Abbreviations

AFP

α-fetoprotein

CEA

carcinoembryonic antigen

CRP

C-reactive protein

CT

computed tomography

FNH

focal nodular hyperplasia

Gd-BOPTA

gadobenate dimeglumine

GGT

γ-glutamyl transpeptidase

HCA

hepatocellular adenoma

HCC

hepatocellular carcinoma

HNF1α

hepatocyte nuclear factor 1α

HUMARA

human androgen receptor assay

IL-6

interleukin 6

Mn DPDP

manganese dipyridoxal diphosphate

MRI

magnetic resonance imaging

NRH

nodular regenerative hyperplasia

RES

reticuloendothelial system

SAA

serum amyloid A

SPECT

single-photon emission CT

Introduction

Benign liver tumors are a heterogeneous group of nodular lesions originating from different cell lines. These tumors include hemangiomas, which are the benign nodes found most commonly in the liver, and hepatocellular neoplasms, which are clinically more relevant lesions ( Table 49-1 ). For hepatocellular lesions a descriptive nomenclature was set out by an international panel of experts sponsored by the World Congress of Gastroenterology 1994. The American College of Gastroenterology has recently released guidelines for the management of benign tumors of the liver that were crafted by the rating of the strength of recommendations and the quality of evidence while highlighting the added value of collecting a detailed history and performing a physical examination of the patient during the evaluation of a focal liver lesion identified with an imaging modality. This chapter focuses on hemangioma, focal nodular hyperplasia (FNH), hepatocellular adenoma (HCA), and large nodular regenerative hyperplasia (NRH)—lesions that have become more frequently diagnosed following the widespread use of imaging tests which led to the recognition of an increased number of affected patients. A molecular classification has recently been proposed for HCA which has changed our understanding of this condition with respect to diagnosis, prognostication, and management.

TABLE 49-1
Classification of Benign Nodular Lesions of the Liver
Hepatocellular Regenerative lesions
Monoacinar regenerative nodule
Diffuse nodular hyperplasia without fibrous septa (nodular regenerative hyperplasia)
Diffuse nodular hyperplasia with fibrous septa or in cirrhosis
Multiacinar regenerative nodule
Lobar or segmental hyperplasia
Focal nodular hyperplasia
Dysplastic or neoplastic lesions
Hepatocellular adenoma
Dysplastic focus
Dysplastic nodule
Biliary Bile duct adenoma
Biliary hamartoma
Biliary cystoadenoma
Biliary papillomatosis
Vascular Hemangioma
Infantile hemangioendothelioma
Hereditary hemorrhage telangiectasia
Lymphangiomatosis
Mesenchymal Leiomyoma
Lipoma
Myeloid lymphoma
Angiomyolipoma
Pseudolymphoma
Fibrous mesothelioma
Hamartoma
Benign teratoma

Hepatic Hemangiomas

Hepatic hemangiomas are benign, vascular tumors of the liver, the second most common liver mass after metastatic cancer.

Epidemiology

Hemangiomas affect 0.7% to 1.5% of the general population, being typically discovered incidentally during evaluation of nonspecific abdominal complaints. The prevalence of hemangiomas has been overestimated in autopsy series (from 0.4% to as high as 20%) because of overrepresentation of elderly patients with comorbid illnesses ( Table 49-2 ). The age at diagnosis is typically between 30 years and 50 years but hemangiomas can be diagnosed in all age groups. The reported female : male ratios range from 1.2 : 1 to 6 : 1. Most hemangiomas are small (< 4 cm). Liver nodules larger than 4 cm are defined as cavernous hemangiomas.

TABLE 49-2
Epidemiology of Liver Nodules in Normal or Near-Normal Liver
Liver Nodule Setting Prevalence
Hemangioma Autopsy
Clinical series
0.4-20.9%
0.7-1.5%
Focal nodular hyperplasia Autopsy
Clinical series
0.3-3.0%
0.03%
Hepatocellular adenoma Females who have never used OCs
Female long-term OC users
Males
0.001-0.0013%
0.003-0.004%
Very rare
Nodular regenerative hyperplasia Autopsy series 2.1-2.6%
OC, Oral contraceptive.

Pathogenesis

The pathogenesis of hemangioma is unknown. Congenital hamartoma is a likely candidate. Alternatively, hepatic hemangiomas could result from dilation of existing blood vessels in tissues that developed normally. A pathogenic role of sex hormones has been postulated, because of consistent female predominance in larger tumors and tumor enlargement/recurrence in hysterectomized women under estrogen replacement therapy and in patients with a long-term use of oral contraceptives. Cavernous hemangiomas have been observed to increase in size during pregnancy and often display estrogen receptors, whereas no direct causal link between oral contraceptive use and hepatic hemangioma was observed in a case-control study. Moreover, tumor growth was also induced or influenced by drugs such as metaclopramide.

