Primary liver tumors in the pediatric population are rare entities. Based on UK children’s cancer registry data (0–16 years old), in the period 2006 to 2008, a total of 55 cases of hepatic tumors (hepatoblastoma and hepatocellular carcinoma) were registered, equating to an annual incidence of approximately 2.0 per million children. The Surveillance, Epidemiology, and End Results (SEER) data from the United States describe the estimated incidence of primary hepatic tumors in children to about 1.6 per million children. Overall, primary hepatic tumors in children account for only 1% of all pediatric malignancies.
Historically, primary hepatic tumors have been regarded as epithelial or mesenchymal in origin, encompassing both malignant and benign tumors ( Table 38.1 ). Refinements to this basic classification of hepatic tumors, specifically with primary malignant epithelial tumors of the liver, have allowed the development of risk stratification and prognostic indicators, along with improvements in treatment strategies.
|Type of Tumor||Malignant||Benign|
|Hepatocellular carcinoma||Focal nodular hyperplasia|
|Undifferentiated sarcoma of the liver||Hemangioma|
In the pediatric setting, the combination of chemotherapy and/or surgery, including transplantation, remains the fundamental tenet of the radical approach to the treatment of primary liver tumors. Strategies using interventional radiology approaches, such as transarterial chemoembolization and radiofrequency ablation, are well established in the management of liver tumors (primary and secondary tumors) in the adult oncology field. The true benefit of such approaches in the pediatric oncology domain is yet to be determined.
Epidemiology and Genetics
Hepatoblastoma accounts for approximately 80% of all primary liver tumors in children, with the peak incidence within the first 4 years of life and a median age of presentation of 18 months. The data describe an almost doubling of the incidence of hepatoblastoma over the periods 1973 to 1977 to 1993 to 1997 (0.61–1.18/million, respectively), with a subsequent annual percentage increase in incidence of hepatoblastoma of 4%. The cause of hepatoblastoma, like most pediatric malignancies, is unknown, and most tumors will arise de novo ; however, there are a number of preexisiting underlying medical conditions/syndromes associated with an increased risk of developing hepatoblastoma ( Table 38.2 ).
|Condition With Predisposition to Developing Hepatoblastoma||Reference|
|Familial adenomatous polyposis|
|Trisomy 18 (Edward syndrome)|
A number of studies across industrialized nations (United Kingdom, United States, China, Japan, Scandinavian countries) have confirmed that there is a clear increased risk of hepatoblastoma with very low birth weight (VLBW), most commonly defined as a birth weight of less than 1500 g.
The exact cause of this increased risk of hepatoblastoma with VLBW remains unclear; however, the possible effects of other perinatal and environmental factors, such as preeclampsia, high maternal pregnancy weight, hydramnios (oligo- and poly-), assisted conception, parental tobacco use, parental occupational exposure to petroleum, metals, and paint on the risk of developing hepatoblastoma have been postulated.
Survival from hepatoblastoma has seen dramatic improvements over the past four decades, from around 35% in the 1970s to current survival figures from 50% to 100%, depending on the disease characteristics. Performing a complete resection is absolutely critical to achieving a cure in hepatoblastoma, but approximately 60% of tumors are deemed unresectable at presentation.
For the staging of hepatoblastoma, the system originally developed by the International Society of Paediatric Oncology Epithelial Liver tumor study group (SIOPEL) and validated in their first international trial, SIOPEL 1 as prognostically valuable, PRETEXT (pretreatment extent of disease) has been universally adopted as the standard radiological approach to describing the extent of the tumor at diagnosis ( Fig. 38.1 ). This includes the addition of a number of annotation factors that describe venous-portal involvement of the tumor (V, P), involvement of contiguous extrahepatic organs, and the presence of distant metastases (E, M). Revisions to these annotation factors recognize the worse prognosis associated with multifocal tumors (F) or tumor rupture (R), along with involvement of the caudate lobe (C) and local lymph nodes (N); this is now reflected in the annotation factors used following recommendations by the Childrens Hepatic tumors International Collaboration (CHIC).
Hepatoblastoma is frequently associated with elevated serum levels of alpha-fetoprotein (AFP) and currently represents the only tumor marker that can be used to monitor response to treatment and possible recurrence in the disease surveillance follow-up phase. Both the Children’s Oncology Group (COG) and ICLTS group have affirmed that a low AFP level (< 100 ng/mL) at diagnosis is associated with a poorer prognosis; this has often been related to particular histological subtypes, in particular the small cell undifferentiated subtype. Taking into consideration the relative AFP levels in early infancy and the prognostic implications of low AFP levels, the CHIC analysis has led to the recommendation of three ranges of AFP to be taken into account when considering risk stratification (< 100, 101–1000, and > 1000 ng/mL). The only other additional factor that has been included in this new risk stratification is age.
