Liver transplantation for non–hepatocellular carcinoma malignancy: Indications, limitations, and analysis of the current literature


  • Eric J. Grossman,

    1. Section of Transplantation, Department of Surgery, University of Chicago Medical Center, Chicago, IL
    Search for more papers by this author
  • J. Michael Millis

    Corresponding author
    1. Section of Transplantation, Department of Surgery, University of Chicago Medical Center, Chicago, IL
    • Section of Transplantation, Department of Surgery, University of Chicago Medical Center, 5841 South Maryland Avenue, Chicago IL, 60637
    Search for more papers by this author
    • Telephone: 773-702-6319; FAX: 773-702-7511


Orthotopic liver transplantation (OLT) is currently incorporated into the treatment regimens for specific nonhepatocellular malignancies. For patients suffering from early-stage, unresectable hilar cholangiocarcinoma (CCA), OLT preceded by neoadjuvant radiotherapy has the potential to readily achieve a tumor-free margin, accomplish a radical resection, and treat underlying primary sclerosing cholangitis when present. In highly selected stage I and II patients with CCA, the 5-year survival rate is 80%. As additional data are accrued, OLT with neoadjuvant chemoradiation may become a viable alternative to resection for patients with localized, node-negative hilar CCA. Hepatic involvement from neuroendocrine tumors can be treated with OLT when metastases are unresectable or for palliation of medically uncontrollable symptoms. Five-year survival rates as high as 90% have been reported, and the Ki67 labeling index can be used to predict outcomes after OLT. Hepatic epithelioid hemangioendothelioma is a rare tumor of vascular origin. The data from single-institution series are limited, but compiled reviews have reported 1- and 10-year survival rates of 96% and 72%, respectively. Hepatoblastoma is the most common primary hepatic malignancy in children. There exist subtle differences in the timing of chemotherapy between US and European centers; however, the long-term survival rate after transplantation ranges from 66% to 77%. Fibrolamellar hepatocellular carcinoma is a distinct liver malignancy best treated by surgical resection. However, there is an increasing amount of data supporting OLT when resection is contraindicated. In the treatment of either primary or metastatic hepatic sarcomas, unacceptable survival and recurrence rates currently prohibit the use of OLT. Liver Transpl 16:930-942, 2010. © 2010 AASLD.

The role of orthotopic liver transplantation (OLT) in the therapy of hepatic malignancies has evolved dramatically over the last 2 decades. Despite the limited success of early trials involving OLT for patients with primary liver malignancies,1 novel chemotherapy regimens combined with appropriate patient selection have led to the widespread acceptance of OLT as an effective treatment for hepatocellular carcinoma (HCC). Presently, the indications for OLT in patients with HCC are expanding,2 and liver transplantation is becoming an essential treatment component for less common hepatic tumors. This review focuses on the indications and limitations of liver transplantation and immunosuppression in the treatment of cholangiocarcinoma (CCA), hepatic metastases from neuroendocrine tumors (NETs), hepatoblastoma (HB), hepatic epithelioid hemangioendothelioma (HEHE), and sarcomas.


CCA is the second most common primary hepatobiliary malignancy in the United States.3 Worldwide, it accounts for 3% of all primary gastrointestinal malignancies and 10% of primary hepatobiliary malignancies.4 The mortality rate for untreated CCA ranges from 50% to 70% within 12 months5; therefore, aggressive treatment is warranted. Surgical resection is the mainstay of treatment for periductal CCA and yields 5-year survival rates of 27% to 44%.6, 7 However, tumor invasion into the main portal vein, common hepatic artery, or one lobe of the liver with invasion of the contralateral branch of the portal vein or hepatic artery renders the tumor unresectable.8, 9 Additionally, CCA arising in the setting of primary sclerosing cholangitis (PSC; Fig. 1) is associated with a likelihood of multifocal disease and a high risk of recurrence following resection; thus, such patients are prohibited from undergoing resection.7, 10 OLT, however, has the potential to readily achieve a tumor-free margin, accomplish a radical resection, and treat underlying PSC when present; this makes OLT an attractive treatment option.5

Figure 1.

Explanted liver with PSC.

