Other Malignant Hepatic Tumors

Epidemiology

A recent study described the epidemiology of FLHCC in the United States using the Surveillance, Epidemiology, and End Results (SEER) database, which covers approximately 26% of the U.S. population. A total of 46,392 HCC cases were identified between 2000 and 2010, whereas only 191 FLHCC cases (0.4% of HCC cases) were identified. The incidence rate of FLHCC was reported to be 0.2 per million person-years. There was no significant change in the incidence rates between 2000 and 2010. There was a male preponderance, with a male to female case ratio of 1.7. The incidence rates are similar between different racial and ethnic groups. FLHCC occurs in young individuals; 60% of patients presented when they were younger than 40 years, whereas only 2% of patients with HCC presented when they were younger than 40 years. FLHCC accounts for 11% of liver cancer patients younger than 40 years. Although data on the global epidemiology of FLHCC are scarce, there appears to be some geographic variation in the epidemiology of FLHCC. For example, the proportion of FLHCCs among primary liver cancers was as high as 5% in a study from Mexico.

Pathogenesis

The pathogenesis of FLHCC is unique and distinct from that of classic HCC. There are no established risk factors for FLHCC. Frequently detected mutations in classic HCC, such as TP53 and CTNNB1 mutations are not usually found. Overexpression of anterior gradient 2, which was shown to inhibit the function of p53, was reported in most FLHCCs, whereas overexpression of anterior gradient 2 was very rare in patients with HCC. Overexpression and extra copy numbers of EGFR, overactivation of mammalian target of rapamycin, Ras, mitogen-activated protein kinase, and phosphatidylinositol-3-kinase, and xenobiotic degradation pathways have also been described in the pathogenesis of FLHCC.

The molecular pathogenesis of FLHCC remained poorly understood until recently, when a number of key studies shed substantial light on the molecular pathogenesis of FLHCC. First, a novel chromosomal deletion was identified on chromosome 19, resulting in a chimeric transcript encompassing portions of the messenger RNAs of PRKACA and DNAJB1 . This DNAJB1 - PRKACA chimeric transcript was not detected in any nontumor samples. Western blot analysis confirmed the translation of this chimeric RNA transcript exclusively in the tumor. Although the mechanistic roles of the chimeric protein in the pathogenesis of FLHCC remain to be further investigated, the presence of the DNAJB1-PRKACA chimeric transcript in 100% of the FLHCCs tested so far raises the possibility that this genetic change plays a major role in the pathogenesis of FLHCC.

Another study analyzed 78 FLHCC samples using a whole-transcriptome single-nucleotide polymorphism array and next-generation sequencing. Unsupervised gene expression clustering revealed three molecular subclasses of FLHCC, designated the proliferation, inflammation, and unannotated classes. Several copy number variations, such as amplification at 8q24.3 and deletions at 19p13 and 22q13.32, were reported. An eight-gene signature was reported that predicted overall survival in patients with FLHCC. Further genomic profiling of FLHCC will provide key information for understanding the molecular pathogenesis of FLHCC and developing targeted treatment for FLHCC.

Pathology

The distinct pathologic features of FLHCC were first reported approximately 60 years ago. Histopathologically, FLHCC typically shows well-differentiated, large polygonal malignant cells with eosinophilic hyaline bodies and granular cytoplasm that are embedded in lamellar bands of thick fibrous tissue in the absence of pathologic changes in the surrounding liver ( Fig. 48-1 ). Rarely, glandular type differentiation with mucin production or clear cell–type changes are reported as a variant form of FLHCC. FLHCC shares both hepatocellular and cholangiocellular differentiation. FLHCC tissues are positive for albumin messenger RNA by in situ hybridization, hepatocyte paraffin 1, and polyclonal carcinoembryonic antigen, confirming their hepatocellular differentiation. On the other hand, they stain positive for cytokeratins 7 and 19 and epithelial cell adhesion molecule, suggesting a bile ductular differentiation.

