Fax: (011) 31 20566-9287
Review Article
Where do we stand with hepatoblastoma?
A review
Article first published online: 16 JUL 2003
DOI: 10.1002/cncr.11585
Copyright © 2003 American Cancer Society
Additional Information
How to Cite
Schnater, J. M., Köhler, S. E., Lamers, W. H., von Schweinitz, D. and Aronson, D. C. (2003), Where do we stand with hepatoblastoma?. Cancer, 98: 668–678. doi: 10.1002/cncr.11585
Publication History
- Issue published online: 1 AUG 2003
- Article first published online: 16 JUL 2003
- Manuscript Accepted: 19 MAY 2003
- Manuscript Revised: 27 APR 2003
- Manuscript Received: 31 DEC 2002
Funded by
- TerMeulen Fund
- Royal Netherlands Academy of Arts and Sciences
- Abstract
- Article
- References
- Cited By
Keywords:
- pediatric tumors;
- phenotypic characteristics;
- staging systems;
- treatment;
- prognosis
Abstract
Hepatoblastoma (HB) is the most common pediatric liver malignancy, comprising approximately 1% of all pediatric cancers. The disparate clinical staging systems and histologic classifications that were developed during the last decades, nevertheless, reflect the remaining difficulties and uncertainties in characterizing HB. Furthermore, the combination of surgery and (neo)adjuvant chemotherapy has improved patient outcomes dramatically. A poor prognosis is associated with large tumor size, multifocality, extrahepatic disease, and metastatic spread. The exact etiology of HB remains unknown, but the cytogenetic alterations, phenotypic features, and biologic aspects that accompany this neoplasm yield more and more insight into its pathogenesis. New cell-biologic and molecular-biologic insights may lead to the development of new treatment modalities, especially for patients with a bad prognosis. This review summarizes the different aspects of this intriguing tumor and discusses the current status of research and treatment for patients with HB. Cancer 2003;98:668–78. © 2003 American Cancer Society.
DOI 10.1002/cncr.11585
The first case report of hepatoblastoma (HB) most likely was published more than 100 years ago.1 In the last decade, several excellent clinical reviews regarding liver tumors in children have appeared, but a review of the biologic aspects of HB is lacking.2–6 In addition to diagnostic characteristics and current treatment modalities for patients with HB, this review discuses the currently known phenotypic features, cytogenetic alterations, and possible pathogenetic role of cytokines, β-catenin, and the Wnt signaling pathway of this malignant tumor in more detail.
Epidemiology
Incidence
HB is a malignancy of the liver with a fairly constant annual incidence of 0.5–1.5 diagnoses per 1 million children age younger than 15 years in Western countries,7 although an increase has been reported in the U.S. It comprises 1% of all pediatric malignancies and affects mostly infants and young children between the ages of 6 months and 3 years, although neonates and adolescents with HB also have been reported. After neuroblastoma and nephroblastoma, primary epithelial tumors of the liver are the third most common intraabdominal neoplasms in children.8 HB is the most frequent liver tumor in Western countries. In Asia and Africa, hepatocellular carcinoma (HCC) occurs more frequently than HB, probably as a consequence of the greater prevalence of hepatitis B infection on those continents.
Risk factors
To date, no environmental risk factors for HB have been described; however, HB has been associated recently with prematurity or low birth weight.9, 10 Familial cases have been reported. In this respect, the coincidence of HB with familial adenomatous polyposis (FAP) and Beckwith–Wiedemann syndrome (BWS) is striking and suggests a role in the pathogenesis of HB for chromosomes 5 and 11, respectively.11, 12
Etiology
Cytogenetic alterations
Cytogenetic analysis of HB has not revealed a consistent pattern of chromosomal anomalies. The most common genetic aberrations are extra copies of chromosomes 1q, 2q, 7q, 8, 17q, and 20.12–18 However, to date, cytogenetic alterations have not been linked with a causal factor or with prognosis.19 Of more functional importance, loss of heterozygosity (LOH) of 11p15 has been observed in up to one-third of patients with HB, and LOH of chromosome 1p also in has been observed in approximately 33% of patients with HB.20, 21 LOH at 11p15, which is always of maternal origin,21 is nearly pathognomonic for patients with BWS, who have a greater risk of developing HB, Wilms tumor, and rhabdomyosarcoma.12 Important imprinted genes on 11p15.5 are p57KIP2, insulin-like growth factor 2 (IGF-2), and H19. p57KIP2 and H19 are tumor suppressor genes, whereas IGF-2 is a major fetal mitogen. In this context, it is noteworthy that IGF-2 transcription is affected by β-catenin mutations, which also seem to play a role in the development of HB. p57KIP2 is up-regulated in HB,22 which argues against its role as a suppressor gene. Loss of imprinting is noted for the maternally imprinted IGF-2 gene but, in HB, is not associated with increased expression of IGF-2 or decreased expression of H19, as it is in BWS and its other associated tumors.12 Mutations observed in the tumor suppressor gene p53 that often are reported in HB appear to play no pathogenetic role in the development of HB, because they are found in a large portion of all tumors.12, 23–25 Furthermore, overexpression of p53 was not correlated with patient survival.26
Pathogenesis
The pathogenetic mechanisms responsible for the development of HB remain unclear. HB, as an embryonal tumor, is derived from undifferentiated embryonal tissue. Rapidly growing HBs often are very sensitive to cytotoxic drugs and sometimes are very sensitive to radiotherapy.27, 28
Histologic clues.
The currently accepted hypothesis is that HB cells are derived from pluripotent hepatic stem cells.29–31 These stem cells, or oval cells, have retained the ability to differentiate into both hepatocytes and biliary epithelial cells and, accordingly, express markers for both cell types, a feature also found in HB.32 In addition, extramedullary intratumoral erythropoiesis and thrombopoiesis are found in HB,33 a feature normally present in the fetal liver.
Role of cytokines.
