Hepatic adenoma is a rare, benign tumor of the liver. [1, 2, 3, 4] Two types of hepatic adenoma have been identified, including tumors of bile duct origin and tumors of liver cell origin. Hepatic adenomas of bile duct origin are usually smaller than 1 cm and are not of clinical interest; typically, they are found incidentally on postmortem examinations. Hepatic adenomas of liver origin are larger—on average, they measure 8-15 cm—and are often clinically significant.
See the images below of a patient with hepatic adenoma.
Classification & characteristics
The 3 major subtypes include (1) the inflammatory/telangiectatic subtype, (2) the steatotic (with HNF1-α gene mutation) subtype, and (3) the subtype with β-catenin activation. An additional unclassified/miscellaneous subgroup is also recognized. [5, 6]
The inflammatory subtype is characterized as follows:
Previously misclassified as telangiectatic focal nodular hyperplasia
Comprises 40-55% of hepatocellular adenomas
Increased risk of bleeding but small risk of malignant transformation
Due to mutations involving interleukin 6
Strong expression of serum amyloid-associated protein A2 (SAA-2) and C-reactive protein on immunohistochemistry
Inflammation and peliosis at histology
The steatotic subtype is characterized as follows:
Comprises 35-50% of hepatocellular adenomas
No risk of malignant transformation
Leads to the development of familial adenomatosis and MODY-3 (maturity-onset diabetes of the young)
Lacks of expression of liver fatty acid binding protein (LFABP) on immunohistochemistry
The subtype with β-catenin activation is characterized as follow:
Comprises 10-18% of hepatocellular adenomas
Affects men and women
Increased risk of malignancy
Associated with androgen therapy and glycogen storage disease
Strong diffuse overexpression of glutamine synthetase and nuclear β-catenin staining
Risk of malignant transformation to hepatocellular carcinoma is 4-8%. After stopping the ingestion of oral contraceptives or anabolic steroids, the tumor can regress in size but the risk of malignant transformation remains.
Because of the small risk of life-threatening hemorrhage and the risk of malignant transformation, surgical excision is recommended for hepatic adenomas. Successful percutaneous radiofrequency ablation has been reported to be successful in 3 patients.  Radiofrequency ablation is less invasive and is an option for patients who cannot undergo surgery or who refuse surgery.
A combination of multiphasic computed tomography (CT) scans and gadolinium-enhanced magnetic resonance imaging (MRI) is best to identify and characterize most hepatic lesions. Certain characteristics, such as arterial enhancement and the presence of fat and hemorrhage, suggest that the lesion represents hepatic adenoma. If an enhancing central scar is seen, the diagnosis of focal nodular hyperplasia (FNH) can be made. Nuclear medicine studies can also be helpful. Most hepatic adenomas do not demonstrate uptake on sulfur-colloid and gallium-67 (67 Ga) scans. [8, 9, 10, 11, 12, 13, 14]
Limitations of techniques
Although CT scanning, MRI, and nuclear medicine studies may help characterize lesions as adenomas, the findings are frequently nonspecific, and biopsy and/or resection may be necessary.
Usually, plain radiographs of the abdomen provide no findings to suggest the diagnosis of hepatic adenoma. The liver is usually normal in size. Rarely, coarse calcifications may be present in hepatic adenomas; calcifications may be seen in the right upper quadrant on radiographs, but this finding is nonspecific.
Hepatic adenomas are often discovered incidentally on CT scans that are performed for other reasons. Once identified, a multiphasic CT scan should be performed to better characterize most hepatic tumors. Protocols differ from institution to institution.
See the CT images below of a patient with hepatic adenocarcinoma.
Typically, helical CT scans are obtained, first of the nonenhanced liver. Then, images are obtained in the hepatic arterial phase using intravenous injection of approximately 120-150 mL of nonionic contrast at a rate of 3-5 mL/s with a 25- to 30-second delay. Images are then acquired in the portal venous phase after a scanning delay of 60-80 seconds.
On CT scans, the most consistent finding in hepatic adenomas is the enhancement pattern. Most lesions (90% according to Ichikawa et al  ) show homogeneous enhancement in the hepatic arterial phase. Unfortunately, this feature is not specific to hepatic adenomas, because hepatocellular carcinoma, hypervascular metastases, and focal nodular hyperplasia can demonstrate similar enhancement in the hepatic arterial phase.
