10.1148/rg.2016160037 [DOI] [PubMed] [Google Scholar]
- 3.Campos JT, Sirlin CB, Choi J-Y. Focal hepatic lesions in Gd-EOB-DTPA enhanced MRI: the atlas. Insights Imaging 2012; 3: 451–74. doi: 10.1007/s13244-012-0179-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Fujita N, Nishie A, Asayama Y, Ishigami K, Ushijima Y, Kakihara D, et al. Hyperintense liver masses at hepatobiliary phase gadoxetic Acid–enhanced MRI: imaging appearances and clinical importance. Radiographics 2020; 40: 72–94. Jan-Feb. doi: 10.1148/rg.2020190037 [DOI] [PubMed] [Google Scholar]
- 5.Goodwin MD, Dobson JE, Sirlin CB, Lim BG, Stella DL. Diagnostic challenges and pitfalls in MR imaging with hepatocyte-specific contrast agents. Radiographics 2011; 31: 1547–68. doi: 10.1148/rg.316115528 [DOI] [PubMed] [Google Scholar]
- 6.American College of Radiology Liver Imaging Reporting and Data System version. 2018. Available from: https://www.acr.org/Clinical-Resources/Reporting-and-Data-Systems/LI-RADS [Accessed February 9, 2020].
- 7.Tamada T, Ito K, Higaki A, Yoshida K, Kanki A, Sato T, et al. Gd-EOB-DTPA-enhanced MR imaging: evaluation of hepatic enhancement effects in normal and cirrhotic livers. Eur J Radiol 2011; 80: e311–6. doi: 10.1016/j.ejrad.2011.01.020 [DOI] [PubMed] [Google Scholar]
- 8.Yoon JH, Lee JM, Kang H-J, Ahn SJ, Yang H, Kim E, et al. Quantitative assessment of liver function by using gadoxetic acid-enhanced MRI: hepatocyte uptake ratio. Radiology 2019; 290: 125–33. doi: 10.1148/radiol.2018180753 [DOI] [PubMed] [Google Scholar]
- 9.Nishie A, Asayama Y, Ishigami K, Tajima T, Kakihara D, Nakayama T, et al. Mr prediction of liver fibrosis using a liver-specific contrast agent: superparamagnetic iron oxide versus Gd-EOB-DTPA. J Magn Reson Imaging 2012; 36: 664–71. doi: 10.1002/jmri.23691 [DOI] [PubMed] [Google Scholar]
- 10.Kobayashi S, Matsui O, Gabata T, Koda W, Minami T, Kozaka K, et al. Intrahepatic periportal high intensity on hepatobiliary phase images of Gd-EOB-DTPA-enhanced MRI: imaging findings and prevalence in various hepatobiliary diseases. Jpn J Radiol 2013; 31: 9–15. doi: 10.1007/s11604-012-0136-x [DOI] [PubMed] [Google Scholar]
- 11.Colletti M, Cicchini C, Conigliaro A, Santangelo L, Alonzi T, Pasquini E, et al. Convergence of Wnt signaling on the HNF4alpha-driven transcription in controlling liver zonation. Gastroenterology 2009; 137: 660–72. doi: 10.1053/j.gastro.2009.05.038 [DOI] [PubMed] [Google Scholar]
- 12.Kim M-J, Rhee HJ, Jeong HT. Hyperintense lesions on gadoxetate disodium-enhanced hepatobiliary phase imaging. AJR Am J Roentgenol 2012; 199: W575–86. doi: 10.2214/AJR.11.8205 [DOI] [PubMed] [Google Scholar]
- 13.Khanna M, Ramanathan S, Fasih N, Schieda N, Virmani V, McInnes MDF. Current updates on the molecular genetics and magnetic resonance imaging of focal nodular hyperplasia and hepatocellular adenoma. Insights Imaging 2015; 6: 347–62. doi: 10.1007/s13244-015-0399-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Mohajer K, Frydrychowicz A, Robbins JB, Loeffler AG, Reed TD, Reeder SB. Characterization of hepatic adenoma and focal nodular hyperplasia with gadoxetic acid. J. Magn. Reson. Imaging 2012; 36: 686–96. doi: 10.1002/jmri.23701 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Kuo Y-H, Wang J-H, Lu S-N, Hung C-H, Wei Y-C, Hu T-H, SN L, TH H, et al. Natural course of hepatic focal nodular hyperplasia: a long-term follow-up study with sonography. J Clin Ultrasound 2009; 37: 132–7. Mar-Apr. doi: 10.1002/jcu.20533 [DOI] [PubMed] [Google Scholar]
- 16.Roncalli M, Sciarra A, Tommaso LD. Benign hepatocellular nodules of healthy liver: focal nodular hyperplasia and hepatocellular adenoma. Clin Mol Hepatol 2016; 22: 199–211. doi: 10.3350/cmh.2016.0101 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Grazioli L, Olivetti L, Mazza G, Bondioni MP. Mr imaging of hepatocellular adenomas and differential diagnosis dilemma. Int J Hepatol 2013; 2013: 1–20. doi: 10.1155/2013/374170 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Ferlicot S, Kobeiter H, Tran Van Nhieu J, Cherqui D, Dhumeaux D, Mathieu D, et al. Mri of atypical focal nodular hyperplasia of the liver: radiology-pathology correlation. AJR Am J Roentgenol 2004; 182: 1227–31. doi: 10.2214/ajr.182.5.1821227 [DOI] [PubMed] [Google Scholar]