Complete surgical resection is fundamental in achieving a cure in hepatoblastoma. The current view is that in all but a few very select cases, neoadjuvant chemotherapy is applied to improve the options for complete resection. Assessment of resectability takes the PRETEXT system and applies this in the context of having received neoadjuvant chemotherapy, referred to as POSTTEXT (post-treatment extent of disease). Surgery for hepatoblastoma should be driven by a specialized hepatobiliary surgical team that includes expertise for performing liver transplantation; it should be undertaken using the Brisbane 2000 nomenclature system to describe the planned resection. Nonanatomical resections are associated with an increased risk of post-operative residual tumor and inferior prognosis and, as such, are not recommended surgical strategies.
Treatment of refractory or relapsed disease remains a significant challenge. For patients who are anthracycline-naïve at the time of relapse or progression, the addition of an anthracycline and further surgery may provide salvage options for about one-third of patients. A number of alternative chemotherapeutic agents have been investigated, including etoposide, ifosfamide, oxaliplatin, and irinotecan. Irinotecan has been shown to have some activity in hepatoblastoma and is being further investigated in the current international Pediatric Hepatic Malignancy International Therapeutic Trial (PHITT) in the context of high-risk hepatoblastoma and persisting metastatic disease.
Liver transplantation has become an increasingly used option for patients with extensive disease (with or without metastases) as our technical ability to undertake these procedures and improvements in post-transplantation care have evolved ( Fig. 38.2 ). The superior outcomes for individuals undergoing liver transplantation as the primary surgical intervention, as compared with liver transplantation in the setting of relapsed or recurrent disease following previous local resection, the so-called rescue transplantation (82% vs. 30% survival at 5 years), might suggest that we should be more liberal in selecting patients for primary liver transplantation as the primary surgical option for hepatoblastoma. The presence of distant metastases (usually lung) at diagnosis is not an absolute contraindication for consideration of liver transplantation provided that clearance of metastases can be achieved with neoadjuvant chemotherapy and/or with surgical resection. Currently, liver transplantation should be considered as the primary surgical solution for patients with multifocal (F +) POSTTEXT IV tumors. Multifocal PRETEXT IV without distant metastasis should be an indication for liver transplantation. In the current PHITT trial, it is recommended to make a referral for transplantation after limited chemotherapy. When considering large, solitary, POSTTEXT IV tumors, preservation of adequate normal liver parenchyma needs to be considered in defining the most appropriate surgical approach. POSTTEXT III tumors that demonstrate vascular or portal involvement (V +, P +) should be considered for liver transplantation, although this latter group of patients may be regarded by some as suitable for extreme conventional resection. Progressive or persistent extrahepatic disease and/or lung metastases that cannot be cleared with neoadjuvant chemotherapy and/or surgery is a contraindication for liver transplantation.
Other Treatment Considerations
Notwithstanding the clear role of transplantation in the management of hepatoblastoma, there remain a number of unanswered questions, such as those considered in the following sections.
Prolonged Chemotherapy Pre-transplantation
Prolonged chemotherapy administration, which attempts to render tumors resectable by conventional resection, should be avoided. Several studies indicate that continuing to administer chemotherapy after four cycles does not increase the likelihood of conventional resectability of the tumour. In addition, the prolonged administration can lead to drug resistance. The disappearance of nodule(s) from one sector of liver following chemotherapy should not change the decision for transplantation if all sectors were involved (multifocal tumor-PRETEXT IV) prechemotherapy because there is a high risk of persistent microscopic viable tumor cells, despite what appears as radiological clearance. Current practice in established centers is to review resectability after two courses and plan liver transplantation after four cycles of chemotherapy.
Metastectomy in Patients With Disseminated Disease
There is agreement regarding exclusion of extrahepatic disease in children considered for transplant. Although chemotherapy is very successful in the eradication of lung metastases in hepatoblastoma patients, a 90% response rate was observed in the recent SIOPEL 4 study. If a child is to be considered for transplantation, pulmonary metastases that persist following chemotherapy should be excised (not too small to find at surgery) before liver transplantation to provide unequivocal evidence that there is no viable tumor. Similarly, if there is any extrahepatic disease, it has to be completely excised before considering for transplantation.
Another challenge faced by physicians dealing with transplantation for hepatoblastoma is the combined toxicity of immunosuppressive drugs for antirejection and chemotherapy resulting in increased risk of complications, especially sepsis. Sepsis can be a major factor responsible for increased mortality post-liver transplantation in patients with hepatoblastoma. Therefore, early minimization of antirejection medications is a reasonable strategy. A cautious approach is still necessary because under conditions of identical immunosuppression, the incidence of rejection was similar when children who underwent liver transplantation for hepatoblastoma were compared with children who underwent transplantation for nonmalignant indications.