Because hepatic adenomas are histologically composed of uniform hepatocytes, most are isoattenuating relative to the healthy liver tissue on nonenhanced scans in the portal venous phase. In a fatty liver, hepatic adenomas usually are hyperattenuating.
The finding of hemorrhage as an area of high attenuation can be seen in as many as 40% of patients. Fat deposition within adenomas is identified on CT scans in only approximately 7% of patients.
Typically, hepatic adenomas have well-defined borders and do not have lobulated contours. A low-attenuation pseudocapsule can be seen in as many as 25% of patients. Coarse calcifications are seen in only 5% of patients.
Magnetic Resonance Imaging
Hepatic adenomas tend to be hyperintense or isointense relative to the liver tissue on T1-weighted images (up to 93% in a series by Paulson et al  ). High signal intensity on T1-weighted images probably relates to the presence of fat or, less commonly, to hemorrhage within the lesion.
Chemical-shift imaging that shows loss of signal on out-of-phase images can confirm the presence of fat. Unfortunately, hepatocellular carcinoma (HCC) is known to contain fat in as many as 40% of lesions; therefore, the presence of fat does not help differentiate the lesions.
Other hepatic lesions can be hyperintense on T1-weighted images, such as melanoma metastases and cavities containing proteinaceous material. On T2-weighted images, hepatic adenomas are most often slightly hyperintense relative to liver tissue. This finding is not specific because many hepatic lesions, including HCC and metastases, are hyperintense on T2-weighted images.
Heterogeneity, defined as any difference of a signal within a lesion on T1-weighted or T2-weighted images, is seen in approximately one half of patients. Heterogeneity relates to the presence of either hemorrhage or necrosis. This finding is not specific as HCC and metastases can bleed and become necrotic. Although uncommon, focal nodular hyperplasia (FNH) also can be hemorrhagic.
A peripheral rim corresponding histologically to a pseudocapsule is seen in 17-31% of patients. Signal characteristics of the rim are variable. Most often, the peripheral rim, when seen, is of low signal intensity on T1-weighted images, is of variable intensity on T2-weighted images, and usually does not enhance.
After gadolinium administration, the pattern of enhancement is similar to that of CT scans. Most hepatic adenomas show intense enhancement in the arterial phase and are isointense relative to the liver tissue on delayed imaging.
Gadolinium-based contrast agents have been linked to the development of nephrogenic systemic fibrosis (NSF) or nephrogenic fibrosing dermopathy (NFD). NSF/NFD has occurred in patients with moderate to end-stage renal disease after being given a gadolinium-based contrast agent to enhance MRI or MRA scans. NSF/NFD is a debilitating and sometimes fatal disease. Characteristics include red or dark patches on the skin; burning, itching, swelling, hardening, and tightening of the skin; yellow spots on the whites of the eyes; joint stiffness with trouble moving or straightening the arms, hands, legs, or feet; pain deep in the hip bones or ribs; and muscle weakness.
Hepatic adenomas, unlike FNH, do not have a central scar. If a low signal intensity scar is seen on T1-weighted images and the scar enhances after gadolinium is administered, the diagnosis of FNH is strongly favored. A central scar has never been reported in a hepatic adenomas.
On routine MRI of the liver consisting of T1-weighted and T2-weighted images, chemical-shift imaging, and dynamic gadolinium-enhanced imaging, distinguishing among hepatic adenomas, HCC, and hypervascular metastases is usually not possible.
Studies have been performed to determine if MRI (superparamagnetic iron oxides) enhancement with the use of ferumoxides may help to better distinguish FNH from hepatic adenomas and HCC in indeterminate cases.  Ferumoxides are taken up by the reticuloendothelial cells in a healthy liver. Because FNH contains Kupffer cells, the ferumoxides are taken up by healthy liver tissue and by FNH, which results in marked reduction in the signal intensity of healthy liver tissue and FNH on T2-weighted images.
Usually, no other lesions show significant signal loss on T2-weighted images. Lesions such as HCC, hepatic adenomas, and metastases usually become conspicuous because they lack a significant number of Kupffer cells. Lesions such as FNH drop out almost as much as healthy liver tissue; however, some hepatic adenomas and some well-differentiated HCCs show some signal intensity loss, which may be explained by the presence of some Kupffer cells in the lesions.
Mangafodipir trisodium (formerly termed Mn-DPDP) is a hepatobiliary MRI contrast agent that is taken up by hepatocytes and excreted into bile. Because hepatic adenoma, FNH, and HCC all contain hepatocytes, they may demonstrate enhancement with this agent. Metastases and hemangiomas do not contain hepatocytes and do not enhance; therefore, this agent can help differentiate hepatic adenoma, which enhances, from metastases, which do not enhance.