HB cells secrete interleukin 1β (IL-1β), which, in turn, induces IL-6 production in the surrounding fibroblasts and endothelial cells.34 Secreted IL-6 may induce the production of the acute phase protein β-2-microglobulin. Both IL-1β and IL-6 can stimulate the secretion of hepatocyte growth factor (HGF). HGF is expressed in childhood HB by fibroblasts and endothelial cells and functions as a paracrine growth factor for HB cells that express the HGF receptor c-met.35, 36 The rapid growth of recurrent, disseminated tumor and/or metastases that can occur in children with incompletely resected, nonpretreated HB may result from increased HGF secretion after they undergo hepatic resection.37
Pathogenetic pathway.
Epidemiologic studies revealed that HB occurs more often in families affected by FAP.38 FAP is caused by inactivation of the adenomatous polyposis coli (APC) tumor-suppressor gene that is localized on chromosome 5. The function of this gene is the down-regulation of β-catenin. In greater than 67% of patients with sporadic HB, alterations of the APC gene were observed.39 However, similar numbers of patients also show activating mutations of the β-catenin gene.40–42 Together, these data suggest that the Wnt signaling pathway plays a role in the development of HB. APC binds to β-catenin, thereby promoting its NH2-terminal phosphorylation of β-catenin by glycogen synthase kinase-3β (GSK-3β). This phosphorylation targets β-catenin for degradation by the proteasome system. Signaling through the WNT signal-transduction pathway (see Fig. 1) and deletions of the NH2-terminus of β-catenin inhibit this phosphorylation by inhibiting GSK-β3 and, thus, cause accumulation of unphosphorylated β-catenin in the cytosol. So-called activating mutations of β-catenin, in which the amino terminal phosphorylation target sequences are mutated or deleted, also cause accumulation of β-catenin in the cytosol. This β-catenin then translocates to the nucleus, where it interacts with the HMG-box transcription factors lymphoid enhancer factor 1, T-cell factor 3 (Tcf3), Tcf4, and pangolin to modulate the transcription of target genes such as c-Myc, cyclin D1, matrix metalloprotease-7, immunoglobulin transcription factor-2, and fibronectin (Fig. 1). Whereas the absence or presence of mutations in β-catenin is not of prognostic value, a predominantly nuclear (as opposed to cytosolic) localization of β-catenin shows significant correlation with shorter survival time in these patients.43 This most likely is because overexpression of the target genes of β-catenin may support HB progression.44 In this respect, it is intriguing that HGF also can induce transient β-catenin translocation to the nucleus in a WNT independent manner (Fig. 1).45 This may be one of the ways in which HGF exerts its described growth-promoting effect on HB.36
Diagnosis
Clinical features
HB is a tumor that typically affects infants and children younger than 3 years. There is a male predominance, and the tumor most likely occurs more frequently among white patients and in the right lobe of the liver.3, 46 HB usually presents as an asymptomatic abdominal mass. Weight loss, anorexia, emesis, and abdominal pain indicate advanced disease.5 Distant metastases, which are found in approximately 20% of patients at diagnosis, occur mostly in the lungs,47 but metastases of the central nervous and even eye metastases have been described.48, 49
Laboratory findings
Anemia and thrombocytosis are common findings in patients with HB.2 This finding most likely is related to the ability of HB cells to secrete IL-1β, which, in turn, induces IL-6 production in the surrounding fibroblasts and endothelial cells (see Role of cytokines, above). A recent study including patients with (non-HB) malignancies and a mouse model showed an IL-6 dependent increase in thrombopoietin (TPO) in both patients and mice and a concomitant increase of platelets in the mice.34, 50 This finding appears to explain the increased TPO levels and the commonly found thrombocytosis in patients with HB.2
A sensitive but nonspecific marker for the presence of HB is α-fetoprotein (AFP). Approximately 90% of patients with HB have highly elevated serum levels of AFP,51 which makes AFP a useful clinical marker for monitoring treatment effectiveness and tumor recurrence. Physiologically, high levels of AFP are expressed in the fetus, with concentrations declining to adult levels in the first 6 months after birth. AFP concentrations can be elevated in patients who have liver diseases associated with liver regeneration, i.e., hepatitis, cirrhosis, hemangioendotheliomas, HCC, germ cell tumors, testicular tumors, and gall bladder carcinomas.
Radiologic findings
Imaging techniques play an important role in the diagnosis, staging, and treatment of patients with HB. Because complete surgical resection is the cornerstone of permanent cure, exact localization and assessment of tumor extent is a prerequisite. Often, the initial diagnosis is made by abdominal ultrasound. HB presents on abdominal ultrasound as a well defined, hyperechoic, solid, usually noncystic, intrahepatic mass, frequently (60–70%) located in the right lobe of the liver.52 Characteristically, abdominal computed tomography (CT) scanning reveals a delineated mass with low attenuation compared with the surrounding normal liver parenchyma.53 Vascular involvement can be assessed with contrast enhancement. Both CT scanning and magnetic resonance imaging also can assess the segmental extension of the tumor and the exact topography of the hepatic vasculature. Angiography can be valuable, although the majority of centers reserve angiography for patients with more complicated disease. Lung metastases are detected with chest X-ray or CT scanning.47 Bone scanning is not recommended as a routine investigation because bone metastases are rare and bone scans may produce potentially misleading results.54
Diagnostic biopsy
A diagnostic biopsy often is omitted if the intention is to treat a tumor that is confined to a single liver lobe with surgery only.55, 56 Nevertheless, a biopsy often is recommended to all patients for accurate diagnosis. First, it may be unethical to give chemotherapy if there is no tissue diagnosis. Second, the elevated physiologic expression of AFP may persist after the age of 6 months. Finally, HCC, although it is very rare, has been reported in patients younger than 3 years, and the prognosis for patients with HCC is much worse compared with the prognosis for patients with HB. Thus, we believe that it is advisable for all patients to undergo a biopsy. The risk associated with a biopsy is low using current techniques, although complications like bleeding and infection do occur in 5–10% of patients.57 Furthermore, tumor spill after a percutaneous needle biopsy has been described.58
Histologic examination is the only way to ascertain the diagnosis, especially in patients with nonclassic tumors. Furthermore, to develop new treatment strategies and to improve outcome, pretreatment phenotypical classification of the primary tumor seems imperative.