Chemotherapy in the Post-transplantation Period
Currently, there is no evidence base indicating the benefit of further chemotherapy in the post-transplantation period, especially if there is no extrahepatic disease pre-transplantation. Several studies have demonstrated equally good results regardless of whether or not subsequent chemotherapy was given after transplantation. Kaliciński et al. compared 65 patients who received chemotherapy after orthotopic liver transplantation (OLT) for hepatoblastoma with 82 patients who did not. There were no significant differences in 5-year survival rates (77% vs. 70%) between the two groups. Many transplantation centers still opt to provide chemotherapy post-transplantation, but we choose not to do this unless there was doubt regarding the extrahepatic disease at the time of listing for transplantation or unless dictated by clinical trial protocol.
Use of Living Versus Cadaveric Donors
The use of living versus cadaveric donors for liver transplantation is dependent on local practices and the availability of both types of donors. The benefits of living donor liver transplantation include control of the timing of transplantation, resulting in potentially shorter wait times and thereby eliminating dependence on cadaveric liver availability—but engender risk of a major operative procedure to a healthy donor. Benefits to cadaveric donor transplantation include longer blood vessels from cadaveric donor grafts, which allow for easier vascular reconstruction.
Recurrence in Post-transplant Transplantation
Cruz et al. published the US experience of transplantation in hepatoblastoma between 1988 and 2010. Recurrence of hepatoblastoma occurred in 53 children (15.9%). Recurrences were more frequent in children with preliver transplantation lung metastases compared with children with hepatic disease only (50% vs. 7.4%; P = .016). Recurrences were more frequent after segmental cadaveric or living donor grafts compared with cadaveric whole liver grafts.
Another interesting finding was the relationship of tumor necrosis following chemotherapy and recurrence—18 patients who had more than 50% tumor necrosis after chemotherapy were disease-free in contrast to one-third of children with less than 50% necrosis who developed recurrent hepatoblastoma ( P = .013; not significant). Management of these recurrences varied depending on the site of recurrence and response to chemotherapy. Recently, it was reported that with cisplatin-based chemotherapy, the presence of a microscope-positive resection margin (micro-PRM) does not influence the outcome, even without additional local treatment. Although complete resection remains the aim, micro-PRM does not necessitate mandatory second-look surgery. A wait-and-see policy was recommended if post-operative chemotherapy is administered and AFP levels and imaging become normal.
Management of Centrally Located Hepatoblastomas
Practice varies among centers, depending on factors such as surgical expertise and availability of a transplantation option. There cannot be fixed rules, so these cases should be considered individually by a multidisciplinary team to determine the best surgical option. Although some authors claim that aggressive resection is possible, others do not recommend extensive liver resection.
Replacement of Inferior Vena Cava
This practice varies, especially in centers where cadaveric vessels are available. Some authors recommend that the retrohepatic vena cava should be removed en bloc with the liver during the transplantation. The cava can be reconstructed using a donor iliac vein (deceased donor) or donor jugular vein (living related transplantation). Others preserve the native retrohepatic vena cava in selected transplant recipients, provided that there is no evidence of direct tumor involvement.
Interval Between Chemotherapy and Transplantation
Héry et al. reviewed their experience with liver transplantation for hepatoblastoma and reported a wait time of 1 to 50 days for liver transplantation (median, 16 days). The authors concluded that the delay in timing from the last chemotherapy to transplantation should be kept as short as possible.
Therefore, most established transplant programs (e.g., United Kingdom, United States) give priority to children listed for transplantation. The concept is to avoid extra chemotherapy to optimize the chances of transplantation in the window between cycles of chemotherapy.
Liver Autotransplantation for Hepatoblastoma
Liu et al. describe one case. This type of surgery may be considered for patients with unresectable hepatoblastoma, POSTTEXT stage III or IV, or those with tumor in the major vessels when no liver donors are available, because the technique is reproducible only by experienced pediatric liver transplantation surgeons.