Venkatesh et al assessed the potential for MR elastography (MRE) to characterize solid liver tumors.  After 44 tumors were identified on T2- and T1-weighted and gadolinium-enhanced T1-weighted images, a stiffness map (elastogram) was obtained, and the stiffness value of tumor-free hepatic parenchyma was calculated. The results appeared to be promising for MRE as a potential noninvasive technique in the evaluation of solid liver tumors: malignant liver tumors had significantly greater mean shear stiffness than benign tumors, fibrotic liver, and normal liver, and fibrotic livers had stiffness values that overlapped both the benign and the malignant tumors. The authors reported that a "cutoff value of 5 kPa accurately differentiated malignant tumors from benign tumors and normal liver parenchyma in this preliminary investigation." 
In another study, Chang and Thoeni evaluated the effect of T1 shortening on T2-weighted MRI sequences before and after the administration of gadolinium by comparing conspicuity of 118 pathologically proven or serially followed focal liver lesions in 84 patients  "On gadolinium-enhanced T2-weighted images, 21 (17.8%) of 118 of the lesions had improved conspicuity, 86 (72.9%) had no difference in conspicuity, and 11 (9.3%) appeared worse." However, there was no statistically significant difference between the unenhanced and enhanced images, although the authors noted a trend toward improved conspicuity with the gadolinium enhancement MRIs. With subgroup analysis the gadolinium-enhanced T2-weighted images, demonstrated significantly better visualization of solid hepatic lesions (28.3%) compared with nonsolid lesions (9.2%). 
Giovanoli et al evaluated the use of gadoxetic acid, a newly developed liver-specific MR contrast agent, in 3 patients with histologically proven liver adenomatosis.  The initial results were not promising: in all 3 patients, >100 adenomas showed no or limited uptake of gadoxetic acid in the delayed phase, thus not making it possible for the differentiation of adenomas from dysplastic or malignant lesions.
Degree of confidence
Generally, on routine MRI of the liver using T1-weighted, T2-weighted, chemical-shift, and dynamic gadolinium-enhanced imaging, certain hepatic masses can be diagnosed with confidence, whereas others cannot.
If a hepatic mass contains a low signal central scar on T1-weighted images that enhance after gadolinium administration, the diagnosis of focal nodular hyperplasia (FNH) is fairly certain.
However, overlap exists in the imaging and enhancement characteristics of hepatic adenomas, hepatocellular carcinoma (HCC), and hypervascular metastases such as melanoma. Clinical correlation in such cases is most helpful. A history of cirrhosis and high alpha-fetoprotein levels favor an HCC diagnosis. A history of melanoma or other primary tumors favors the diagnosis of metastases. In otherwise healthy young women using oral contraceptives, the diagnosis of hepatic adenoma is favored. Patients with glycogen storage disease, hemochromatosis, or acromegaly, as well as males on anabolic steroids, are also more prone to developing hepatic adenomas.
Although most hepatic adenomas are hyperintense relative to normal liver on T1-weighted images, this is not a specific finding. Other hepatic masses, such as hepatocellular carcinoma, melanoma, metastases, and protein material in hepatic abscess cavities, can be hyperintense on T1-weighted images as well.
On ultrasound, hepatic adenomas demonstrate variable echogenicity (see the first image below). The lesions may be hypoechoic, isoechoic, or hyperechoic relative to liver parenchyma. Usually, differentiating hepatic adenomas from other liver lesions such as focal nodular hyperplasia or hepatocellular carcinoma is not possible based on either gray scale or Doppler ultrasonographic characteristics. 
In April 2016, the FDA approved sulfur hexafluoride (Lumason) for ultrasonography of the liver in adults and children to enhance the characterization of focal liver lesions. Contrast-enhanced ultrasound (CEUS) greatly improved diagnosis as compared to ultrasound without contrast. In this study, CEUS allowed a correct diagnosis in more than 80% of focal liver lesions and led to a change in the diagnostic workup in 131/157 patients (83.4%) and in the therapeutic workup in 93/157 patients (59.2%). 
Cherqui et al described increased intralesional venous structures with a paucity of intra-arterial structures in hepatic adenomas  ; however, Rumack et al failed to replicate this finding, and it is not a reliable differentiating feature.  The primary role of ultrasound is to screen patients with hepatic masses that are discovered incidentally or who have a clinical history of abnormal liver function test results. Further imaging is then indicated using MRI, CT scanning, and/or nuclear medicine.