Although the exact carcinogenic mechanism of CCA is not known, current evidence suggests that hypermethylation of specific gene promoters is required for the transformation, promotion, and progression of CCA.11 The multitude of genes methylated in CCA and their potential roles in cholangiocarcinogenesis were recently reviewed by Sandhu and colleagues.12 Interestingly, hypermethylation and inactivation of specific cell cycle inhibitors are present in over 80% of CCAs.13, 14 Additionally, methylation of the tumor suppressor genes Semaphorin 3B, p16INK4A, and Ras,14–16 the DNA repair gene human mutL homologue-1,14 and the proinflammatory gene interleukin-617 has been implicated in the progression of CCA. In addition to DNA methylation, histone acetylation and deacetylation may also promote tumor growth. Despite promising in vitro evidence showing that inhibiting histone deacetylase retards CCA proliferation,18 the specific role of histone modification in the carcinogenesis of CCA remains an area of ongoing research.

Nonetheless, after nearly 20 years of surgical trials and clinical research, OLT preceded by neoadjuvant radiotherapy has emerged as the standard of care for patients suffering from early-stage, unresectable hilar CCA. This section reviews the milestone achievements and publications supporting OLT and outlines the current areas of investigation.


5-FU, 5-fluorouracil; CA19-19, carbohydrate antigen 19-9; CCA, cholangiocarcinoma; FLHCC, fibrolamellar hepatocellular carcinoma; HB, hepatoblastoma; HCC, hepatocellular carcinoma; HEHE, hepatic epithelioid hemangioendothelioma; MVTx, multivisceral transplantation; NET, neuroendocrine tumor; OLT, orthotopic liver transplantation; OPTN, Organ Procurement and Transplantation Network; PSC, primary sclerosing cholangitis; SIOPEL, Childhood Liver Tumour Strategy Group of the International Society of Paediatric Oncology; UNOS, United Network for Organ Sharing; VEGF, vascular endothelial growth factor.

Early Experiences

The evolution of this treatment modality and the improvements in survival after OLT for CCA are well documented in a recent review by Castaldo and Wright Pinson.19 In the late 1980s and early 1990s, the average 1-year survival rate was 30%, whereas the 5-year survival rate was a dismal 5% to 10%.19 In 1991, the largest series at that time revealed 2- and 5-year survival rates of 30% and 17%, respectively.20 Assuming that these disappointing results were due to local recurrence and residual disease persisting after transplantation, researchers at the University of Pittsburgh pioneered a radical approach using cluster abdominal transplantation. Unfortunately, this technique also yielded an equally poor 3-year survival rate of 20% and a recurrence rate of 57%.21

In 2000, the Cincinnati registry examined the results of 207 patients who underwent liver transplantation for CCA. The overall 1-, 2-, and 5-year survival rates were 72%, 48%, and 23%, respectively. Among those patients who suffered recurrence, the recurrence was detected within 2 years for 84%; recurrence occurred in the liver allograft for 47% and in the lungs for 30%.22 Similarly, in 2004, Spanish liver transplant centers reported a 30% 3-year survival rate for 36 patients.23 Overall, the early results for OLT were poor, and many surgeons and oncologists questioned the future of OLT in the treatment of CCA.

Adjuvant and Neoadjuvant Therapy

With the number of published trials increasing, the aggregate data indicated superior outcomes for patients who underwent OLT and were discovered to have negative margins and an absence of regional lymph node metastases.24, 25 In order to extend these observations, the University of Nebraska and the Mayo Clinic instituted a protocol of strict patient selection as well as the addition of neoadjuvant or adjuvant radiation and chemotherapy, and they reported drastically improved results.