- 19.Mathieu D, Kobeiter H, Maison P, Rahmouni A, Cherqui D, Zafrani ES, et al. Oral contraceptive use and focal nodular hyperplasia of the liver. Gastroenterology 2000; 118: 560–4. doi: 10.1016/S0016-5085(00)70262-9 [DOI] [PubMed] [Google Scholar]
- 20.Grazioli L, Bondioni MP, Haradome H, Motosugi U, Tinti R, Frittoli B, et al. Hepatocellular adenoma and focal nodular hyperplasia: value of gadoxetic acid-enhanced MR imaging in differential diagnosis. Radiology 2012; 262: 520–9. doi: 10.1148/radiol.11101742 [DOI] [PubMed] [Google Scholar]
- 21.Cha DI, Yoo S-Y, Kim JH, Jeon TY, Eo H. Clinical and imaging features of focal nodular hyperplasia in children. AJR Am J Roentgenol 2014; 202: 960–5. doi: 10.2214/AJR.13.11856 [DOI] [PubMed] [Google Scholar]
- 22.Cox VL, Bhosale P, Varadhachary GR, Wagner-Bartak N, Glitza IC, Gold KA, et al. Cancer genomics and important oncologic mutations: a contemporary guide for body imagers. Radiology 2017; 283: 314–40. doi: 10.1148/radiol.2017152224 [DOI] [PubMed] [Google Scholar]
- 23.Kitao A, Matsui O, Yoneda N, Kozaka K, Kobayashi S, Sanada J, et al. Hepatocellular carcinoma with β-catenin mutation: imaging and pathologic characteristics. Radiology 2015; 275: 708–17. doi: 10.1148/radiol.14141315 [DOI] [PubMed] [Google Scholar]
- 24.Fujita N, Nishie A, Asayama Y, Ishigami K, Ushijima Y, Kakihara D, et al. Hyperintense liver masses at hepatobiliary phase gadoxetic acid-enhanced MRI: imaging appearances and clinical importance. Radiographics 2020; 40: 72–94. Jan-Feb. doi: 10.1148/rg.2020190037 [DOI] [PubMed] [Google Scholar]
- 25.An HS, Park HS, Kim YJ, Jung SI, Jeon HJ. Focal nodular hyperplasia: characterisation at gadoxetic acid-enhanced MRI and diffusion-weighted MRI. Br J Radiol 2013; 86: 20130299. doi: 10.1259/bjr.20130299 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Demarco MP, Shen P, Bradley RF, Levine EA. Intraperitoneal hemorrhage in a patient with hepatic focal nodular hyperplasia. Am Surg 2006; 72: 555–9. [PubMed] [Google Scholar]
- 27.Gupta RT, Iseman CM, Leyendecker JR, Shyknevsky I, Merkle EM, Taouli B. Diagnosis of focal nodular hyperplasia with MRI: multicenter retrospective study comparing gadobenate dimeglumine to gadoxetate disodium. AJR Am J Roentgenol 2012; 199: 35–43. doi: 10.2214/AJR.11.7757 [DOI] [PubMed] [Google Scholar]
- 28.Katabathina VS, Menias CO, Shanbhogue AKP, Jagirdar J, Paspulati RM, Prasad SR. Genetics and imaging of hepatocellular adenomas: 2011 update. Radiographics 2011; 31: 1529–43. doi: 10.1148/rg.316115527 [DOI] [PubMed] [Google Scholar]
- 29.Garcia-Buitrago MT. Beta-Catenin staining of hepatocellular adenomas. Gastroenterol Hepatol 2017; 13: 740–3. [PMC free article] [PubMed] [Google Scholar]
- 30.Hui CL, Lee ZJ. Hepatic disorders associated with exogenous sex steroids: MR imaging findings. Abdom Radiol 2019; 44: 2436–47. doi: 10.1007/s00261-019-01941-4 [DOI] [PubMed] [Google Scholar]
- 31.Dhingra S, Fiel MI. Update on the new classification of hepatic adenomas: clinical, molecular, and pathologic characteristics. Arch Pathol Lab Med 2014; 138: 1090–7. doi: 10.5858/arpa.2013-0183-RA [DOI] [PubMed] [Google Scholar]
- 32.McInnes MDF, Hibbert RM, Inácio JR, Schieda N. Focal nodular hyperplasia and hepatocellular adenoma: accuracy of gadoxetic acid-enhanced Mr Imaging-A systematic review. Radiology 2015; 277: 927. doi: 10.1148/radiol.2015154033 [DOI] [PubMed] [Google Scholar]
- 33.van Aalten SM, Thomeer MGJ, Terkivatan T, Dwarkasing RS, Verheij J, de Man RA, et al. Hepatocellular adenomas: correlation of Mr imaging findings with pathologic subtype classification. Radiology 2011; 261: 172–81. doi: 10.1148/radiol.11110023 [DOI] [PubMed] [Google Scholar]