Morin et al, however, report that ultrasound can be used with specific contrast media and specialized imaging techniques to fully characterize the enhancement pattern of hepatic lesions, which, the authors indicate, are similar to that achieved with contrast-enhanced, multiphasic CT scanning and MRI. 
A combination of radiotracers may help make the diagnosis of hepatic adenomas in equivocal cases (see the third image below). 
On gallium-67 (67 Ga) scans, hepatic adenomas demonstrate decreased uptake compared with healthy liver tissue, which can be explained by the benign nature of the cells. In contrast, hepatocellular carcinoma (HCC) often demonstrates equivocal or greater67 Ga uptake than liver, with studies reporting that 90-95% of HCCs demonstrate uptake or equivocal uptake of67 Ga.
Because hepatic adenomas usually have few or absent Kupffer cells, the lesions show focal defects on sulfur-colloid liver-spleen scans. However, an occasional hepatic adenoma contains enough Kupffer cells to demonstrate normal uptake of sulfur colloid. HCC almost always appears as defects on sulfur-colloid scintigraphy because HCC lacks Kupffer cells. In contrast, focal nodular hyperplasia (FNH) contains Kupffer cells and usually demonstrates uptake of sulfur colloid. In summary, sulfur-colloid uptake strongly favors a diagnosis of FNH. Lack of sulfur-colloid uptake is not specific and can be attributed to many other hepatic lesions, including hepatic adenomas, HCC, and metastases.
When hepatobiliary agents are used, hepatic adenomas usually demonstrate early uptake with subsequent retention of the radiotracer because hepatic adenomas do not contain bile ducts; thus, the radiotracer is not excreted by the lesion, which remains "hot" on delayed images. This is in contrast to HCC, which shows focal defects on early scans. Avid uptake becomes detectable only after 2-5 hours.
The use of positron emission tomography (PET) scanning with fluorine-18-fluorodeoxyglucose (18 FDG) has been shown to be useful in the evaluation of many tumors. Malignant tumors usually show uptake of18 FDG, whereas benign tumors do not. Occasionally, benign lesions such as sarcoid lesions, inflammatory processes, and abscesses can show uptake. A case has been reported of18 FDG uptake in a hepatic adenoma. 
When hepatic adenoma is radiologically indistinguishable from HCC and FNH, a combination of radionuclide imaging, including sulfur-colloid,67 Ga, and technetium-99m (99m Tc) pyridoxyl-5-methyltryptophan (PMT) uptake, may help establish the correct diagnosis. Most hepatic adenomas demonstrate decreased67 Ga uptake, decreased sulfur-colloid uptake, and early and retained uptake of hepatobiliary agents.
Degree of confidence
Most hepatic adenomas demonstrate decreased67 Ga uptake, decreased colloid uptake, early and retained uptake of hepatobiliary agents, and no uptake on PET scanning; therefore, the diagnosis of hepatic adenoma can often be confidently made with the use of nuclear medicine studies.
Cases have been reported of hot hepatic adenomas on PET18 FDG scans. In addition, reports exist of hepatic adenomas with enough Kupffer cells to demonstrate uptake on sulfur colloid scans.
In the diagnostic workup of hepatic adenomas, angiography does not have a significant role. This modality can be helpful for technical reasons when considering resection. On angiography, hepatic adenomas typically appear as hypervascular masses, with the vascular supply arising peripherally. However, hepatic adenomas may be hypovascular (as many as 50%) or have areas of hypovascularity within the mass that correspond to hemorrhage and necrosis.
In contrast, focal nodular hyperplasia (FNH) is typically hypervascular with dense capillary blushing. In large lesions, a dilated branch of the hepatic artery can enter the center of the mass and then divide into small branches that radiate in a manner similar to the spokes on a wheel (spoke-wheel appearance). If the spokelike appearance is noted, FNH is the likely diagnosis. Hepatocellular carcinoma (HCC) demonstrates hypervascularity, irregular tumor vessels, and arteriovenous shunting. In patients with HCC, a tumor thrombus in the portal or hepatic veins may also be seen. Most liver metastases are hypervascular with a capillary stain.
Degree of confidence
Angiography is usually not performed for the detection and differentiation of hepatic masses. Angiography can be performed preoperatively to better define the vascular anatomy before resection, although the information can be obtained noninvasively with CT scanning or MR angiography.