In 2000, De Vreede et al.26 demonstrated an 80% 5-year survival rate in highly selected stage I and II patients with CCA. Patients in this series underwent external beam irradiation and received bolus 5-fluorouracil (5-FU); this was followed by brachytherapy with iridium plus a 5-FU infusion, exploratory laparotomy, and, finally, a continuous 5-FU infusion until transplantation. More recently, in 2005, the Mayo group reported a protocol using neoadjuvant external beam radiation therapy, chemosensitization with 5-FU and oral capecitabine, and intraluminal brachytherapy.27 Notably, only patients with unresectable early-stage (<3-cm) CCA tumors and patients with concomitant PSC were enrolled. Regional lymph node metastases, peritoneal metastases, and locally extensive disease were all criteria for exclusion. This multimodality protocol improved survival to 92% at 1 year, 82% at 3 years, and 82% at 5 years. Additionally, 1-, 3-, and 5-year recurrence rates were also improved to 0%, 5%, and 12%, respectively.27 The potential weakness of this report is that 16 of the 38 explanted livers (42%) did not demonstrate CCA. Of the 16 livers lacking the diagnosis of CCA at explantation, 8 had unequivocal histological or cytological confirmation before the administration of neoadjuvant therapy; 3 had suspicious cytology, structures appearing to be malignant, and underlying PSC; 2 had strictures appearing to be malignant and high carbohydrate antigen 19-9 (CA19-9) levels without PSC; and 3 had strictures appearing to be malignant, PSC, and digital image analysis and fluorescent in situ hybridization tests positive for aneuploidy. Using a similar protocol of intense brachytherapy and 5-FU, Sudan et al.28 at the University of Nebraska reported a survival rate of 45% (median follow-up = 7.5 years) and improved locoregional control of cancer in comparison with previous outcomes.

In 2006, Heimbach et al.29 extended the results from the Mayo Clinic and analyzed the specific factors that might serve as predictors of disease recurrence after neoadjuvant chemoradiotherapy. Patients older than 45 years, those with high CA-19.9 levels (>100 U/mL), and those with larger tumors were more likely to develop recurrent disease. Interestingly, underlying PSC was not associated with recurrence. In this retrospective analysis, the overall 1-, 3-, and 5-year patient survival rates were 74%, 60%, and 38%, respectively.30 Not surprisingly, improved survival was associated with a diagnosis of CCA pre-OLT and with OLT after 1993. Ironically, patients who were incidentally found to have CCA had a significantly worse prognosis after transplantation. This is in contrast to a 12-year single-institution retrospective analysis of incidental CCAs treated with transplantation, which demonstrated a 5-year survival rate of 83%.31

Ongoing Areas of Research

Because multimodality treatment protocols have improved survival outcomes, OLT has become accepted as a potential treatment for appropriately selected and preoperatively treated patients with CCA. Nonetheless, there exist numerous complementary areas of questioning and current investigation. With the increasing use of high-dose neoadjuvant chemoradiotherapy in the treatment of CCA, concern has been raised about the potential for vascular complications. Smaller vessels and vascular endothelial cells are sensitive to radiation injury, and the adverse effect of radiotherapy on vascular tissue is well known.32 To address this, Mantel et al.33 reviewed data for patients who underwent liver transplantation for CCA, and they compared the incidences of vascular complications in whole organ and living donor recipient control groups. Despite an increased risk of arterial complications, portal venous complications, and overall vascular complications associated with radiotherapy, these complications did not adversely affect patient and graft survival.

In order to address the concern that residual biliary tree epithelium may lead to recurrence in patients with CCA and concomitant PSC, Wu et al.34 recently described their experience with hepatectomy-whipple en bloc followed by OLT for early-stage CCA. Of the 6 patients diagnosed with CCA who underwent hepatectomy-whipple, 1 died of unrelated causes; however, the remaining 5 were free from recurrence at 5 to 10 years. These results should prompt future consideration of complete excision of the intact biliary tree via OLT-whipple in patients with early-stage hilar CCA complicating PSC.

In an attempt to broaden the patient selection criteria, Rea and colleagues27 examined whether liver transplantation with neoadjuvant chemoradiation is more effective than resection. The 1-, 3-, and 5-year patient survival rates were 92%, 82%, and 82% after transplantation and 82%, 48%, and 21% after resection, and there were fewer recurrences in the transplant patients (13% versus 27%). In contrast, Neuhaus et al.35 reported 60% 5-year survival (with perioperative mortality excluded) after resection.35 Therefore, as additional data are collected, liver transplantation with neoadjuvant chemoradiation may prove to be a viable alternative to resection for patients with localized, node-negative hilar CCA. However, there are presently no other published reports addressing this question.