Clinical staging
Until 1990, all systems for staging primary liver tumors (in children), including HB, were based on the findings at surgery or after surgery. Then, a staging system that was based exclusively on images obtained prior to surgery was developed in the first study of the International Society of Pediatric Oncology Liver Study Group (SIOPEL-1).59 This offered the possibility of including patients who did not undergo surgery, and the information obtained could be used to assess and, if necessary, adjust preoperative therapy. Nevertheless, both preoperative and postoperative staging systems use the same parameters for staging, namely, size, vascular invasion, extension and complexity of the primary tumor, and the absence or presence of metastases.
The postoperative staging systems.
The so-called TNM classification system, originally developed by Pierre Denoix between 1943 and 1952, was adopted by the International Union Against Cancer in 1958.60 The TNM classification system and stages are summarized in Table 1. In 1983, the Japanese committee on the TNM classification system modified the TNM system, defining T classification solely according to by the number of anatomic liver segments involved in the tumor: T1: tumor in one segment; T2: tumor in two segments; T3: tumor in three segments; and T4: tumor in four or more segments.61 A further simplification of the system for staging liver tumors was implemented by the Children's Cancer Study Group (CCSG) and the Pediatric Oncology Group (POG),62, 63 who described the following four stages: Stage I: complete resection of the tumor; Stage II: microscopic residual tumor; Stage III: macroscopic residual tumor; and Stage IV: distant metastases.
| Status/stage | Criteria |
|---|---|
| |
| Tumor classification | |
| T1 | Solitary, ≤ 2 cm, no vascular invasion, 1 lobe, no extrahepatic disease |
| T2 | Solitary, ≤ 2 cm, vascular invasion, 1 lobe, no extrahepatic disease |
| T2 | Not solitary, ≤ 2 cm, no vascular invasion, 1 lobe, no extrahepatic disease |
| T2 | Solitary, > 2 cm, no vascular invasion, 1 lobe, no extrahepatic disease |
| T3 | Solitary, > 2 cm, vascular invasion, 1 lobe, no extrahepatic disease |
| T3 | Not solitary, ≤ 2 cm, vascular invasion, 1 lobe, no extrahepatic disease |
| T3 | Not solitary, > 2 cm, with or without vascular invasion, 1 lobe, no extrahepatic disease |
| T4 | Not solitary, any size, with or without vascular invasion, > 1 lobe, extrahepatic disease |
| Stage grouping | |
| Stage I | T1 N0 M0 |
| Stage II | T2 N0 M0 |
| Stage IIIA, | T3 N0 M0 |
| Stage IIIB | T1–T3 N1 M0 |
| Stage IVA | T4 any N, M0 |
| Stage IVB | Any T, any N, M1 |
A retrospective analysis of 72 patients who were treated in the German Pediatric Liver Tumor Study HB89 showed that both the TNM classification system and the CCSG/POG staging system had highly significant predictive value for survival (P = 0.0001 and P = 0.0009, respectively). In that study, the Japanese TNM staging system had a lower predictive value (P = 0.0161), and its modified T classification (i.e., based on the number of liver segments involved) was irrelevant with regard to outcome (P = 0.1359) (Table 2).64
| Stage | Staging system | |||||
|---|---|---|---|---|---|---|
| TNMb | CCSG/POGc | Japanese TNMd | ||||
| % | No. | % | No. | % | No. | |
| ||||||
| Stage I | NA | 0 | 100 | 21 | 100 | 1 |
| Stage II | 96 | 24 | 50 | 6 | 91 | 23 |
| Stage III | 77 | 34 | 74 | 38 | 71 | 41 |
| Stage IV | 29 | 14 | 29 | 7 | 29 | 7 |
| P valuee | 0.0001 | — | 0.0009 | — | 0.0161 | — |
The preoperative staging system.
In 1990, SIOPEL adopted a new preoperative staging system, Pretreatment Extent of Disease (PRETEXT). This system is based on the branching pattern of the portal vein, which divides the liver into eight segments, and (noninvasive) imaging techniques. Tumors are classified into one of the four categories by determining the number of affected liver sector(s) (Fig. 2). Extrahepatic growth is indicated by adding one or more of the following letters: V: involvement of the hepatic/caval vein; P: involvement of the portal vein; E: the presence of extrahepatic tumor extension; and M: the presence of distant metastases (the VPEM parameters).

Figure 2. The first study of the International Society of Pediatric Oncology (SIOP) Liver Tumor Study Group (SIOPEL-1) Pretreatment Extent of Disease (PRETEXT) grouping system. The system divides the liver into four parts, called sectors. The left lobe of the liver (L) is divided into a lateral sector (Segments 2 and 3) and a medial sector (Segment 4), whereas the right lobe (R) is divided into an anterior sector (Segments 5 and 8) and a posterior sector (Segments 6 and 7). Tumors are classified into one of the four PRETEXT categories, depending on the number of liver sectors affected by the tumor. PRETEXT I: one sector involved; PRETEXT II: two sectors involved; PRETEXT III: two nonadjoining sectors free or three sectors involved; and PRETEXT IV: all four sectors involved. Extrahepatic growth is indicated by adding one or more of the following characters: V: hepatic/caval vein; P: portal vein; E: extrahepatic extension; and M: distant metastases. Reprinted with permission from: Schnater JM, Aronson DC, Plaschkes J, et al. Surgical view of the treatment of patients with hepatoblastoma: results from the first prospective trial of the International Society of Pediatric Oncology Liver Tumor Study Group (SIOPEL-1). Cancer 2002;94:1111–1120. Copyright © 2002 American Cancer Society. Reprinted by permission of Wiley-Liss, Inc., a subsidiary of John Wiley & Sons, Inc.
