FINAL DIAGNOSIS: FOCAL NODULAR HYPERPLASIA LESIONS OF LIVER
Focal nodular hyperplasia (FNH) was first described and defined by Edmondson in 1958. It is a benign, tumor-like lesion consisting of hyperplastic parenchymal nodules with abundant abnormal vessels typically arising from a central stellate-shaped scar. Most (86- 91%) occur in women of reproductive age but convincing cases have also been described in men and children (1-4). In a study of 18 men and 216 women with FNH, mean FNH size was significantly smaller in men (37.5 mm in men vs. 63.4 mm in women) (1). The striking female predominance and size difference suggests a hormonal dependence. In the majority of the cases, the FNH is discovered incidentally. Symptoms that prompt discovery include recent to chronic abdominal pain or discomfort, palpable mass, and constitutional symptoms such as weight loss, weakness, and fever (3,4). FNH lesions are usually solitary, but multiple lesions are not rare (15-24%) (3,4). The characteristic gross appearance of FNH is a circumscribed, but unencapsulated, bulging, tan, lobulated, nodular mass. A central stellate scar with radiating fibrous septa has been emphasized as a hallmark feature of FNH, but approximately half of cases lack this finding (4).
Microscopically, FNH consists of nodules of benign-appearing hepatocytes separated from each other by fibrous septa that contain abnormal arteries, and to a lesser extent, veins. No normal liver architecture can be recognized within the nodule. The fibrous septa consist of variably vascularized, dense connective tissue, which usually contains chronic inflammatory cells. True bile ducts are not usually present.
The major microscopic differential diagnosis for FNH includes hepatocellular adenoma and nodular regenerative hyperplasia. FNH also needs to be distinguished from biliary cirrhosis. Briefly, hepatocellular adenoma usually lacks fibrous septa and cholangioles at the interface zone typically seen in FNH. Nodular regenerative hyperplasia is usually a diffuse disease that does not present with a localized mass. It is characterized by multiple hyperplastic periportal hepatocyte nodules that are not separated from each other by fibrous septae and vary from 0.1 to 1 cm. Presentation as a localized mass is also helpful in distinguishing FNH from biliary cirrhosis, provided that disease in the surrounding or non-lesional liver tissue has been reasonably excluded. The absence of true bile ducts and the presence of abnormal vessels within the fibrous septae also help distinguish FNH from biliary cirrhosis.
Distinguishing between an early FNH and hepatocellular adenoma is especially difficult, and in some cases can be impossible, on routine light microscopy alone. Like hepatocellular adenomas, early (small) FNH lesions are fed by so-called "aberrant" blood vessels or blood vessels without other portal tracts structures or an accompanying fibrous sheath. However, early or small FNH lesions also lack the typical characteristic features of more mature FNH lesions, such as fibrous septa containing chronic inflammation and cholangioles at the interface zone. In our experience, deeper levels, copper and biliary cytokeratin stains can help make the distinction. Copper deposition and biliary cytokeratin-positive cholangiolar metaplasia of hepatocytes at the interface zone between the aberrant vessels/fibrous septae favor a diagnosis of early FNH. Molecular analysis showing loss of heterozygosity or mutations at several loci favors a diagnosis of hepatocellular adenoma over FNH (5-7).
Variants of FNH have been described by Nguyen et al., that can have transitional morphologic features between adenoma and FNH (4). Simultaneous occurrence of adenoma and FNH has also been described (8). Therefore, the diagnosis of FNH based on small biopsy specimens should be made with extreme caution using close clinicopathologic correlation. If the biopsy specimen contains only normal appearing hepatocytes, diagnosis of FNH may be impossible. If fibrous septa are contained, the distinction between cirrhosis is difficult.
The vascular pathogenesis of FNH is now widely accepted. FNH probably represents a local hyperplastic response of hepatocytes to abnormal hepatic circulation. In 1985, Wanles et al proposed that FNH is a hyperplastic response of the hepatic parenchyma to a preexisting arterial spider-like malformation based on their morphometric analysis (9). They hypothesized that greater blood flow to a region compared to the adjacent parenchyma may cause nodular transformation. FNH lesions accompanied by clear congenital vascular anomalies of the liver or other organs have been reported, supporting the congenital theory of origin (10-13).
FNH-like lesions, conversely, have also been described sporadically in acquired liver disorders such as liver cirrhosis, Budd-Chiari syndrome, and portosystemic shunt. According to Quaglia et al. 5 out of 146 explanted cirrhotic livers (3.4%) had FNH-like lesions that were histologically indistinguishable from FNH, with fibrous scars and septa which contained vascular and ductular structures (14). The hepatocellular nodules occurring in the livers from patients with Budd-Chiari syndrome share morphologic features with large regenerative nodules, FNH, and hepatocellular adenoma. Cazals-Hatem et al. and Ibarrola et al. reported FNH-like nodules in 9 out of 17 (53%) and 1 out of 6 (17%) explanted livers of patients with severe Budd-Chiari syndrome, respectively (15,16). In acquired liver disorders, alterations of the vascular architecture occurring in chronic liver disease (cirrhosis) or thrombotic obstruction of portal vein branches (Budd-Chiari syndrome) are supposed to cause local ischemia and consequent hyperplasia and nodule formation in areas with a richer blood flow.
Endogenous estrogens may be associated with enhanced growth of FNH, but the etiologic role of oral contraceptives is controversial. It has been occasionally reported that regression of FNH can occur after withdrawal of the oral contraceptive pills (17). However, a recent comparative study performed by Mathieu et al. demonstrated that the number and size of lesions were independent of the use of oral contraceptives of any type and for any duration (18). Weimann et al. suggested that pregnancy was also not associated with FNH changes or complications (19).
Very few follow-up studies with objective size measurements have been performed to date, and therefore, the natural history of FNH remains largely unknown. According to the long-term follow-up study of biopsy-proven 18 lesions by Leconte et al., 6 lesions were stable, 10 decreased in size, and 2 increased (20). Thus, spontaneous regression of FNH seems to be not rare. The risk of hemorrhage and rupture is rare, as is malignant transformation (3). Small FNHs with no symptoms may not require treatment (21, 22). However, if the patient becomes symptomatic, or if a definitive diagnosis of FNH cannot be established, resection should be considered. Hepatic artery ligation and transarterial embolization are other alternative therapies for FNH (23, 24).
In summary, we presented a case of FNH in a 48-year-old-woman with a history of long-term use of oral contraceptives. The lesion presented as a symptomatic large (15cm) mass occupying the right lobe of the liver. In addition to the main mass, smaller nodules of FNH were found in bilateral hepatic lobes. The lesions lacked central stellate scars, which was not typical for FNH.
Contributed by Eizaburo Sasatomi MD, PhD and Anthony J Demetris, MD