Finally, the most controversial issue pertaining to OLT for CCA is the allocation of cadaveric donor livers and the use of living donors. Opponents of the use of cadaveric livers claim that patients with CCA face increased mortality in comparison with other patients awaiting liver transplantation; therefore, transplantation for these patients is a misuse of a scare resource. However, in a 2004 article, Heimbach et al.36 responded by noting that OLT is the only opportunity for survival for patients with CCA and that with neoadjuvant therapy, the survival results are markedly improved. Therefore, it would be unethical not to offer liver transplantation to these patients. Although there have been various case reports describing living-related liver donation for CCA, the only series of such cases was published by Jonas et al.37 in 2005. Seven patients with CCA underwent living-related OLT, and there was 1 posttransplant death; however, the remaining patients were alive after a median of 20 months. Moreover, there was no donor mortality or severe donor morbidity.

Complementing OLT as a treatment for CCA, the use of molecular markers has emerged as a promising method for better understanding tumor pathogenesis, and they potentially provide prognostic information. Dozens of proteins have been associated with CCA, and there are mixed data pertaining to these markers; however, a select few have emerged as potential prognosticators.38 Specifically, the tumor suppressor gene p53, when overexpressed, has been repeatedly demonstrated to be associated with decreased overall survival.39–43 Notably, in a series of 112 patients, Cheng and colleagues44 reported that overexpression of p53 in distal bile duct carcinomas was strongly associated with significantly reduced survival independently of clinicopathological prognostic factors. Cell cycle regulatory proteins, when overexpressed, have also been associated with decreased survival. Two separate Japanese studies45, 46 have demonstrated that the overexpression and gene amplification of cyclin D1 are associated with inferior outcomes in CCA. Finally, elevated levels of the tumor marker CA19-9 appear to portend a worse outcome; however, the precise cutoff level and its implications are not yet known.29, 47–49

In conclusion, liver transplantation following neoadjuvant therapy does not appear to be an inferior cancer operation in comparison with resection and may have more indications than the current standard of care. Nonetheless, transplantation is currently an essential component in the multimodality treatment regimen of CCA, and the current treatment paradigm is outlined in Fig. 2.

Figure 2.

CCA treatment paradigm. [Color figure can be viewed in the online issue, which is available at]


NETs encompass a wide group of neoplasms that typically arise from neuroendocrine cells within the gastrointestinal tract and pancreas.50 Generally, NETs are highly vascularized and aggressive tumors; and understanding the specific molecular basis for the tumor phenotype is an area of active research. Carcinoid tumors have been associated with inherited mutations in tumor suppressor genes (multiple endocrine neoplasia 1, von Hippel–Lindau tumor suppressor, tumor suppressor 2, and neurofibromin 1)51, 52; however, the genetic abnormalities associated with sporadic NETs are less well characterized. Several growth factors, such as basic fibroblast growth factor, transforming growth factor, endothelial growth factor receptor, insulin-like growth factor receptor, and vascular endothelial growth factor (VEGF), have been associated with the development of carcinoid tumors.53–58 Recently, Zhang et al.59 demonstrated that elevated VEGF expression in NETs in humans was inversely correlated with progression-free survival. Interestingly, their data also demonstrated that the VEGF-neutralizing antibody, bevacizumab, inhibited tumor growth in a human carcinoid xenograft. We hope that such promising results are a harbinger of future molecule-based directed therapies.

Unfortunately, the overall 5-year survival rate for patients with untreated liver metastases from NETs is approximately 30% to 40%; however, there are reports that carcinoid tumors have a slightly better prognosis.60, 61 Surgical treatment is not always indicated; symptom control and improved quality and length of life can be achieved by the reduction of the levels of circulating hormones via functional hormonal blockade, tumor debulking, and ablation.61 Because the duration of the clinical response is attributable to the residual tumor burden,62 resection is indicated only if 80% to 90% of the tumor mass can be removed.63 However, if metastases are limited to the liver, OLT is a viable treatment option.61 OLT is currently being applied to patients with unresectable metastases or for palliation of medically uncontrollable symptoms.64 The use of transplantation for metastatic disease is unique to NETs. Outcomes after OLT for metastatic colorectal adenocarcinoma have been poor,65, 66 so transplantation is not a treatment option. In contrast, the outcomes for patients with NETs who undergo OLT have been promising, and in the last 10 years, there has been considerable interest in improving survival outcomes and identifying prognostic indicators.