- 34.Grazioli L, Morana G, Kirchin MA, Schneider G. Accurate differentiation of focal nodular hyperplasia from hepatic adenoma at gadobenate dimeglumine-enhanced MR imaging: prospective study. Radiology 2005; 236: 166–77. doi: 10.1148/radiol.2361040338 [DOI] [PubMed] [Google Scholar]
- 35.Purysko AS, Remer EM, Coppa CP, Obuchowski NA, Schneider E, Veniero JC. Characteristics and distinguishing features of hepatocellular adenoma and focal nodular hyperplasia on gadoxetate disodium-enhanced MRI. AJR Am J Roentgenol 2012; 198: 115–23. doi: 10.2214/AJR.11.6836 [DOI] [PubMed] [Google Scholar]
- 36.Kim DH, Choi SH, Kim SY, Kim M-J, Lee SS, Byun JH, et al. Gadoxetic acid-enhanced MRI of hepatocellular carcinoma: value of washout in transitional and hepatobiliary phases. Radiology 2019; 292: 270. doi: 10.1148/radiol.2019194011 [DOI] [PubMed] [Google Scholar]
- 37.Zech CJ, Ba-Ssalamah A, Berg T, Chandarana H, Chau G-Y, Grazioli L, et al. Consensus report from the 8th International forum for liver magnetic resonance imaging. Eur Radiol 2020; 30: 370–82. doi: 10.1007/s00330-019-06369-4 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Chavhan GB, Shelmerdine S, Jhaveri K, Babyn PS. Liver MR imaging in children: current concepts and technique. Radiographics 2016; 36: 1517–32. Sep-Oct. doi: 10.1148/rg.2016160017 [DOI] [PubMed] [Google Scholar]
- 39.Kitao A, Matsui O, Yoneda N, Kozaka K, Kobayashi S, Sanada J, et al. Hepatocellular carcinoma with β-catenin mutation: imaging and pathologic characteristics. Radiology 2015; 275: 708–17. doi: 10.1148/radiol.14141315 [DOI] [PubMed] [Google Scholar]
- 40.Kitao A, Matsui O, Yoneda N, Kozaka K, Kobayashi S, Koda W, et al. Hypervascular hepatocellular carcinoma: correlation between biologic features and signal intensity on gadoxetic acid-enhanced Mr images. Radiology 2012; 265: 780–9. doi: 10.1148/radiol.12120226 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.Suh YJ, Kim M-J, Choi J-Y, Park YN, Park M-S, Kim KW. Differentiation of hepatic hyperintense lesions seen on gadoxetic acid-enhanced hepatobiliary phase MRI. AJR Am J Roentgenol 2011; 197: W44–52. doi: 10.2214/AJR.10.5845 [DOI] [PubMed] [Google Scholar]
- 42.Tirumani SH, Kim KW, Nishino M, Howard SA, Krajewski KM, Jagannathan JP, et al. Update on the role of imaging in management of metastatic colorectal cancer. Radiographics 2014; 34: 1908–28. Nov-Dec. doi: 10.1148/rg.347130090 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 43.Park YS, Lee CH, Kim JW, Shin S, Park CM. Differentiation of hepatocellular carcinoma from its various mimickers in liver magnetic resonance imaging: what are the tips when using hepatocyte-specific agents? World J Gastroenterol 2016; 22: 284–99. doi: 10.3748/wjg.v22.i1.284 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44.Fujiwara H, Sekine S, Onaya H, Shimada K, Mikata R, Arai Y, et al. Ring-Like enhancement of focal nodular hyperplasia with hepatobiliary-phase Gd-EOB-DTPA-enhanced magnetic resonance imaging: radiological-pathological correlation. Jpn J Radiol 2011; 29: 739–43. doi: 10.1007/s11604-011-0624-4 [DOI] [PubMed] [Google Scholar]
- 45.Kozaka K, Kobayashi S, Yoneda N, Kitao A, Yoshida K, Inoue D, et al. Doughnut-like hyperintense nodules on hepatobiliary phase without arterial-phase hyperenhancement in cirrhotic liver: imaging and clinicopathological features. Eur Radiol 2019; 29: 6489–98. doi: 10.1007/s00330-019-06329-y [DOI] [PubMed] [Google Scholar]
- 46.Yoneda N, Matsui O, Kitao A, Komori T, Kozaka K, Ikeda H, et al. Peri-Tumoral hyperintensity on hepatobiliary phase of gadoxetic acid-enhanced MRI in hepatocellular carcinomas: correlation with peri-tumoral hyperplasia and its pathological features. Abdom Radiol 2018; 43: 2103–12. doi: 10.1007/s00261-017-1437-4 [DOI] [PubMed] [Google Scholar]