Although the PRETEXT system was developed mainly to assess the efficacy of neoadjuvant chemotherapy and to predict surgical resectability, it also had highly prognostic value for both overall survival and event free survival (Table 3).65 Of the VPEM parameters, M (lung metastases) was the only significant parameter that was relevant for survival.47, 65 The predictive, prognostic value for survival of the PRETEXT and TNM classification systems in patients who underwent surgical resection in the SIOPEL-1 study were similar (SIOPEL group, unpublished results). This means that the prognostic value of the PRETEXT system is as good as that of the postoperative staging systems but it also allows assessment of the effects of preoperative therapy.
| Groupb | 5-year OS (%) | 5-year EFS (%) |
|---|---|---|
| ||
| Group I | 100 | 100 |
| Group II | 91 | 83 |
| Group III | 68 | 56 |
| Group IV | 57 | 46 |
| P valuec | 0.001 | 0.0001 |
The available systems appear to agree on the major determinants of a patient's prognosis, in that tumor size, tumor extension, and multifocality, all factors that affect resectability directly, are the primary determinants of long-term survival. Distant metastases affect the prognosis negatively. To be able to compare the results of the different study groups, it was proposed in 1999 that all international groups would use the SIOPEL PRETEXT system along with their own staging system in their studies.66
Histologic classification
Currently, there still is no full agreement on the classification of HBs, and refinement of diagnostic criteria for HB is necessary to provide a reproducible classification system.67 Over 30 years ago, 2 HB subtypes were recognized68: the epithelial type, which contains predominantly epithelial tissue, and the mixed epithelial and mesenchymal type, which also contains tissues of mesenchymal derivation. A classification that was based on the degree of differentiation of HB cells was developed a few years later.69 Three histologic subtypes were distinguished. The poorly differentiated embryonal type was characterized by a tubular or glandular histology and consisted mainly of rosettes of elongated tumor cells as well as varying contributions of fetal cells and anaplastic cells. HB cells in the highly differentiated fetal type resembled normal hepatocytes with rare mitoses and were arranged in two or three cell-thick tumor cords, but a normal lobular architecture was not present. Finally, the anaplastic type, also described as the small cell undifferentiated type,70 was characterized by small cells with densely stained nuclei and scant cytoplasm. Subsequently, a macrotrabecular type of HB characterized by features similar to HCC in adults was added.71 More recently, an elaborate histologic classification with no less than six patterns was developed.72 Currently, most pathologists have returned to the original classification system of Ishak and Glunz68 and distinguish only two morphologic types of HB.73 The epithelial type contains embryonal cells or fetal cells and often contains a mixture of the two. In areas with well differentiated HB cells, extramedullary hematopoiesis often is quite prominent.68, 74 The mixed type contains mesenchymal tissue in addition to the epithelial elements. The simple division of HB into two morphologic types accommodates individual or regional variations within an individual classification and within more elaborate classifications. Irrespective of these extensive and elaborate efforts to develop a histologic classification system for HB, disagreements remain regarding whether a purely fetal histologic phenotype predicts a favorable prognosis56, 63, 64, 70, 75–77 and whether an anaplastic, small cell histologic phenotype predicts an unfavorable prognosis.70, 76, 78 Nevertheless, the purely fetal HB is the only histologic subtype that currently leads to a change in therapy in the current Children's Oncology Group protocol for HB (Protocol 9645), although it has been applied only to patients with Stage I tumors (i.e., completely resected).67
Treatment
The cornerstone of treatment and the only potential cure for patients with HB is complete resection of the tumor. Although this is a long-known truism, dramatic changes in survival were accomplished only in the last 3 decades. Currently, the 5-year survival rate is 75%79, 80; whereas, 30 years ago, this rate was only 35%.2 A short overview of the different treatment modalities used that resulted in this dramatic increase in survival is presented below.
Combined chemotherapy and surgery
The key to improved therapeutic results was the discovery that HB is highly sensitive to cytostatic and cytotoxic drugs, such as vincristine, doxorubicin, cyclophosphamide, 5-fluorouracil,81 and cisplatin.82 Accordingly, it was found that pretreatment with a combination of cisplatin and doxorubicin (the SIOPEL strategy) improved the prognosis of children with HB, and that strategy has remained the main SIOPEL treatment principle. Some study groups opted for primary resection if possible and only began chemotherapy and second-look surgery if primary surgery was not possible.55, 83 In contrast, the prospective SIOPEL-1 trial was the first study that had the intention of treating all patients with preoperative neoadjuvant chemotherapy. This strategy was based on the expectation that preoperative chemotherapy lead to shrinkage of the tumor, rendering the tumor more solid, less prone to bleeding, and better delineated from the healthy liver parenchyma, thus making complete resection more likely.5 In addition, (micro)metastases, if present, would be treated concurrently. The large-scale, prospective SIOPEL-1 trial confirmed the positive results of earlier, smaller scale studies.
The objective of ongoing trials is to improve the prognosis of the 25% of patients who die as a result of their disease. Examples of these new trials are the comparison of two chemotherapy regimens (cisplatin/vincristine/fluorouracil vs. cisplatin/continuous infusion doxorubicin) by the Children's Oncology Group56; an HB trial in the U.S. (Protocol 9645; cisplatin, vincristine, and fluorouracil vs. carboplatin and cisplatin with or without amifostine; available from URL: http://www.cancer.gov/search/clinical_trials/results_clinicaltrials.aspx); the identification of low-risk and high-risk groups, comparing the treatment of patients who have low-risk HB using cisplatin monotherapy with the treatment of patients who have high-risk HB using intensified cisplatin, carboplatin, and doxorubicin (SIOPEL-2 and SIOPEL-3)47; and the application of megatherapy with carboplatin and etoposide by the German Study Group.28 It will be interesting to see which of these strategies produce a further improvement in outcome.