In 1998, Lehnert67 identified 103 patients who underwent OLT for metastatic NETs in the largest review to date. The 2- and 5-year survival rates of 60% and 47% were superior to the findings of an earlier review by Bechstein and Neuhaus,68 who reported a 1-year survival rate of 52%. However, the 5-year disease-free survival rate was only 24%. Lehnert identified 3 positive prognostic factors: patient age less than 50 years, primary tumor location in the lungs or bowel, and pretransplant hormone blockade. In contrast, extensive abdominal operations and hilar lymph node involvement portended a less favorable outcome.67

Using molecular markers to better define prognostic indicators, Rosenau et al.69 reported results from the largest single-center study. Specifically, the explanted livers of 19 patients with metastatic NETs were examined immunohistochemically for the expression of Ki67 and E-cadherin. Ki67 is a nuclear protein expressed in all proliferating cells and is believed to correlate with the clinical course of tumor progression.70 When E-cadherin, an adhesion molecule, is aberrantly expressed, it is hypothesized to influence tumor metastasis.71 Interestingly, low Ki67 indices (<5%) and nonaberrant E-cadherin expression levels have been associated with significantly improved survival. Additionally, overall 1- and 5-year survival rates of 89% and 80% have been reported. In contrast, Florman et al.72 was not able to reproduce such favorable survival outcomes in a 2004 series; the 1- and 5-year survival rates among the reported 11 transplant patients suffering from NETs were 73% and 36%, respectively. Nonetheless, 3 subsequent studies (summarized in Table 1) have yielded promising outcome data and offer new information pertaining to prognostic indicators.

Table 1. NET Outcomes of Selected Series
AuthorYearnOne-Year SurvivalFive-Year SurvivalRecurrenceConclusions
Olausson et al.7420071590%90%20% at 5 years1. The survival of grafts and survival of patients with OLT and MVTx were comparable to survival after benign disease.
2. The inclusion of high-risk patients did not affect the outcome.
van Vilsteren et al.6120061987%77% at 1 year1. OLT in selected patients with hepatic metastases is a viable option.
2. The prognostic ability of Ki67 labeling is not known.
Ahlman et al.73200412100% with OLT 50% with MVTx50% at 30 months1. OLT is an option for well-differentiated tumors and tumors with a low proliferation rate (Ki67 index < 10%).
2. MVTx is associated with a higher mortality rate.

In an attempt to prospectively analyze the prognostic value of Ki67 labeling indices, Ahlman et al.73 selected patients with well-differentiated NETs exhibiting limited proliferation (Ki67 < 10%) to undergo OLT. Of the 12 patients enrolled, 4 underwent multivisceral transplantation (MVTx), and 8 underwent OLT. There were no deaths in the OLT group after a mean follow-up of 30 months (range = 1-66 months). The data for MVTx, however, were much more disappointing: a 50% mortality rate within 4 months of transplantation. Nonetheless, the recurrence-free interval in this selected cohort of survivors was promising; 50% remained disease-free at the time of publication, and this strengthens the prognostic value of the Ki67 labeling index.

In a retrospective analysis, the Mayo Clinic published extremely promising results: an 87% overall estimated 1-year survival rate for the 17 patients included in their series and a 77% estimated 1-year recurrence-free rate for the survivors. In contrast to previous reports, there was no significant correlation found between recurrence and Ki67 labeling indices, despite a statistically significant association between a low Ki67 labeling index (<2%) and improved survival.61 However, because the overwhelming majority of patients in this study had low Ki67 indices (<2%), it is very difficult to extrapolate these data. Interestingly, of the 19 patients analyzed, the patient with the highest Ki67 labeling index developed recurrent cancer and ultimately succumbed to metastatic disease.