Pathology

Macroscopically, the tumors are ovoid, soft, reddish-purple or blue masses separated from the surrounding parenchyma by a fibrous pseudocapsule. Various degrees of fibrosis, hyalinization, calcification, thrombosis, and shrinking are seen. Extensive fibrosis and hyalinization, with narrowing or obliteration of vessels, are typical for sclerosed hemangiomas. Calcifications are rare, punctiform, and either central or peripheral. They attest to the presence of phleboliths, most often visible in sclerosed hemangiomas or giant reshaped hemangiomas. Microscopically, hemangiomas are vascular abnormalities characterized by multiple blood-filled sinusoidal spaces and vascular lakes lined by endothelial cells. Vascular channels are separated by a fibrous tissue. They are fed by hepatic artery branches, and their internal circulation is slow. Blood vessels and arteriovenous shunting may be seen in large septa. The tumor seems to grow by ectasia rather than by hyperplasia or hypertrophy.

Clinical Features

Characteristically, hemangioma is an incidental finding on hepatic imaging. Diagnosis was incidentally made in 23% to 62% of cases, whereas clues to diagnosis were abdominal pain in 23% to 53% of cases, suspected metastases in 3% to 16% of cases, palpable mass in 0% to 2% of cases, and nonspecific complaints in 5% to 9% of cases. Most patients will have a single tumor node. However, a few patients may present with isolated diffuse hemangiomatosis, and this may happen in association with Rendu-Osler-Weber disease or skeletal hemangiomatosis. Cavernous hemangiomas are typically silent, benign clinically, and rarely expanding or symptomatic tumors. The few patients with symptoms complain of one of the following: abdominal swelling, upper-right abdominal pain, early satiety, dyspepsia, anorexia, abdominal mass, and hepatomegaly. The presence of symptoms correlated with the size of hemangiomas in one study but not in another study. In addition, there seems to be no correlation between symptoms and the number of tumors. Atypical hemangiomas exist that form arteriovenous shunts that may cause severe symptoms, including heart failure. Other unusual clinical presentations of a hepatic hemangioma include hemobilia, caval thrombosis, portal hypertension and torsion of a pedunculated tumor, inflammatory pseudotumor, recurrent intranodular thrombosis (rarely associated in giant hemangioma with thrombocytopenia) microangiopathic hemolytic anemia, and consumption coagulopathy, so-called Kasabach-Merritt syndrome.

Diagnosis

Liver function test findings are typically normal in patients with hemangiomas. A few patients with cavernous hemangiomas may present with coagulopathy (i.e., thrombocytopenia and hypofibrinogenemia). Fine-needle aspiration biopsy is considered reasonably safe in hepatic hemangiomas as long as the aspiration route is through a layer of normal liver tissue, but lacks sensitivity. The aspirate consists mainly of blood with only few noncharacteristic spindle cells of benign appearance. In 36 consecutive fine-needle aspiration biopsies of liver hemangiomas in Helsinki, a cytologic diagnosis of hemangiomas was obtained in 21 cases (58%). One patient (3%) had uneventful intraperitoneal bleeding after the aspiration. The sensitivity of percutaneous biopsy (microhistology) for the diagnosis of hemangioma is high (75% to 91%), with a specificity of 100%. Biopsy should be avoided if the radiologic features of a hemangioma are present.

Imaging Findings

Ultrasonography and Contrast-Enhanced Ultrasonography

The most common ultrasonographic appearance of hemangioma is that of a sharply demarcated lesion with uniformly increased echogenicity relative to normal liver ( Fig. 49-1 ). This pattern is observed in approximately 70% of hemangiomas detected by ultrasonography. The remaining cases show atypical ultrasound patterns, and appear either as hypoechoic lesions with an hyperechoic border or as lesions with heterogeneous internal structure. Heterogeneity is commonly observed in large hemangiomas. Although there are no vascular patterns that can be used to reliably diagnose hemangioma with conventional color or power Doppler ultrasonography, early clinical experience with ultrasound contrast agents has suggested that contrast-enhanced ultrasonography can provide useful information. Most liver hemangiomas (78% to 93%) show peripheral nodular enhancement during the early phase of the contrast-enhanced study, with progressive centripetal fill-in. Diffuse contrast enhancement with homogeneous fill-in or persistent hypoechoic appearance due to absent contrast enhancement can be observed in small, high-flow hemangiomas or thrombosed hemangiomas, respectively.

