Focal nodular hyperplasia is the second most common benign lesion of the liver after hemangioma. It represents up to 86% of nonhemangiomatous benign lesions and occurs 3 to 10 times more frequently than hepatic adenomas.¹, ², ³ Although focal nodular hyperplasia is seen in both sexes throughout the age spectrum, these lesions are 10 times more common in women, with 75% occurring in young women between the ages of 20 and 50 years (median 41 years).⁴ Less commonly, they are seen in children and adolescents, where they comprise up to 2% of all hepatic lesions in children.⁵

Focal nodular hyperplasia is usually solitary, although multiple nodules have been reported in 20% to 30% of cases.⁶ Most cases are asymptomatic and discovered incidentally at the time of surgery, on unrelated imaging studies, or at autopsy. Patients who do report symptoms describe abdominal discomfort, pain, or a palpable mass.⁶,⁷ Focal nodular hyperplasia can be evaluated by computed tomography or magnetic resonance imaging because imaging studies show highly characteristic features and are diagnostic in 90% of cases.⁸,⁹ In these cases, a biopsy is usually unnecessary for diagnosis. Unlike hepatic adenomas, focal nodular hyperplasias do not carry any risk for malignant transformation and they only rarely present with acute infarction, necrosis, or hemorrhage.¹⁰,¹¹

Focal nodular hyperplasias are the result of parenchymal hyperplasia in response to increased blood flow and arterial shunting, usually as a result of vascular malformations.¹² In fact, 23% of patients with focal nodular hyperplasia have associated hemangiomas.¹³ The relationship
between these lesions and vascular shunting is further supported by the presence of focal nodular hyperplasia lesions in patients with hereditary hemorrhagic telangiectasia (Osler-Weber-Rendu disease)¹⁴ as well as hemihypertrophy and vascular malformations (Klippel-Trénaunay-Weber syndrome).¹⁵

A tenuous relationship between oral contraceptives and focal nodular hyperplasia has been reported,¹⁶ but the data arguing against a causal relationship are strong. For example, focal nodular hyperplasias occur in men and children who do not use oral contraceptives.¹⁷ Furthermore, focal nodular hyperplasia was first described in the early 1900s, long before the advent of oral contraceptives, and its incidence has remained steady after the introduction of oral contraceptives in the 1960s.¹⁸,¹⁹ This observation is in sharp contrast to the dramatic rise in the incidence of hepatic adenomas in the same time period.¹⁹ Although likely not causal, estrogens may play a role in the growth of focal nodular hyperplasia because there have been reports of focal nodular hyperplasia enlarging in the setting of oral contraceptives or pregnancy.¹⁶

In the pediatric population, focal nodular hyperplasias are often preceded by chemotherapy for malignancies of other organs and typically occur many years after the treatment.

By definition, the background liver is noncirrhotic. The lesions are usually solitary (80% to 95%), well demarcated, without a capsule, and can occasionally be pedunculated.⁶ They are found more commonly in the right lobe and are usually <5 cm in diameter, with only 3% larger than 10 cm; rare reports describe focal nodular hyperplasias as large as 19 cm.⁶,²⁰ On cut section, the lesions are lobulated and have a firm rubbery texture. They may be lighter in color than the surrounding liver. A characteristic finding is the presence of a central stellate scar, or multiple scars, with radiating fibrous bands separating lobules of parenchyma (Fig. 27.1), imparting the “spoke and wheel” architecture commonly seen on imaging studies. Smaller lesions, particularly those that are <3 cm, may lack this characteristic scar, which is seen in 62% of lesions overall.⁶,²¹

Histologically, focal nodular hyperplasias are well demarcated but unencapsulated lesions composed of benign hepatocytes with characteristic fibrous bands radiating from a central scar-like area (Figs. 27.2 and 27.3). These fibrous bands can encompass cords or nodules of hepatic parenchyma, resulting in the appearance of focal cirrhosis, a potential pitfall in needle biopsy specimens. However, the scar-like tissue within focal nodular hyperplasia nodules does not contain true portal tracts, but instead contains abnormal large feeder arteries that may have medial hypertrophy and intimal fibrosis (Fig. 27.4), at times even occluding the lumen. The native bile ducts are replaced by a prominent bile ductular proliferation traveling along the edge of the fibrous septae (Fig. 27.5). These characteristics may be less developed in smaller lesions, instead appearing more like bridging fibrosis than focal cirrhosis.

The hepatocytes within the lesion are essentially indistinguishable from those in the background noncirrhotic parenchyma and show no cytological or architectural atypia. However, the poor bile drainage in these lesions can result in cholate stasis, copper deposition, feathery degeneration of hepatocytes, and
Mallory hyaline accumulation in areas adjacent to the fibrous bands. Hepatocytes within the focal nodular hyperplasia can also show steatosis, and lymphocytic inflammation may be found in the fibrous bands.