In the most recent series, which was published in 2007, Olausson et al.74 described an 18-year experience with OLT and MVTx in 15 patients (5 with MVTx and 10 with OLT) with metastatic spread of an NET to the liver. The Ki67 labeling index did not exceed 10% for any tumor specimen, and nearly two-thirds of the patients displayed an index less than 5%. However, in contrast to previously established guidelines in which large tumors are considered a contraindication to OLT,75 12 of the 15 patients in this series demonstrated tumor burdens exceeding 50% of the liver volume. The overall 5-year survival rate was 90%, and the recurrence-free survival rate for both multivisceral and liver transplantation was 20% at 5 years. Despite the inclusion of higher risk patients, these results further support the safety and efficacy of OLT for patients with unresectable liver metastases from NETs. The data supporting MVTx, although encouraging, will need further validation.

Finally, because of the conflicting data regarding the predictive value of the Ki67 labeling index, we here report the first meta-analysis analyzing the use of this prognosticator. In order to quantify the implications of the Ki67 labeling index, data from the 4 largest series61, 69, 73, 74 were combined, and patients were divided into a low-risk group (Ki67 index < 2%) and a high-risk group (Ki67 index ≥ 2%). Fifty-five low-risk patients were identified; of these, 37 had at least 1 year of follow-up data, whereas 18 were followed for 3 years. In contrast, 22 of the high-risk patients were identified with 1-year data, and only 9 had 3 years of follow-up data. In this analysis, survival rates included all-cause mortality, and only survivors were included in the calculation of recurrence rates. Unfortunately, individual patient Ki67 indices were not available from the Mayo Clinic series61; however, in that series, tumor specimen indices varied from 0.06% to 2.35% with an average of 0.69%. In this meta-analysis, these patients were included in the low-risk category.

These results (Fig. 3) provide striking data regarding the prognostic value of the Ki67 labeling index with respect to survival and recurrence. Low-risk patients demonstrated improved 1- and 3-year survival and improved 1- and 3-year recurrence-free survival in comparison with high-risk patients. Notably, only 17% of high-risk patients were free from recurrence at 3 years versus 52% of low-risk patients. This meta-analysis strongly supports the use of the Ki67 labeling index in patient prognostication.

Figure 3.

Meta-analysis of Ki67 prognosis.


HEHE is a rare tumor of vascular origin with variable malignant potential.76 Histologically distinguishing HEHE from angiosarcomas is difficult; however, immunostaining with cytokeratin and endothelial markers such as CD31, CD34, factor VIII, and podoplanin have improved the diagnostic capabilities.77 Little is known about the pathogenesis of HEHE, although there has been speculation of an association with the use of oral contraceptives,78 exposure to vinyl chloride,79 and previous hepatitis B infection.80

The clinical course of HEHE is unpredictable, and reliable prognostic clinical and histopathological features are limited.81, 82 Appropriate treatment is dictated by the extent of disease upon presentation, and only for the atypical patient with limited intrahepatic disease is liver resection a reasonable treatment choice. Nearly 90% of patients present with bilobar and multifocal disease, and 37% present with extrahepatic involvement82; thus, less than 10% of patients are candidates for resection82 (Fig. 4).

Figure 4.

Representative computed tomography images of the livers of 2 patients with HEHE who underwent OLT, with both scans illustrating bilobar disease (top), and explanted livers with HEHE (bottom).

Unfortunately, the rarity of this disease (1 or 2 cases per 100,000 people per year)50 limits the amount of current and relevant data available for analysis. Nonetheless, over the past 2 decades, there have been multiple publications reporting the results of single-institution series. One of the earliest and largest series, published in 1995 by Madariaga et al.,83 reported actual 1- and 5-year survival rates of 100% and 71.3% for 16 patients who underwent OLT for HEHE. Additionally, the involvement of hilar lymph nodes or vascular invasion did not affect survival. Subsequently, 6 single-institution reports of OLT for HEHE have been published.84–89 However, none of these describes a series involving more than 7 patients; therefore, they are unable to report meaningful survival statistics. However, nearly all these single-institution series endorsed the proposition that limited extrahepatic disease should not be considered an absolute contraindication to OLT. Despite the limited outcome data available from single-institution studies, long-term survival data from 3 sizeable recent reviews are very promising: the 1-year survival rate is 80% or higher, and 5-year survival rates range from 54.5% to 83%. Both are comparable to survival outcomes in patients with hepatocellular cancer (Table 2).