Clinical Features

Because of their young age and low probability of cancer, most patients with FLHCC are not diagnosed until the tumor progresses sufficiently to stretch the liver capsule and exert a sufficient mass effect to cause symptoms. As a result, less than 20% of patients present with primary tumors smaller than 5 cm. The most common symptoms include abdominal pain, abdominal distension, and systemic cancer–related symptoms such as fatigue, weight loss, anorexia, and nausea. Patients often present with pain from metastatic disease in other organs, such as joint or bone pain from skeletal metastasis. They may present with ascites due to peritoneal metastatic disease. Rarely, FLHCC patients present with fulminant hepatic failure due to massive replacement of the liver by the tumor. Patients may also present with thrombophlebitis of the lower extremities or hyperthyroidism as paraneoplastic phenomena.

There are no established serum tumor biomarkers sensitive or specific for FLHCC. Most patients do not have an elevated serum alpha fetoprotein (AFP) level. Although one study showed that all 10 patients with FLHCC studied had elevated des-γ-carboxyprothrombin levels, the diagnostic and prognostic value of des-γ-carboxyprothrombin requires further study. The levels of transaminases and bilirubin are often normal or mildly elevated at initial presentation.

Diagnosis

When young patients present with liver masses without characteristic radiologic features of HCC or cholangiocarcinoma or evidence of chronic liver disease, FLHCC should be suspected. FLHCC typically appears as a sharply demarcated lobulated heterogeneous mass with a central scar. Radiologic features of FLHCC may overlap with those of other scar-producing lesions, including focal nodular hyperplasia. Thus, ultrasound- or CT-guided liver biopsy should be strongly considered for the diagnostic confirmation of FLHCC. This will exclude hepatic metastasis from other primary sites or other rare primary liver cancers or benign tumors. Cross-sectional images (CT or MRI) are useful to delineate the extent of the disease and decide on the optimal treatment approach ( Fig. 48-2 ). Following the recent discovery of the DNAJB1 - PRKACA fusion transcript, which is specific to FLHCC, both fluorescence in situ hybridization and DNA sequencing assays have been developed to allow molecular confirmation of the diagnosis of FLHCC.

Treatment and Prognosis

Although limited treatment modalities exist for the management of FLHCC, overall the prognosis is better in patients with FLHCC compared with those with usual HCC ( Table 48-2 ). A study using the SEER database reported a 5-year survival probability of 33% in patients with FLHCC compared with 16% in patients with usual HCC. It is not clear whether the better survival in FLHCC patients is due to the lack of underlying liver disease or to a less aggressive cancer biology. The prognosis of nonmetastatic FLHCC patients was similar to that of patients with noncirrhotic HCC but better than that of patients with cirrhotic HCC.

The treatment of choice in patients with FLHCC is surgical resection. Among the FLHCC patients in the SEER database, approximately 41% had surgical resection. The technical feasibility of surgical resection dictates the overall prognosis. The 5-year survival probability in patients with FLHCC who had surgical resection was 58%, which is superior to the probability of 44% of their usual HCC counterparts who had surgical resection. On the other hand, the prognosis was extremely poor in FLHCC patients who did not have surgical resection, with a 5-year survival probability of only 7%.

Similar results were observed in a meta-analysis of 575 FLHCC patients. The 5-year overall survival probability was 44% in all FLHCC patients, and it was higher, at 70%, in patients who had surgical resection. The feasibility of surgical resection, free surgical borders, and the absence of vascular invasion are good prognostic indicators. Other studies reported hepatic dysfunction, larger tumor size, multiple liver masses, lymph node invasion, extrahepatic metastasis, and vascular invasion as poor prognostic indicators in patients with FLHCC. Unfortunately, most patients develop recurrent FLHCC after surgical resection, with a reported 5-year recurrence-free survival rate of 18%. Late recurrence is not uncommon, and re-resection should be considered whenever possible because of the lack of effective alternative treatments.

Liver transplantation (LT) does not appear to be as effective as surgical resection. The 5-year overall survival probability was 34% in 35 patients who underwent LT for FLHCC, with a median survival of 32 months, which was shorter than the median survival of 222 months in 90 patients who underwent surgical resection ( p < 0.001). This appears to be due to the extensive tumor burden at the time of LT for FLHCC. As opposed to the usual selection of patients with early-stage HCC for LT, LT has tended to be performed in patients with advanced, unresectable FLHCC, resulting in a high rate of tumor recurrence after LT.