Fig. 49-1, Hemangioma, ultrasonography. The lesion shows typical features and appears as a round, well-defined, hyperechoic nodule ( arrow ).

Computed Tomography

The standard spiral computed tomography (CT) protocol for suspected hemangioma includes baseline and contrast-enhanced scanning in the arterial, portal venous, and delayed phases. Most hemangiomas are hypoattenuating on baseline scans. There is discontinuous peripheral nodular enhancement of vascular attenuation on arterial-phase imaging isoattenuating to aorta with progressive centripetal fill-in and hyperattenuation with possible central hypoattenuation or isoattenuation to the vascular space in the delayed phase. This pattern has a sensitivity of 67% to 86% and a specificity of 99% to 100% for the diagnosis of hemangioma. Atypical CT features are observed in hemangiomas with either high flow or very slow flow. High-flow hemangiomas show rapid filling after contrast agent administration, resulting in homogeneous enhancement during the hepatic arterial or portal venous phase. This feature is relatively common in small hemangiomas. Differentiation of high-flow hemangioma from hypervascular malignant tumors may be difficult, and relies on attenuation equivalent to that of the aorta during all phases of CT, including the delayed phase. Very slow flow hemangiomas appear either as nonenhanced lesions or as lesions with weak peripheral enhancement without centripetal progression. These features may be related to thrombosis or abundant fibrosis, and mimic a hypovascular malignant tumor.

Magnetic Resonance Imaging

The magnetic resonance imaging (MRI) protocol for characterizing suspected hemangioma includes gradient-echo T1-weighted sequences, fast spin-echo T2-weighted sequences with short and long (> 200-msec) echo times, and serial dynamic gadolinium-enhanced gradient-echo T1-weighted sequences. Hemangioma appears as a homogeneous focal lesion with smooth, well-defined margins. The lesion is hypointense compared with liver parenchyma on T1-weighted MRIs and strongly hyperintense on T2-weighted MRIs relative to spleen. The high signal intensity on heavily T2-weighted (long echo time) MRIs gives hemangiomas a consistent light-bulb pattern with 100% sensitivity and 92% diagnostic specificity. Dynamic contrast-enhanced MRI shows a quite typical perfusion pattern in hemangioma; that is, discontinuous peripheral nodular enhancement in the early phase with centripetal progression to uniform or almost uniform enhancement during the portal venous and the delayed phase ( Fig. 49-2 ). Such a characteristic enhancement pattern has a sensitivity of 77% to 91% and a specificity of 100% for the diagnosis of hemangioma. However, very small (< 1.5 cm), high-flow hemangiomas frequently exhibit a hypervascular pattern, with uniform enhancement in the arterial phase, which may persist in the portal venous and delayed phases ( Fig. 49-3 ). In these cases, diagnostic assessment may be difficult and requires careful analysis of baseline and contrast-enhanced images. Hemangioma shows a peculiar feature after the injection of reticuloendothelial system (RES)–targeted MR agents (i.e., lesion hyperintensity on T1-weighted postcontrast MRIs). This is due to the T1 effect of superparamagnetic iron oxide particles trapped within the slow-flow vascular channels of the lesion. MRI is preferred to CT in cases where there is a nodule smaller than 3 cm or where a nodule is found close to the heart or intrahepatic vessels.

Fig. 49-2, Hemangioma, magnetic resonance imaging (MRI).

Fig. 49-3, Hemangioma, magnetic resonance imaging (MRI).

Technetium-99m-Labeled Red Blood Cell Scintigraphy

Technetium-99m-pertechnetate-labeled red blood cell scintigraphy is a relatively specific examination for characterizing hemangioma. With this method there is decreased activity on early dynamic images and increased activity on delayed blood pool images. Comparison between 99m Tc-pertechnetate-labeled red blood cell single-photon emission CT (SPECT) and MRI showed that MRI had higher sensitivity and specificity than SPECT, especially for lesions smaller than 2 cm in diameter.