Epidemiology and Genetics
Hepatocellular carcinoma (HCC) represents the second most frequently occurring primary malignant liver tumor seen in childhood and adolescence. The incidence varies, depending on geographic location, and is explained by some of the causative factors for HCC in the pediatric population. UK-based registry data has reported an incidence of HCC as low as 0.09/million; however, this seems to be much lower than other series reported in the literature. The SEER data indicate that HCC represents 0.3% of all cancers seen in childhood and, over the period 1973 to 1997, had an estimated incidence of 0.4 per million. The incidence has remained reasonably stable over the past 40 years; this is in sharp contrast to that of hepatoblastoma, which has doubled over the same time frame. HCC is generally regarded as a tumor of older children and adolescents and is exceptionally rare in those younger than 5 years, although the incidence rises to 0.8 per million in the 15- to 19-year age group. There is a male preponderance reported in most published studies. The incidence in population-based studies in the Far East has indicated an overall higher incidence than that noted earlier, with an overall incidence of 0.67 per 100,000 population. This has fallen to an overall incidence of 0.38/100,000 population following the introduction of a hepatitis B (HBV) vaccination program. HBV vaccination has had a bigger impact on the incidence in males compared with females.
When considering the causes and risk factors for the development of HCC, the presence or absence of cirrhosis is of importance. However, a minority of cases of HCC in the pediatric and adolescent population have a background of cirrhotic liver disease as compared with 70% to 90% of HCC in adults. In the absence of cirrhosis, HCC can arise de novo or with pre-existing noncirrhotic liver disease. HCC arising de novo accounts for up to 50% of cases of HCC in the pediatric and adolescent age groups in areas where hepatitis B is not endemic.
The impact on the incidence of HCC following the implementation of an HBV vaccination program in certain areas worldwide, described earlier, gives a clear indication of the role of HBV in the development of HCC. In the pediatric population, this is very much a result of perinatally acquired HBV. There is a 100-fold greater incidence of HCC in HBV chronic carriers in comparison to noncarriers and, in select groups, clear evidence of HBV infection can be found in two-thirds to 100% of cases of HCC in children. Despite this clear association, it remains unclear what the malignant transforming event is in the context of HBV infection.
The role of hepatitis C (HCV) infection is also a well-described risk factor for the development of HCC in adults but is exceptionally rare in children. HCV infection is estimated to affect between 0.2% and 0.4% of children and is principally acquired via vertical transmission. HCC is associated with a number of inherited metabolic disorders that lead to chronic liver disease—not all necessarily resulting in cirrhosis—and conditions characterized by congenital cholestasis.
The international consensus on the classification of pediatric liver tumors describes three groups of HCC:
Fibrolamellar HCC (FL-HCC)
Hepatocellular neoplasm NOS (not otherwise specified)
It has been recognized that interpreting the histopathological features of HCC can be challenging because a proportion of these tumors arise on the background of pre-existing hepatic pathology. However, pediatric HCC cases have a very similar appearances to those seen in adult HCC cases (classic HCC). FL-HCC is regarded as a very distinct entity, typically presenting in teenagers and young adults and always arising on the background of a normal liver. The female preponderance of FL-HCC seen in the adult population is not sustained in the pediatric population, with no significant difference in the rates between males and females. There was an initial suggestion that FL-HCC had a better overall prognosis compared with classic HCC, but this has not been found to be true in other studies. Typically, FL-HCC is not associated with elevated AFP levels. Late relapses with FL-HCC are more frequently described. A very small proportion of HCC cases also demonstrate component features that are very similar to those of hepatoblastoma, and these have traditionally been referred to as transitional cell liver tumors . However, Lopez-Terrada et al. proposed that these tumors to be termed hepatocellular neoplasm NOS.
Because of the rarity of pediatric HCC, this type of tumor has been included in all the respective collaborative group trials listed above for hepatoblastoma as opposed to separate, dedicated HCC trials. The improvement in survival for hepatoblastoma has not been mirrored in HCC. The greatest opportunity to achieve a long-term cure is based on the ability to obtain a complete surgical resection, including the option of liver transplantation. Staging and attempts at risk stratification have followed exactly the same route as those for hepatoblastoma, discussed earlier. The PRETEXT system described by the SIOPEL group is used to describe the potential resectability of HCC ( Fig. 38.3 ), taking into account the annotation factors described with hepatoblastoma (VPEFRCNM). There is agreement from collaborative groups and their respective studies that the minority of tumors are resectable at the time of diagnosis and there is evidence that chemoresponsiveness is higher in pediatric HCC compared with adult HCC cases. However, all the collaborative groups have used the same chemotherapy backbone for HCC as used for hepatoblastoma, but with disappointing results. Although studies have reported responses to chemotherapy of about 40% to 50%, this is not reflected in an increase in the proportion of tumors becoming resectable because most HCC cases present as advanced disease (PRETEXT III and IV). Given the relative lack of efficacy demonstrated with traditional cisplatin-based chemotherapy regimens, the search for alternative options has continued. Based on data from large studies in adults with HCC, the combination of gemcitabine and oxaliplatin has demonstrated responses and is actively being investigated in the current PHITT study (EudraCT number, 2016-002828-85).