Table 2. HEHE Survival Outcomes of Selected Reviews
AuthorYearSourceOLT (n)Survival (%)
1 Year3 Years5 Years10 Years
Rodriguez et al.762008UNOS/OPTN126806864
Lerut et al.812007European Liver Transplant Registry59938372
Mehrabi et al.822006Comprehensive review of the literature1289654.5

There are limited data regarding precise prognosticators; however, the need for pretransplant inpatient hospitalization and intensive care unit management significantly affects posttransplant survival,76 as do microvascular invasion and combined microvascular and macrovascular invasion.81 Survival is not influenced by pretransplant adjuvant therapy, lymph node invasion, or limited extrahepatic disease.81

As additional studies surface, we hope that the value of complementary and adjuvant therapies such as chemotherapy, radiotherapy, radiofrequency destruction, hormone therapy, arterial embolization, and chemoembolization will be assessed. However, because of the aggressive nature of HEHE and the likelihood of recurrent disease despite resection, the existing data support OLT in appropriately selected patients.


HB is the most common primary hepatic malignancy of children and accounts for 60% to 85% of all childhood hepatic tumors; it has an annual incidence of 1 per million90, 91 (Fig. 5). HB is classified histologically as either wholly epithelial or mixed-epithelial; the fetal subtype has the best prognosis.92 The cellular source of undifferentiated epithelial HB, which carries the worst prognosis, is not fully understood, although on the basis of ultrastructural and immunohistochemical data, hepatic stem cells have been proposed as a potential source.93 Biochemical and molecular markers related to growth and the regulation of the cell cycle, such as Polo-like kinase-1, have been implicated in the pathogenesis of HB.94, 95 Alterations of the beta-catenin signaling pathway represent the most common molecular alteration in sporadic HB95; interestingly, these variants display a distinct spatial distribution pattern. Often, peripheral tumor cells will exhibit aberrant beta-catenin reactivity in contrast to more central locations, in which this abnormal activity will be absent.95 These genetic discrepancies among cells existing within a single tumor may provide insight into tumorigenesis and improve modalities for cell-directed therapy.

Figure 5.

Representative computed tomography images and gross explanted livers with HBs. The upper image presents the liver of an 8-year-old boy with a low attenuation mass, which was found primarily in the right lobe of the liver and measured approximately 8 cm in its greatest dimension. This patient had chemoresponsive lung metastases and, therefore, underwent OLT. The middle scout images demonstrate diffuse bilobar disease. The lower left image shows an explanted liver with an untreated HB; in contrast, the lower right image shows an explanted liver from a patient who underwent pretransplant chemotherapy.

Fortunately, patients with resectable tumors have an overall survival rate of 80% at 5 years96, 97; however, there are subtle differences between North American and European hospitals with respect to the optimal treatment algorithm.

In North America, HBs are staged according to the residual tumor after surgery, whereas in Europe, tumors are staged according to the degree of intrahepatic tumor extension, which is classified by the Pretreatment Extent of Disease scoring system.92, 98 Additionally, there are variations in the timing of chemotherapy. In Europe, studies coordinated by the Childhood Liver Tumour Strategy Group of the International Society of Paediatric Oncology (SIOPEL) have emphasized preoperative chemotherapy99 because of the high frequency (90%-95%) of HB chemosensitivity.100 Chemotherapy has the potential to downstage HB tumors and thus enable resection for patients whose tumors were previously determined to be unresectable.101, 102 However, in a 2002 review, Finegold103 argued that 40% of HBs are primarily resectable, and those with pure fetal histology and a low mitotic rate do not require toxic chemotherapy. Additionally, preoperative chemotherapy leads to increased surgical complications, including bleeding, infection, and cardiac and renal toxicity. Finegold therefore concluded that patients with potentially resectable HBs should be treated with surgical resection and be spared the unwanted effects of chemotherapy. Further investigation is warranted to better define the role of preoperative chemotherapy.