The utility of nonsurgical modalities for treatment of FLHCC has not been well investigated. Only a small proportion of patients receive nonsurgical treatment. No chemotherapeutic regimens or targeted treatments have been shown to improve survival of patients with FLHCC in a randomized controlled trial (RCT). The use of chemotherapeutic agents has been mostly based on anecdotal experiences of a few cases. Although Phase II clinical trials of fluorouracil and recombinant interferon alfa-2b showed somewhat promising results (62% treatment response, including complete and partial responses) with tolerable adverse effects, the small number of patients with FLHCC ( n = 9) makes it hard to generalize this finding. Experiences with transarterial chemoembolization (TACE) and transarterial radioembolization (TARE) have been reported recently. One study showed that TACE is well tolerated and led to a reduction in the size of the tumor, which subsequently permitted safe hepatic resections in two patients. The therapeutic efficacy of locoregional and systemic chemotherapy or targeted therapy requires further investigation to standardize treatment and improve the outcomes of patients with FLHCC. With the discovery of the DNAJB1 - PRKACA fusion transcript, there is an active effort to determine whether agents targeting this fusion will show clinical utility.

Epidemiology

Although hepatoblastoma is the most common pediatric hepatic malignancy, the incidence rate is very low. Analysis of SEER data showed an incidence rate of 1.1 per million person-years between 1973 and 1997. The incidence rate increased from 0.6 per million person-years to 1.2 per million person-years between the 1973-1977 and 1993-1997 time periods ( p < 0.001). A more recent study showed that the incidence rate increased further to 1.4 per million person-years between 1992 and 2004, with an annual percent change of 4.3%. In the United States and other countries with low rates of early acquisition of hepatitis B virus infection, the incidence rate of hepatoblastoma is two to three times higher than the incidence rate of HCC in children. Most hepatoblastomas are diagnosed in children younger than 5 years. Consequently, the incidence rate of hepatoblastoma is approximately 20-fold higher than the incidence rate of HCC in children younger than 5 years, and hepatoblastoma accounts for 91% of primary hepatic malignancies in children younger than 5 years. By contrast, the incidence rate of HCC increases in older children, and HCC accounts for 87% of primary hepatic malignancies in individuals aged 15 years to 19 years. The incidence rates of hepatoblastoma are 1.3-fold higher in boys than in girls.

Pathogenesis

A number of genetic syndromes have been linked to an increased risk of hepatoblastoma, including Beckwith-Wiedemann syndrome, Edwards syndrome, and familial adenomatous polyposis. Aberrant activation of the Wnt signaling pathway is a key feature in the molecular pathogenesis of hepatoblastoma. Gain-of-­function mutation of β-catenin contributes to Wnt pathway activation, and nuclear accumulation of β-catenin was strongly correlated with poorly differentiated hepatoblastomas. A study using microarray analysis identified two subclasses of hepatoblastoma on the basis of a 16-gene signature that also predicts patient prognosis. Up-regulation of the Myc signaling pathway was associated with the immature subtype of hepatoblastomas, whereas downregulation of the Myc signaling pathway was associated with impaired tumorigenesis. High expression of telomerase reverse transcriptase is associated with Wnt signaling pathway activation in the absence of β-catenin mutation and was also associated with an aggressive phenotype of hepatoblastomas.