Multidrug resistance is a major problem in the therapy of patients with advanced and recurrent HB. This has been linked to an increased expression of the multidrug resistance gene 1 (MDR-1) and the concomitant increase of its gene product, P-glycoprotein (P-gp), after each course of chemotherapy.84 P-gp is an ATP dependent membrane channel protein that actively transports drugs out of the cell. An inhibitor of P-gp, the chemosensitizer PSC833, significantly improved the effects of chemotherapy in an HB cell culture model and in animals xenotransplanted with human HB.85 Several inhibitors of P-gp are undergoing late-stage clinical trials, and promising results have been obtained in patients with hematologic malignancies, although results for patients with solid tumors have been negative.86, 87 Nevertheless, development of P-gp inhibitors is ongoing, and clinical trials are being refined, taking into account the complex interactions between the inhibitor and the target cytotoxic drug.88 A clearer picture of the clinical relevance of P-gp inhibitors and their use in the treatment of patients with HB should emerge over the next few years.
Liver transplantation
The first liver transplantation for malignant liver disease was reported in 1968.89 Although transplantation as treatment for patients neoplastic disease appeared more promising in children than in adults, frequent tumor recurrence was a major problem. In view of the shortage of donor organs, at first, transplantation as treatment for malignant disease generally was not accepted. This attitude has changed recently after results of orthotopic liver transplantation (OLT) demonstrated additional value in the treatment of patients with HB.90–93 OLT has become a treatment option for children who have multifocal, bilobar, otherwise unresectable HB without extrahepatic extension of the tumor that responds to chemotherapy.94–96
Other treatment options
Chemoembolization has been promoted as a pretreatment option to render an unresectable HB resectable,97 even when this tumor is located in the caudate lobe.98 These results must be interpreted with caution. First, to our knowledge the majority of reports describe only a single patient or very few patients, and no prospective, randomized results are available. Second, postoperative complications occurred significantly more often after patients underwent major hepatectomy after chemoembolization than after patients who underwent hepatectomy alone.99 Third, and most important, the recently published study of children who were treated with chemoembolization demonstrated that only two of six children with HB survived with no evidence of disease, and one of those patients underwent OLT.100 Similarly, radiotherapy and brachytherapy for patients with HB play a minor role as treatment options.27, 101
The potential side effects of preoperative systemic chemotherapy, such as cardiotoxicity, nephrotoxicity, ototoxicity, and bone marrow depression, warrant the search for other, less toxic modalities. Additional treatment options for patients who do not respond to chemotherapy or who develop drug resistance also are necessary. An attractive strategy that already has been tested in in vivo HCC models may be the suicide gene therapeutic approach.102, 103 The strategy behind this approach is to kill tumor cells selectively by expressing a gene that can convert a membrane-permeable, nontoxic substance (the prodrug) into a toxic agent (the suicide drug) in tumor cells only, thus avoiding the toxic effects of systemic chemotherapy. This can be achieved by targeting the tumor cells with, for example, a replication-deficient adenovirus carrying suicide genes such as E. coli cytosine deaminase or Herpes simplex virus thymidine kinase. In addition to the natural predilection of adenoviruses for hepatocytes, the second level of specificity is formed by using a tumor cell specific promoter/enhancer, like the AFP promoter/enhancer, to express the suicide gene in HB only. The developed in vivo HB models could be used to test this strategy for HB.104–106
Prognosis
The dramatic improvement of the prognosis for children who present with HB in the last 35 years has shifted the attention to improve survival from therapy to the identification and evaluation of risk factors. It has become clear that children with an extrahepatic tumor extension, multifocality, vascular invasion, DNA aneuploidy, and distant metastases have a poor prognosis and, thus, are at high risk.47, 64, 65 Whether intensified chemotherapeutic regimens or a switch to new chemotherapy drugs will improve the prognosis of patients with high-risk disease remains to be seen. Resectable tumor, a decline in circulating AFP levels during chemotherapy,107 and pure fetal histology64, 75 were correlated positively with prognosis and probably characterize the patients with low-risk disease. Trials also are ongoing in which patients with low-risk HB are pretreated using less intensive chemotherapy (SIOPEL-3). These new trials are of particular interest, in that they will evaluate whether late toxic effects can be decreased in patients with low-risk HB and whether the efficacy of treatment can be increased in patients with high-risk HB.
Conclusions
HB is an uncommon liver malignancy that is seen mostly in children younger than 3 years. The dramatic increase in survival that has been observed in the last 3 decades is due mainly to the combination of chemotherapy and surgery. Currently, approximately 75% of children with HB can be cured completely, although large tumor extent, multifocality of the tumor, and metastatic spread are associated with a poor prognosis. Cellular-biologic and molecular-biologic studies are revealing the biologic properties of this embryonal tumor; however, to date, they have not led to the discovery of reliable prognostic factors. The development of new treatment modalities may be the prerequisite for further improvements in the survival of patients with HB.
REFERENCES
- 1. A case of teratoma hepatis. J Pathol Bacteriol. 1898; 5: 128–137.Direct Link:
- 2, , . Liver tumors in children in the particular reference to hepatoblastoma and hepatocellular carcinoma: American Academy of Pediatrics Surgical Section Survey—1974. J Pediatr Surg. 1975; 10: 329–337.
- 3
- 4. Hepatoblastoma in children: a review. J Pediatr Hematol Oncol. 1997; 19: 418–422.
- 5, . Liver tumours. Eur J Cancer. 1999; 35: 953–958; discussion 958–959.
- 6
- 7, , , et al. Malignant hepatic tumours in children: incidence, clinical features and aetiology. Paediatr Perinat Epidemiol. 1990; 4: 276–289.Direct Link:
- 8, , . Primary liver tumors in infancy and childhood. 41 cases variously treated. Arch Surg. 1974; 109: 143–147.