Diagnostic Workup

The recommended diagnostic workup for suspected hemangioma is dependent on the clinical scenario. If a hemangioma with typical ultrasound features is incidentally detected in a patient with neither a history of malignancy nor chronic liver disease, no additional investigation may be required. It has been shown that in this clinical setting the risk of misinterpreting a malignant tumor for an hemangioma is negligible (0.5%). Conversely, in incidental lesions with ultrasound features atypical of hepatic hemangioma, further diagnostic workup is recommended. Additional investigation is also mandatory—regardless of the ultrasound features—for any lesion detected in a patient at increased risk of malignancy. MRI is currently the most accurate technique for diagnostic confirmation of suspected hemangioma, and it is preferred in cases where the lesion is smaller than 3 cm or close to intrahepatic vessels or the heart. Despite promising results obtained in recent investigations, contrast-enhanced ultrasonography is at an early stage of clinical application and it can increase both sensitivity and specificity in diagnosis. On the other hand, spiral CT has limitations in achieving a reliable diagnosis of small lesions, especially in the setting of cirrhosis. Indeed, high-flow hemangiomas that are enhanced homogeneously in arterial-phase CT scans may not be confidently distinguished from small hypervascular carcinoma. Because MRI has greater sensitivity to small differences in contrast enhancement and because several fast MR sequences can be used to track the passage of a small, tight bolus of contrast material, MRIs may show the characteristic enhancement patterns of hemangioma and hepatocellular carcinoma (HCC) better than CT images. In addition, MRI, besides the information provided by the dynamic gadolinium-enhanced study, can offer improved capability of lesion characterization through the analysis of lesion signal intensity on baseline sequences, especially heavily T2-weighted sequences. In the setting of cirrhosis, diagnosis of hemangiomas should meet strict CT or MRI criteria, whereas percutaneous biopsy can be used to solve an uncertain diagnosis. Echo-guided liver biopsy should be avoided if the radiologic features of a hemangioma are present, whereas in small nodules of uncertain diagnosis, a close follow-up with imaging techniques is more prudent than a biopsy.

Treatment

A conservative approach is recommended for the management of hepatic hemangiomas. The large majority of these lesions remain asymptomatic and stable over time, making surgical treatment for the prevention of rare complications not justified. There is some suggestion that very large lesions present an added risk of spontaneous rupture or rupture because of trauma, but reports in the literature of such events are sparse.

Treatment is not indicated for asymptomatic tumors that are smaller than 5 cm. Current indications for surgical management of these benign liver masses include uncertain diagnosis with a suspicion of malignancy, severe or progressive symptoms due to size, and less commonly, risk of hemorrhage or rupture. Treatment of symptomatic patients with impaired quality of life can be done with surgical or nonsurgical therapeutic modalities by an experienced team. Elective surgical resection has been advocated in selected young patients with an asymptomatic, larger than 10 cm hemangioma to eliminate the risk of hemorrhage, thrombosis, and rupture. Enlarging tumors and those that became symptomatic have been successfully treated by resection, with relief of symptoms in 90% of cases. Patients with multiple hemangiomas and those with extensive hilar involvement may be considered for either angiographic embolization or liver transplant. The former treatment is indicated for patients with one or few tumors that have favorable vascular anatomy as a debulking therapy before surgery to reduce blood loss at the time of surgery or to treat atypical tumors with arteriovenous shunts. When local ethanol injection therapy was delivered to 37 patients with symptomatic hemangiomas (41% with multiple nodes, 60% with cavernous tumors), the tumors shrank in 27 of the patients (73%) and pain disappeared in 10 of the 29 patients (35%) with symptoms.

Liver transplant may be indicated for large, unresectable tumors, for extensive multiple tumors, or when surgical resection is not feasible. Hepatic resection and transcatheter hepatic embolization are effective treatment modalities for Kasabach-Merritt syndrome, although in a few instances unresectable cavernous hemangiomas with this complication may be an indication for liver transplant. There is little evidence for the efficacy of radiotherapy, which carries a risk of radiation hepatitis. In a PubMed MEDLINE search, 32 patients with a spontaneously ruptured hemangioma (range 6-25 cm in diameter) were identified. Thirteen patients (59%) underwent hepatic resection, five patients (23%) underwent suture, and four patients (18%) underwent tamponade. Three of the 13 patients who underwent resection, 2 of the 5 patients who underwent suture, and 3 of the 4 patients who underwent tamponade died.

Treatment of symptomatic patients with impaired quality of life can be done with surgical or nonsurgical therapeutic modalities by an experienced team.

Prognosis and Natural History

Limited series of longitudinal studies assessing changes in the volume of hemangiomas during follow-up showed a decrease in size ranging from 0% to 7%, and an increase ranging from 0% to 11%. The observed increase in size of the lesions is thought to result from progressive ectasia rather than hyperplasia or hypertrophy.