Pathology

Hepatoblastoma is an embryonal cancer that arises from primitive hepatocyte precursor cells that have the potential to differentiate into different morphologic features, such as fetal and/or embryonal hepatocytes, cartilage, bone, striated muscle fibers, and squamous epithelium. Grossly, tumors are well-circumscribed, nodular masses with foci of hemorrhage and necrosis with or without mesenchymal components such as bone or cartilage. Recently, the Children's Oncology Group Liver Tumor Committee proposed a histopathologic classification of hepatoblastoma ( Table 48-3 ). The two main histologic subtypes are an epithelial subtype and a mixed epithelial and mesenchymal subtype. The epithelial subtype accounts for two thirds of hepatoblastomas, and can be further subclassified into fetal, embryonal, macrotrabecular, small cell undifferentiated, and cholangioblastic subtypes. Fetal hepatoblastoma can be further subclassified into well-differentiated (with uniform round nuclei organized into cords with minimal mitotic activity and extramedullary hematopoiesis), crowded (a mitotically active subtype), pleomorphic (with poorly differentiated cells, hyperchromatic nuclei, and scant cytoplasm), and anaplastic (with marked nuclear enlargement and pleomorphism) subtypes. The embryonal subtype is characterized by small, basophilic, darkly stained cells with uniform, hyperchromatic nuclei and scant cytoplasm and is often accompanied by extramedullary hematopoiesis. The macrotrabecular subtype is characterized by clustered growth of either fetal or embryonal cells between sinusoids. Small cell undifferentiated hepatoblastoma, which has a pathologic appearance similar to that of neuroblastoma or other small blue cell tumors, represents 2% to 5% of hepatoblastomas. This subtype of tumor is highly invasive, with poor responsiveness to chemotherapy. Most children with this tumor subtype die within 2 years of diagnosis. Lastly, the cholangioblastic subtype is characterized by a predominance of biliary cells at the periphery of epithelial islands. The mixed subtype can be further classified into mixed types with stromal derivatives or teratoid features. The stromal derivative subtype is characterized by the formation of blastema, bone, skeletal muscle, and cartilage. The teratoid subtype is characterized by mixtures of primitive endoderm, neural derivatives, melanin, and squamous and glandular elements.

Clinical Features

Patients with hepatoblastoma typically present with abdominal distension. Because of rapid growth, the tumor may rupture, leading to hemoperitoneum. Sexual precocity may be present because of the ectopic synthesis of gonadotropin. Hepatoblastomas commonly occur as a single hepatic mass. Serum AFP levels are markedly elevated in most patients. AFP is a clinically useful diagnostic and prognostic tumor marker in the management of hepatoblastoma. Both normal and markedly elevated AFP levels are associated with poor prognosis. Serial monitoring of AFP is particularly useful in monitoring treatment response and tumor recurrence after curative treatment.

Diagnosis

Doppler ultrasonography is useful for identifying liver masses and to determine the presence of tumor vascular invasion. Once a liver mass has been detected, contrast-enhanced CT or MRI should be performed for anatomic delineation of the tumor extent to determine the potential resectability of the tumor. MRI with a hepatobiliary specific contrast medium such as gadoxetate disodium has been shown to be highly effective in characterizing tumor extent and the presence of satellite nodules or vascular invasion. The diagnosis of hepatoblastoma should be confirmed by biopsy and histology. Because of the heterogeneity of hepatoblastomas, it is recommended to obtain a minimum of 5 to 10 cores for adequate tissue diagnosis.

Prognosis and Treatment

The prognosis of hepatoblastoma is better than that of HCC but worse than that for other pediatric malignancies. In 2003 the 5-year survival probability was reported to be 52% in children with hepatoblastoma versus 18% in children with HCC ( p < 0.0001). Surgical resection with or without neoadjuvant and/or adjuvant chemotherapy is the primary treatment strategy for hepatoblastoma. Tumors composed of pure populations of fetal cells carry a better prognosis than embryonal or mixed epithelial-mesenchymal hepatoblastomas or small cell undifferentiated hepatoblastomas. Thus adjuvant chemotherapy is not recommended in patients with early-stage pure fetal subtype tumors with low mitotic activity, as surgical resection alone is a sufficient treatment. In patients presenting with advanced disease, neoadjuvant chemotherapy can be administered to downstage the tumor to achieve a complete resection. In patients with early-stage disease, adjuvant chemotherapy is typically recommended after surgical resection. LT can be considered in patients with extensive disease burden in whom complete resection of the tumor is not feasible. LT is still feasible in patients who initially present with metastatic disease, as long as the metastatic lesions regress with neoadjuvant chemotherapy; however, it is contraindicated in patients with chemotherapy-resistant, progressive metastatic disease.

Platinum-based treatment regimens are typically used as neoadjuvant or adjuvant chemotherapy for hepatoblastoma. Cisplatin is the most commonly used chemotherapeutic agent, administered alone or in combination with 5-fluorouracil, vincristine, or doxorubicin.