- 9, , . Association between hepatoblastoma and very low birth weight: a trend or a chance [see comments]? J Pediatr. 1997; 130: 557–560. Comment in: J Pediatr. 1997;130:516–517; J Pediatr. 1998;132:750; J Pediatr. 1998;133:585–586.
- 10, . Hepatoblastoma and low birth weight: a trend or chance observation? Med Pediatr Oncol. 2002; 39: 508–509.Direct Link:
- 11, , , . Constitutional interstitial deletion of 11p11 and pericentric inversion of chromosome 9 in a patient with Wiedemann–Beckwith syndrome and hepatoblastoma. Cancer Genet Cytogenet. 1986; 23: 95–104.
- 12, , . Genetics of Beckwith–Wiedemann syndrome-associated tumors: common genetic pathways. Genes Chromosomes Cancer. 2000; 28: 1–13.Direct Link:
- 13, , . Significance of extra copies of chromosome 20 and the long arm of chromosome 2 in hepatoblastoma. Cancer Genet Cytogenet. 1996; 91: 65–67.
- 14, , , , . The first recurring chromosome translocation in hepatoblastoma: der(4)t(1;4)(q12;q34). Genes Chromosomes Cancer. 1997; 19: 291–294.Direct Link:
- 15, , , et al. Cytogenetic characterization of childhood hepatoblastoma. Cancer Genet Cytogenet. 2000; 119: 32–36.
- 16, , , et al. Cytogenetics of hepatoblastoma: further characterization of 1q rearrangements by fluorescence in situ hybridization: an international collaborative study. Med Pediatr Oncol. 2000; 34: 165–170.Direct Link:
- 17, , , . Characterization of genomic alterations in hepatoblastomas. A role for gains on chromosomes 8q and 20 as predictors of poor outcome. Am J Pathol. 2000; 157: 571–578.
- 18, , , . Cytogenetic findings in two new cases of hepatoblastoma. Cancer Genet Cytogenet. 2002; 133: 179–182.
- 19, , , , , . Prognostic impact of molecular genetic alterations in hepatoblastoma. Med Pediatr Oncol. 2002; 38: 104–108.Direct Link:
- 20, , , , . Loss of heterozygosity on chromosome 1 in human hepatoblastoma. Int J Cancer. 1996; 67: 467–471.Direct Link:
- 21, , , , , . Loss of maternal alleles on chromosome arm 11p in hepatoblastoma. Cancer Res. 1994; 54: 5041–5044.
- 22, , , et al. p57(KIP2) is not mutated in hepatoblastoma but shows increased transcriptional activity in a comparative analysis of the three imprinted genes p57(KIP2), IGF2, and H19. Am J Pathol. 2000; 157: 1393–1403.
- 23, , , , . Analysis of the p53 gene in European hepatocellular carcinomas and hepatoblastomas. Oncogene. 1993; 8: 2303–2306.
- 24, , . Absence of p53 gene mutation and infrequent overexpression of p53 protein in hepatoblastoma. J Pathol. 1995; 176: 243–247.Direct Link:
- 25, , , , , . Mutation analysis of p53 gene in childhood malignant solid tumors. J Pediatr Surg. 1997; 32: 1175–1180.
- 26, , , et al. Primary malignant epithelial tumors of the liver in children: a study of DNA content and oncogene expression. Pediatr Dev Pathol. 1998; 1: 270–280.
- 27, , , et al. Is there a place for radiation therapy in the management of hepatoblastomas and hepatocellular carcinomas in children [see comments]? Int J Radiat Oncol Biol Phys. 1992; 23: 525–531. Comment in: Int J Radiat Oncol Biol Phys. 1992;23:675–676.
- 28. Identification of risk groups in hepatoblastoma—another step in optimising therapy [editorial]. Eur J Cancer. 2000; 6: 1343–1346.
- 29, , , , . Hepatic stem-like cells in hepatoblastoma: expression of cytokeratin 7, albumin and oval cell associated antigens detected by OV-1 and OV-6. Histopathology. 1997; 31: 324–329.Direct Link:
- 30, . Hepatic stem cells: a review. Pathology. 2001; 33: 130–141.
- 31, . Stem-like cells in hepatoblastoma. Med Pediatr Oncol. 2002; 39: 504–507.Direct Link:
- 32. Hepatoblastoma with cholangioblastic features (“cholangioblastic hepatoblastoma”) and other liver tumors with bimodal differentiation in young patients. Med Pediatr Oncol. 2002; 39: 487–491.Direct Link:
- 33, , , , . Extramedullary hematopoiesis and intratumoral production of cytokines in childhood hepatoblastoma. Pediatr Res. 1995; 38: 555–563.
- 34, , , , . Production of interleukin-1 beta and interleukin-6 in hepatoblastoma. Int J Cancer. 1993; 53: 728–734.Direct Link:
- 35, , , et al. Human hepatocyte growth factor stimulates the growth of HUH-6 clone 5 human hepatoblastoma cells. Cell Biol Int Rep. 1992; 16: 145–154.
- 36, , , et al. Hepatocyte growth-factor-scatter factor can stimulate post-operative tumor-cell proliferation in childhood hepatoblastoma. Int J Cancer. 2000; 85: 151–159.
- 37, , , . The occurrence of liver growth factor in hepatoblastoma. Eur J Pediatr Surg. 1998; 8: 133–136.
- 38, . Risk of hepatoblastoma in familial adenomatous polyposis. Am J Med Genet. 1992; 43: 1023–1025.Direct Link:
- 39, , , , . Somatic mutations of the APC gene in sporadic hepatoblastomas. Cancer Res. 1996; 56: 3320–3323.
- 40, , , , , . Childhood hepatoblastomas frequently carry a mutated degradation targeting box of the beta-catenin gene. Cancer Res. 1999; 59: 269–273.