Although resection is the preferred treatment when the tumor burden is not prohibitive, liver transplantation remains the only treatment option for unresectable HBs. The overall survival rate after transplantation increased from 50% in the late 1980s104 to 77.8% in 2008.105 In 2004, the SIOPEL-1 study reported an 85% overall survival rate at 10 years post-OLT for 7 children who underwent primary OLT and a 40% overall survival rate for 5 children who underwent rescue OLT.106 In a 2006 review of 135 patients identified from the United Network for Organ Sharing (UNOS) database, the overall 1-, 5-, and 10-year patient survival rates were 80%, 69%, and 66%, respectively.107 Finally, in 2008, the results from a large single-institution series105 showed overall actuarial patient and graft survival to be 91%, 77.6%, and 77.6% at 1, 5, and 10 years, respectively, with no retransplantation at a median follow-up of 60 months. Collectively, OLT has achieved survival statistics similar to those for patients with HCC.107

Currently, the only contraindication to transplantation for HBs is the persistence of 1 or more sites of extrahepatic disease unresponsive to chemotherapy.105 Unlike other malignancies in which transplantation is a potential treatment option, the presence of lung metastases at diagnosis is not an absolute contraindication. Lung metastases often respond well to chemotherapy and can disappear completely or become resectable with preoperative chemotherapy.106, 108, 109

Tumor recurrence is the most common cause of death after transplantation,96, 104 and this has prompted some centers to administer posttransplant chemotherapy. In 2004, Otte et al.106 reviewed the published experience of liver transplantation for HB and reported that posttransplant chemotherapy did not provide a statistically significant survival advantage and, in combination with immunosuppression, might lead to added morbidity. However, different centers use different treatment protocols, and further study is needed to better address this issue.


FLHCC typically presents as a large solitary tumor and is histologically characterized by fibrous stroma arranged in thin parallel lamellae around tumor cells.110 This malignancy was once believed to be a variant of HCC; however, it is now considered a distinct form of liver cancer.111 Unlike HCC, FLHCC is not typically associated with chronic liver disease,112 so resection is the preferred treatment option.113 When partial hepatectomy is contraindicated, OLT has been proposed as an alternative.

Because of the rarity of this tumor and the infrequent need for transplantation, patients with FLHCC are not prioritized for liver transplantation114; this limits the available data. Nonetheless, over the last 25 years, there have been reports of OLT for FLHCC. Starzl and colleagues115 were some of the first to promote the idea of aggressive surgical resection and transplantation; however, outcomes after transplantation have been modest, and most published series have been underpowered. Pinna et al.116 reported a 45% 3-year survival rate and a 35% 5-year survival rate in 13 patients who underwent OLT. Pichlmayr117 reported similar outcomes, with 1-, 5-, and 10-year survival rates of 63%, 38%, and 25%, respectively, and a median survival time of 35 months. In a 2000 series,118 9 patients with FLHCC underwent OLT with a median follow-up of 25 months. The overall survival rates at 1, 3, and 5 years were 90%, 75%, and 50%, respectively. Recently, living donor liver transplantation has emerged as a resourceful treatment option for FLHCC because of the shortage of cadaveric livers and improved posttransplant survival statistics.119, 120

Although there are limited existing data, outcomes have improved over the last 2 decades. With improving split liver techniques and the potential use of marginal donor organs, further research regarding liver transplantation for FLHCC is warranted.


Unlike the promising data pertaining to the previously mentioned non-HCC tumors, the data regarding OLT for either primary or metastatic sarcoma of the liver are disappointing. In one of the few published reviews, Husted and colleagues121 examined records of all patients in the US component of the Israel Penn International Transplant Tumor Registry and identified 19 patients who underwent OLT for sarcoma of the liver. In addition to a recurrence rate of nearly 100% in 18 months, the 5-year survival rate was 5%. Therefore, at this time, OLT is not a recommended treatment option for sarcoma of the liver.


Liver transplantation is currently an integral component of treatment regimens for specific nonhepatocellular malignancies. In appropriately selected patients suffering from CCA, NETs, HB, or HEHE, OLT provides the best chance for cure or survival. Additionally, an increasing amount of data supports immunosuppression regimens that use conversion from calcineurin inhibitors to the macrolide antibiotic sirolimus because of its decreased nephrotoxicity and antitumor properties. In conclusion, the indications for liver transplantation for non-HCC malignancy and its limitations have evolved dramatically over the past decades and will continue to be redefined through future research and investigation.