Hepatic adenomas (HAs) are benign lesions of the liver. Potential complications of HA include hemorrhage, rupture, and malignant transformation. HAs and hepatic carcinomas have been associated with exogenous androgen use in the past. To our knowledge, there is only 1 report in the literature of HA being associated with endogenous androgens. That was in a 29-year-old man with HA, focal nodular hyperplasia, and elevated levels of both androgens and estrogens (1). In polycystic ovarian syndrome women have overproduction of ovarian and sometimes adrenal androgens. We report here on an adolescent girl with endogenously elevated androgen levels in whom an HA developed.
The patient was an obese Hispanic girl, 13 years and 10 months old, who was brought for evaluation of polycystic ovarian syndrome and type 2 diabetes mellitus. History revealed pubic hair development at 8 years of age, breast development at 11 years, excess of facial hair, and deepening of her voice. Review of her systems was positive for easy bruising, fatigue after running, rare temporal/frontal headaches, and intermittent nocturia but no polyuria. She described having vague abdominal pain, not associated with nausea, vomiting, diarrhea, or weight loss. She did not describe taking any medications at the time of presentation. She was in the 7th grade in school, receiving special assistance in reading and other subjects. Her medical history was significant for sleep apnea, requiring tonsillectomy and adenoidectomy. Her mother was also obese (body mass index [BMI] 45 kg/m2) and had a history of irregular menses, but no fertility problems. There was a family history of type 2 diabetes mellitus in a maternal uncle.
On assessment, the girl's weight was 89.6 kg (>95%), her height was 151.8 cm (10%), and her BMI was 38.9 kg/m2 (>95%). Her blood pressure was normal for her age. She had hyperpigmented velvety plaques localized to her inner elbows, neck, and axillae. She was hirsute, with marked truncal, back, and facial hair. Temporal balding was evident. The results of abdominal examination were noncontributory, with no apparent hepatosplenomegaly. She had Tanner stage V breasts and pubic hair and normal external female genitalia, without clitoromegaly. The results of physical examination were negative for acne, papillary edema, thyromegaly, obvious abdominal masses, skin ecchymoses, or congenital anomalies.
Laboratory test results included an aspartate aminotransferase of 23 IU/L (normal 15–45), alanine aminotransferase of 26 IU/L (normal 7–35), total bilirubin of 10.3 μmol/L (normal 5–21), cholesterol of 209 mg/dL (normal <170), triglycerides of 216 mg/dL (normal <120), low-density lipoprotein of 132 mg/dL (normal <110), high-density lipoprotein of 34 mg/dL (desirable 45), follicle-stimulating hormone of 11 mIU/mL (normal 3.7–12.9), luteinizing hormone of 12 mIU/mL (normal 0.9–14), free thyroxine of 1.4 ng/dL (normal 0.7–1.7), thyroid-stimulating hormone of 2.15 mIU/mL (normal 0.6–6.3), dehydroepiandrosterone sulfate of 66 μg/dL (normal 138 ± 51), androstenedione 217 ng/dL (normal 80–240), fluorescence in situ hybridization for Prader-Willi normal, α-fetoprotein of 1.6 (normal <15) μg/L, carcinoembryonic antigen of <0.4 μg/L (normal 0–3.4). Total and free testosterone and sex hormone binding globulin levels before and after surgical and medical treatment are shown in Table 1. The patient underwent an oral glucose tolerance test, the result of which was significant for evidence of insulin resistance (Table 2), and a dexamethasone suppression test, which showed suppression of cortisol but not testosterone, suggesting an ovarian source for at least part of the hyperandrogenemia (Table 3).
Ultrasound showed the right ovary to be enlarged with multiple follicles (ovarian volume of 14.7 mL). No ovarian or adrenal tumor was seen on computed tomography; however, there was an ovoid mass lesion at the right liver lobe measuring 4.3 × 4.4 cm. Magnetic resonance imaging demonstrated uniform high signal intensity on T1-weighted imaging, suggesting that the lesion contained fat. With administration of gadolinium contrast material, uniform enhancement of the lesion was seen on T1 fat-saturated imaging sequence, excluding focal steatosis and making the diagnosis of HA more likely (Fig. 1). The liver mass was surgically resected without complications. Histopathological examination of the lesion was diagnostic of HA (Fig. 2). Immunochemical staining of the HA for androgen receptors and estrogen receptors was negative. Analysis of the girl's TCF1 gene, a gene that encodes hepatocyte nuclear factor 1-α, identified a DNA sequence variant (change of the amino acid threonine to alanine at nucleotide position 586) of unknown significance.
The patient's androgens remained elevated after resection of the HA (Table 1). Estrogen and progesterone treatments were not recommended, given their possible associations with HA. Therefore, treatment was started with metformin. Her current dose is 1000 mg twice per day. With good compliance and toleration of metformin, spotting has begun to occur monthly. She reports increased exercise and healthier eating patterns. Hair growth has not worsened or improved. Her BMI has slightly increased to 39.8 kg/m2. Her laboratory findings now show marked improvements, but not normalization of androgens (Table 1). We are following endometrial stripe by ultrasonography, and if it increases we will consider rectal progesterone to induce menses. Repeat imaging has shown no further evidence of HA or ovarian tumors.