- 41, , , , , . Beta-catenin mutations and protein accumulation in all hepatoblastomas examined from B6C3F1 mice treated with anthraquinone or oxazepam. Cancer Res. 2000; 60: 2864–2868.
- 42, , , , . High frequency of beta-catenin mutations in hepatoblastoma. Pediatr Surg Int. 2001; 17: 508–512.
- 43, , , et al. Nuclear localization of beta-catenin is an important prognostic factor in hepatoblastoma. J Pathol. 2001; 193: 483–490.Direct Link:
- 44, , , et al. Frequent deletions and mutations of the beta-catenin gene are associated with overexpression of cyclin D1 and fibronectin and poorly differentiated histology in childhood hepatoblastoma. Clin Cancer Res. 2001; 7: 901–908.
- 45, , , et al. Hepatocyte growth factor induces Wnt-independent nuclear translocation of beta-catenin after Met-beta-catenin dissociation in hepatocytes. Cancer Res. 2002; 62: 2064–2071.
- 46, , , , , . Epidemiological and familial aspects of hepatoblastoma. Med Pediatr Oncol. 1990; 18: 103–109.Direct Link:
- 47, , , et al. Hepatoblastoma presenting with lung metastases. Cancer. 2000; 89: 1845–1853.Direct Link:
- 48
- 49, , . Neonatal hepatoblastoma metastatic to the choroid and iris. Arch Ophthalmol. 1996; 114: 757–761.
- 50, , , et al. Interleukin-6 stimulates thrombopoiesis through thrombopoietin: role in inflammatory thrombocytosis. Blood. 2001; 98: 2720–2725.
- 51, , . Hepatoblastoma. A clinical and pathologic study of 54 cases. Am J Surg Pathol. 1982; 6: 693–705.
- 52, . Ultrasound of primary hepatic tumours in childhood. Pediatr Radiol. 1988; 19: 19–24.
- 53, . Periportal low attenuation at CT in childhood. Radiology. 1992; 183: 685–688.
- 54, , , . Potentially misleading bone scan findings in patients with hepatoblastoma. Clin Nucl Med. 1993; 18: 1026–1031.
- 55, , . Is laparatomy the first step in treatment of childhood liver tumors? The experience from the German Cooperative Pediatric Liver Tumor Study HB-89. Eur J Pediatr Surg. 1994; 4: 82–86.
- 56, , , et al. Randomized comparison of cisplatin/vincristine/fluorouracil and cisplatin/continuous infusion doxorubicin for treatment of pediatric hepatoblastoma: A report from the Children's Cancer Group and the Pediatric Oncology Group. J Clin Oncol. 2000; 18: 2665–2675.
- 57, , . Review articles: liver biopsy. N Engl J Med. 2001; 344: 495–500.
- 58, , , , . Hepatic rupture after puncture biopsy: value of embolization. Apropos of a case with hepatoblastoma. Pediatrie. 1988; 43: 531–533.
- 59, . A common language for childhood liver tumours. Pediatr Surg Int. 1992; 7: 325–326.
- 60SobinLH, WittekindC, editors. TNM classification of malignant tumors, 5th edition. New York: John Wiley & Sons, 1997.
- 61, , , et al. The proposed Japanese TNM classification of primary liver carcinoma in infants and children. Jpn J Clin Oncol. 1983; 13: 361–369.
- 62, , , et al. Effective treatment of unresectable or metastatic hepatoblastoma with cisplatin and continuous infusion doxorubicin chemotherapy: a report from the Children's Cancer Study Group. J Clin Oncol. 1991; 9: 2167–2176.
- 63, , , , . Cisplatin, vincristine, and fluorouracil therapy for hepatoblastoma: a Pediatric Oncology Group study. J Clin Oncol. 1993; 11: 96–99.
- 64, , , . Prognostic factors and staging systems in childhood hepatoblastoma. Int J Cancer. 1997; 74: 593–599.Direct Link:
- 65, , , et al. Pretreatment prognostic factors for children with hepatoblastoma—results from the International Society of Paediatric Oncology (SIOP) study SIOPEL 1. Eur J Cancer. 2000; 36: 1418–1425.
- 66
- 67. Hepatoblastoma: assessment of criteria for histologic classification. Med Pediatr Oncol. 2002; 39: 478–483.Direct Link:
- 68, . Hepatoblastoma and hepatocarcinoma in infancy and childhood. Report of 47 cases. Cancer. 1967; 20: 396–422.Direct Link:
- 69, . Histologic classification of liver-cell carcinoma in infancy and childhood and its clinical evaluation. A study of 70 cases collected in Japan. Cancer. 1970; 25: 551–563.Direct Link:
- 70, . Hepatoblastoma: an analysis of the relationship between morphologic subtypes and prognosis. Am J Pediatr Hematol Oncol. 1988; 10: 301–307.
- 71, , . Hepatoblastoma. Attempt at characterization of histologic subtypes. Am J Surg Pathol. 1982; 6: 599–612.
- 72
- 73, , , et al. Immunohistochemical evaluation of hepatoblastomas with use of the hepatocyte-specific marker, hepatocyte paraffin 1, and the polyclonal anti-carcinoembryonic antigen. Mod Pathol. 1998; 11: 934–938.
- 74, , , , . Successful transplantation of human hepatoblastoma into immunodeficient mice. J Pediatr Surg. 1996; 31: 1241–1246.
- 75, , , et al. Histopathology and prognosis in childhood hepatoblastoma and hepatocarcinoma. Cancer. 1989; 64: 1082–1095.Direct Link:
- 76, , , , . Hepatoblastoma: the prognostic significance of histologic type. Pediatr Pathol. 1992; 12: 167–183.
- 77, , , et al. Clinico-pathological criteria with prognostic relevance in hepatoblastoma. Eur J Cancer. 1994; 8: 1052–1058.