The risk that a patient with an asymptomatic hemangioma will develop abdominal pain is negligible. In a series of patients presenting with abdominal pain, pain disappeared in most patients either after treatment of comorbidities or without specific treatment. Pain persisted in two thirds of patients who underwent treatment of hemangioma with hepatic resection, embolization, or artery ligation. The mechanisms for development of symptoms or pain are unclear but may include expansion of tumor size with pressure effects on adjacent hepatic parenchyma or Glisson's capsule. In a few cases, symptoms relate to intralesional hemorrhage, localized thrombosis, or torsion of a pedunculated hemangioma. Kasabach-Merritt syndrome (i.e., disseminated intravascular coagulation in the setting of a cavernous hemangioma of the liver and cutaneous hemangiomas) presents with abdominal pain and signs of bleeding. During a follow-up of 3 months to 180 months, a tiny minority of patients (10%) with a hemangioma showed a decrease of tumor size, whereas in a similar percentage of patients the tumor grew in size ( Table 49-3 ).

TABLE 49-3
Longitudinal Studies Assessing Changes of Volume of Hemangiomas During Follow-Up
Authors No. of Patients Median Follow-up (mo) * TUMOR SIZE
Decreased Increased
Gandolfi et al. 123 22 (12-60) 0 1 (1%)
Farges et al. 78 92 0 1 (1%)
Weimann et al. 104 32 (7-123) 7 (7%) 11 (11%)
Terkivatan et al. 78 45 (24-72) 0 1 (1%)
Okano et al. 64 19 (6-58) 1 (2%) 0

* The range is given in parentheses.

The risk of rupture of hepatic hemangiomas in negligible, the only reports involving patients with cavernous hemangiomas undergoing trauma-induced or spontaneous rupture. Rupture is associated with sudden onset of severe abdominal pain, abdominal distension, hypotension or shock, and increased serum levels of transaminases and prothrombin time. A PubMed MEDLINE search has identified up to 32 cases of spontaneous rupture of hepatic hemangioma (mostly cavernous hemangiomas) in adults without a history of trauma.

According to the American College of Gastroenterology recommendations, pregnancy and the use of oral contraceptives or anabolic steroids are not contraindicated in patients with liver hemangioma, and follow-up imaging is not required in cases of classic hemangioma.

Focal Nodular Hyperplasia

FNH of the liver is a rare, completely benign lesion characterized by nodular hyperplasia of hepatic parenchyma around a central scar containing an anomalous artery, usually occurring in a normal liver and frequently multinodular.

Epidemiology

FNH is the second most common benign tumor of the liver, with an estimated prevalence of 0.4% to 3% in unselected autopsy series, whereas clinically relevant cases of FNH are rare, with a reported prevalence of 0.03%. The tumor has a female : male ratio of between 2 :1 and 2.6 :1, and the average age at presentation is between 35 years and 50 years. FNH is frequently associated with liver hemangioma (20% of cases), with other FNH (multiple FNH in 20% to 30% of cases), and more rarely with HCA.

Pathogenesis

FNH appears to be the result of a hyperplastic response of the hepatic parenchyma, mainly composed of reactive polyclonal proliferating hepatocytes, to an arterial lesion and/or portal venous malformation. Clonality studies, mainly using human androgen receptor assays (HUMARAs), showed the polyclonal origin of FNH in 50% to 100% of cases. The consequent formation and enlargement of arterial to venous shunts causes hyperperfusion in local arteries, resulting in oxidative stress that triggers a response from hepatic stellate cells to produce the central scar typically seen in cases of FNH. FNH can be associated with other vascular abnormalities, such as hepatic hemangioma and hereditary hemorrhagic telangiectasia (Rendu-Osler-Weber disease) or the congenital absence of the portal vein.

The finding of an unbalanced expression of the angiopoietin 1 and angiopoietin 2 genes coupled with expression of angiopoietin 1 protein by the endothelial cells of dystrophic vessels suggests a role of angiopoietin genes in FNH. Clonality studies and overexpression of important genes involved in cell homeostasis such as the Bcl-2 and transforming growth factor α genes support the important role of hepatocellular proliferation in FNH. Conversely, the role of oral contraceptives in FNH is disputed. A hospital-based case-control study in women with histologically proven FNH showed a quantitatively proportional increase in the risk of FNH in patients who had ever used oral contraceptives. Their use has been associated with an increase in size and vascularity of FNH nodes, and tumor regression was observed after drug withdrawal. However, the association between pregnancy, estrogen, and FNH was negated by an 8-year study in 216 women in Paris.

You're Reading a Preview

Become a Clinical Tree membership for Full access and enjoy Unlimited articles

Become membership

If you are a member. Log in here