HA is a rare benign hepatic tumor in the pediatric patient population. It is more commonly seen in adolescent girls who take oral contraceptives but has been described in younger children (2) and even in utero (3). The pathogenesis of HA remains uncertain. The association between HA and oral contraceptives was first proposed in the 1970s (4,5). HAs are often silent, and the control data were not thoroughly studied, making the data difficult to interpret. Regression of HA after withdrawal of oral contraceptives, however, has been observed (6). Exogenous androgens, which have been most commonly used in aplastic anemias as well as in many other conditions, such as endocrine abnormalities, hereditary angioedema, immune thrombocytopenia, and body building, have been associated with HA. This evidence again supports a role for hormonal regulation in HA development.
To our knowledge, this is the first case report of a patient with polycystic ovarian syndrome and HA. One can ask whether the patient's elevated androgen levels were an effect of the tumor or a potential mechanism in its pathogenesis. Adrenal rest tumors (7) and hepatic carcinomas can rarely secrete androgens (8), but we are unaware of any report to date of HAs secreting androgens. In this case there was no decrease in the level of androgens after resection of the HA. This makes HA as the source of the hyperandrogenemia unlikely.
The role of androgens and their receptors (AR) in the pathogenesis of HA is unclear. In 1988 Ostrowski et al (9) showed a positive correlation between liver tumor development and AR expression. In 1997 Eagon et al (10) described a higher proportion of nuclear AR (biologically active form) in the HA compared with normal liver. A year later, Cohen et al (11) studied the immunohistochemical expression of AR in HA and adjacent liver in 18 patients. They also demonstrated nuclear AR in 9 of 27 HA and 2 of 11 adjacent normal livers of similar immunostain intensity but in a larger percentage of nuclei in HA than in the adjacent normal liver. These findings may suggest a possible role for AR and androgens in the pathogenesis of HA.
Histologically, HA is composed of cells larger than normal hepatocytes arranged in cords, separated by prominent sinusoids, and is characterized by absence of portal triads. There is a report in the literature of androgen-dependent adenomas (12) with distinct histological features such as nuclear atypia (occasionally large nucleoles, sometimes binucleated nuclei), well-defined acinar formation, and plugs of bile. AR staining was not done in that study. The histological appearance of the HA in this case report resembles the description by Touraine et al (12) for androgen-dependent adenomas (Fig. 2), yet AR staining in our patient was negative.
HAs have been related to exogenous androgen previously. We discuss an adolescent girl with excess endogenous androgen production and an HA with histological features consistent with hyperandrogenemia. Treatment of her underlying insulin resistance will benefit her glucose homeostasis and also, we hope, continue to improve her hyperandrogenemia, thereby possibly decreasing the risk for future HAs.
1. Grange JD, Guechot J, Legendre C, et al
. Liver adenoma and focal nodular hyperplasia in a man with high endogenous sex steroids. Gastroenterology 1987; 93:1409–1413.
2. Wheeler DA, Edmondson HA, Reynolds TB. Spontaneous liver cell adenoma in children. Am J Clin Pathol 1986; 85:6–12.
3. Resnick MB, Kozakewich HP, Perez-Atayde AR. Hepatic adenoma in the pediatric age group: clinicopathological observations and assessment of cell proliferative activity. Am J Surg Pathol 1995; 19:1181–1190.
4. Rooks JB, Ory HW, Ishak KG, et al
. Epidemiology of hepatocellular adenoma: the role of oral contraceptive use. JAMA 1979; 242:644–648.
5. Edmondson HA, Henderson B, Benton B. Liver-cell adenomas associated with use of oral contraceptives. N Engl J Med 1976; 294:470–472.
6. Edmondson HA, Reynolds TB, Henderson B, et al
. Regression of liver cell adenomas associated with oral contraceptives. Ann Intern Med 1977; 86:180–182.
7. Wallace EZ, Leonidas JR, Stanek AE, et al
. Endocrine studies in a patient with functioning adrenal rest tumor of the liver. Am J Med 1981; 70:1122–1125.
8. Galifer RB, Sultan C, Margueritte G, et al
. Testosterone-producing hepatoblastoma in a 3-year-old boy with precocious puberty. J Pediatr Surg 1985; 20:713–714.
9. Ostrowski JL, Ingleton PM, Underwood JC, et al
. Increased hepatic androgen receptor expression in female rats during diethylnitrosamine liver carcinogenesis: a possible correlation with liver tumor development. Gastroenterology 1988; 94:1193–1200.
10. Eagon PK, Elm MS, Stafford EA, et al
. Androgen receptor in human liver: characterization and quantitation in normal and diseased liver. Hepatology 1994; 19:92–100.
11. Cohen C, Lawson D, DeRose PB. Sex and androgenic steroid receptor expression in hepatic adenomas. Hum Pathol 1998; 29:1428–1432.
© 2007 Lippincott Williams & Wilkins, Inc.
12. Touraine RL, Bertrand Y, Foray P, et al
. Hepatic tumours during androgen therapy in Fanconi anaemia. Eur J Pediatr 1993; 152:691–693.