- 78, , . Small cell undifferentiated histology in hepatoblastoma may be unfavorable. Cancer. 2001; 92: 3130–134.Direct Link:
- 79, , , et al. Cisplatin, doxorubicin and delayed surgery for childhood hepatoblastoma a successful approach. Results of the first prospective study of the International Society of Pediatric Oncology (SIOP)—SIOPEL 1. J Clin Oncol. 2000; 18: 3819–3828.
- 80, , , et al. Surgical view of the treatment of patients with hepatoblastoma. Results from the first prospective trial of the International Society of Pediatric Oncology Liver Tumor Study Group (SIOPEL-1). Cancer. 2002; 94: 1111–1120.
- 81, , , , , . Combination chemotherapy (vincristine, Adriamycin, cyclophosphamide, and 5-fluorouracil) in the treatment of children with malignant hepatoma. Cancer. 1982; 50: 821–826.Direct Link:
- 82, , , , , . Effective cisplatin (DDP) based chemotherapy in the treatment of hepatoblastoma. Med Pediatr Oncol. 1985; 13: 187–190.Direct Link:
- 83. Conversion of unresectable to resectable hepatoblastoma and long-term follow-up study. World J Surg. 1995; 19: 814–816.
- 84, , , , , . The role of the MDR1 gene in the development of multidrug resistance in human hepatoblastoma: clinical course and in vivo model. Cancer. 2002; 95: 1795–1801.Direct Link:
- 85, , , , . MDR1 modulators improve the chemotherapy response of human hepatoblastoma to doxorubicin in vitro. J Pediatr Surg. 2002; 37: 1579–1584.
- 86, , , et al. Treatment of poor prognosis AML patients using PSC833 (valspodar) plus mitoxantrone, etoposide, and cytarabine (PSC-MEC). Adv Exp Med Biol. 1999; 457: 47–56.
- 87. Multidrug resistance: clinical relevance in solid tumours and strategies for circumvention. Curr Opin Oncol. 1998; 10 (Suppl 1): S15–S19.
- 88
- 89, , , et al. Extended survival in 3 cases of orthotopic homotransplantation of the human liver. Surgery. 1968; 63: 549–563.
- 90, , , et al. Liver transplantation for hepatoblastoma. The American experience. Ann Surg. 1991; 213: 118–121.
- 91, , , et al. Unresectable hepatic tumors in childhood and the role of liver transplantation. J Pediatr Surg. 1996; 31: 1563–1567.
- 92, , , et al. Preoperative chemotherapy, major liver resection, and transplantation for primary malignancies in children. Transplant Proc. 1996; 28: 2393–2394.
- 93, , , , , , et al. Orthotopic liver transplantation for unresectable hepatoblastoma: a single center's experience. J Pediatr Surg. 1999; 34: 1261–1264.
- 94, , , . Indications and role of liver transplantation for malignant tumors. Oncologist. 1997; 2: 164–170.
- 95, , , et al. Liver transplantation and chemotherapy for hepatoblastoma and hepatocellular cancer in childhood and adolescence. J Pediatr. 2000; 136: 795–804.
- 96, , , et al. Liver transplantation for hepatoblastoma. Results from the International Society of Pediatric Oncology (SIOP) Study SIOPEL-1 and review of the world experience. Med Pediatr Oncol. In press.
- 97, , , et al. Effectiveness of preoperative transarterial chemoembolization in presumed inoperable hepatoblastoma. J Vasc Interv Radiol. 1999; 10: 1275–1280.
- 98, , , et al. Resection after intraarterial chemotherapy of a hepatoblastoma originating in the caudate lobe. Surgery. 1990; 107: 231–235.
- 99, , , , , . Frequency, technical aspects, results, and indications of major hepatectomy after prolonged intra-arterial hepatic chemotherapy for initially unresectable hepatic tumors. J Am Coll Surg. 1995; 180: 213–219.
- 100, , , . Feasibility and toxicity of chemoembolization for children with liver tumors. J Clin Oncol. 2000; 18: 1279–1284.
- 101, , , , . A 10-year experience of pediatric brachytherapy. Int J Radiat Oncol Biol Phys. 1995; 32: 451–455.
- 102, , , et al. In vivo gene therapy for alpha-fetoprotein-producing hepatocellular carcinoma by adenovirus-mediated transfer of cytosine deaminase gene. Cancer Res. 1997; 57: 461–465.
- 103, , , et al. Gene therapy for hepatocellular carcinoma based on tumour-selective suicide gene expression using the alpha-fetoprotein (AFP) enhancer and a housekeeping gene promoter. Eur J Cancer. 2001; 37: 140–147.
- 104, , , , , . Characterization of the continuous cell line HepT1 derived from a human hepatoblastoma. Lab Invest. 1996; 74: 809–818.
- 105, , , , . Comparative activity of cisplatin, ifosfamide, doxorubicin, carboplatin, and etoposide in heterotransplanted hepatoblastoma. Cancer. 1998; 83: 2400–2407.Direct Link:
- 106, , , et al. Ein Tiermodell für das humane Hepatoblastom. In: SiewertJR, editor. Chirurgisches Forum 2002 für experimentelle und klinische Forschung, Volume 31, 1st edition. Berlin: Springer-Verlag, 2002: 543–537.
- 107, , , et al. Timing and magnitude of decline in alpha-fetoprotein levels in treated children with unresectable or metastatic hepatoblastoma are predictors of outcome: a report from the Children's Cancer Group. J Clin Oncol. 1997; 15: 1190–1197.

1097-0142/asset/olbannerleft.gif?v=1&s=ca681f5719430b26e1bc15e9ea4c9fc0a7110104)
1097-0142/asset/olbannerright.gif?v=1&s=8142566facf7e76aef9be6c51162a2e920b3b9f9)
1097-0142/asset/cover.gif?v=1&s=a7299bc18f075294c232ade468773cd0672bd470)