Because most diagnoses of focal nodular hyperplasias can be made on radiologic examination, excision or biopsy occurs only when radiologic features are atypical and the diagnosis is unclear, such as in smaller lesions that do not exhibit classic stellate scars. The lack of classic features can also prove diagnostically challenging for pathologists, and special stains or immunohistochemistry can aid in the diagnosis. Glutamine synthetase, a surrogate marker for β-catenin activation, is perhaps the most useful immunostain in the diagnosis of focal nodular hyperplasias. One study demonstrated an increase in confident diagnosis to 90% of cases with the addition of this immunostain, as compared with only 50% of cases using morphology alone.²² Glutamine synthetase results in a “map-like” pattern of staining in 90% of focal nodular hyperplasias (Fig. 27.6), with strong cytoplasmic reactivity in central areas, contrasted by broad geographic areas of nonreactive hepatocytes near the fibrous bands.²² By comparison, the normal liver shows glutamine synthetase staining limited to the few hepatocytes directly adjacent to the central vein. This map-like staining quality can be useful in differentiating focal nodular hyperplasias from hepatic adenoma (see section Differential Diagnosis). Other immunostains commonly used to subtype hepatic adenomas include liver fatty acid-binding protein (L-FABP), β-catenin, serum amyloid A, and C-reactive protein. In focal nodular hyperplasias, immunostains for L-FABP show normal retained expression, β-catenin is nonreactive for nuclear expression (but shows normal cytoplasmic or membranous reactivity), and serum amyloid A and C-reactive protein are negative for strong and diffuse staining, though patchy staining can be seen in some cases of focal nodular hyperplasias.²³

Other stains commonly used in the evaluation of liver biopsies include reticulin, which will highlight a normal reticulin framework with cell plates 1 to 2 hepatocytes in width, and copper stains such as Rhodamine, which can highlight copper accumulation in hepatocytes adjacent to the fibrous bands, a result of impaired bile drainage. Copper is not specific for focal nodular hyperplasia, as it is also positive in some inflammatory adenomas and hepatocellular carcinomas.

In the setting of a noncirrhotic patient, the diagnosis of focal nodular hyperplasia can be suggested if hepatocytes are accompanied by bile duct epithelium and, sometimes, by fibrous tissue. The presence of bile duct epithelium essentially rules out adenomas and hepatocellular carcinomas.

Hepatic adenomas are typically solitary lesions that arise more commonly in the right lobe of the liver and can range in size from 1 to 30 cm. Patients may be asymptomatic, have mild abdominal discomfort, or present acutely with tumor rupture and hemorrhage.³² Larger lesions are more likely to present with symptoms such as abdominal pain,³³ and lesions >5 cm have higher risk of bleeding—a sequela which is accompanied by sudden severe pain and hypotension. On physical examination, an abdominal mass can be identified in up to 30% of patients, whereas hepatomegaly is present in approximately 25%. Elevations in liver function tests are rare, but elevations of alkaline phosphatase and jaundice have been described, presumably reflecting compression of intrahepatic bile ducts by an enlarging tumor. γ glutamyl transpeptidase (GGT) may be elevated, particularly in patients with intratumoral bleeding or multiple adenomas. α fetoprotein is normal except in some cases of malignant transformation. Whereas up to half of patients can have multiple adenomas,³² about 10% to 15% of

patients have 10 or more lesions—a condition called hepatic adenomatosis.³⁴

Hepatic adenomatosis is defined formally by the presence of 10 or more adenomas in the liver. Originally, hepatic adenomatosis was thought to be a genetically distinct entity from cases with a single or few hepatic adenomas, but this is no longer generally thought to be true. Early definitions excluded cases arising in glycogen storage diseases, or in the setting of steroid use, or that improved upon steroid withdrawal. Hepatic adenomatosis in some studies has a more equal distribution in men and women, and cases typically show elevation of alkaline phosphatase and GGT levels.³⁵,³⁶ Currently, hepatic adenomatosis is no longer considered to be distinct entity.³⁷

Hepatic adenomatosis can be associated with congenital or acquired abnormalities of hepatic vasculature (e.g., congenital absence of portal vein, portal venous thrombosis with cavernous transformation, or intrahepatic portosystemic shunts).³⁸ Lesions in adenomatosis appear to have a higher rate of hemorrhage when compared with traditional hepatic adenomas (62% vs. 26% in one study),³⁶ in particular those >4 cm and subcapsular in location. The adenomas found in hepatic adenomatosis may be of any subtype, including inflammatory, hepatocyte nuclear factor 1 α (HNF1α) inactivated, β-catenin activated, or unclassified.³⁹ In some cases of adenomatosis, subclassification can show a mixture of separate inflammatory adenomas, β-catenin-activated adenomas, and inflammatory adenomas with secondary β-catenin activation. Malignant transformation has been well documented.³⁸,³⁹

Hepatic adenomas are uncommon tumors, with 85% of cases occurring in young to middle-aged women with a history of oral contraceptive use.⁴⁰, ⁴¹, ⁴² The annual incidence is 30 to 40 per million in long-term users (>2 years), compared with 1 per million in women who have never used oral contraceptives.⁴²,⁴³ Hepatic adenomas were rarely reported before the advent of oral contraceptives in the 1960s, and subsequent studies have shown that dose and duration of hormonal therapy directly correlates with the size, number, and frequency of bleeding in these tumors.⁴¹,⁴²,⁴⁴, ⁴⁵, ⁴⁶, ⁴⁷, ⁴⁸ Regression of adenomas has been observed after discontinuation of oral contraceptive, with recurrence during readministration or pregnancy.⁴⁹,⁵⁰ The current formulations of low-dose estrogen oral contraceptive have led to a reduction in tumor incidence, but tumor development is not limited to oral contraceptive use because any population receiving estrogen for medical purposes can develop adenomas.