A, Patient survival at 1, 5, and 10 years is 93%, 89%, and 84%, respectively, for HBL and 90%, 90%, and 72%, respectively, for HCC (log-rank test, P = .23). B, Disease-free survival at 1, 5, and 10 years is 93%, 82%, and 82%, respectively, for HBL and 90%, 78%, and 78%, respectively, for HCC (log-rank test, P = .11). C, Graft survival at 10, 40, and 70 months is 86%, 83%, and 83%, respectively, for HBL and 100%, 100%, and 85%, respectively, for HCC (log-rank test, P = .11). We found no significant difference between patients with HBL and HCC in all 3 comparisons as determined by log-rank test.
Customize your JAMA Network experience by selecting one or more topics from the list below.
Pham TA, Gallo AM, Concepcion W, Esquivel CO, Bonham CA. Effect of Liver Transplant on Long-term Disease-Free Survival in Children With Hepatoblastoma and Hepatocellular Cancer. JAMA Surg. 2015;150(12):1150–1158. doi:10.1001/jamasurg.2015.1847
Hepatoblastoma (HBL) and hepatocellular cancer (HCC) are the most common primary hepatic malignant neoplasms in childhood. Given the rarity of these childhood tumors and their propensity to present at advanced stages, updated long-term data are needed.
To determine the efficacy of liver transplant in children with HBL or HCC.
Design, Participants, and Setting
This single-institution retrospective medical record review and analysis spanned from January 1, 1997, through September 17, 2014, at Stanford University School of Medicine. A total of 40 patients younger than 18 years underwent liver transplant for treatment of HBL (n = 30) or HCC (n = 10) during the study period, with follow-up until September 17, 2014. Patients who underwent transplant for HCC included those with tumors that were greater in size than what is proposed by the Milan (a single tumor measuring ≤5 cm or ≤3 nodules measuring ≤3 cm) and University of California, San Francisco (single tumor measuring ≤6.5 cm or ≤3 nodules measuring ≤4.5 cm and a total diameter of ≤8 cm), criteria.
Main Outcomes and Measures
Disease-free and overall patient survival and graft survival.
Using a Kaplan-Meier survival analysis, 1-, 5-, and 10-year disease-free survival after liver transplant was 93%, 82%, and 82%, respectively, for 30 patients with HBL and 90%, 78%, and 78%, respectively, for 10 patients with HCC. Risk factors associated with HBL recurrence after transplant included having pretreatment extent of disease stage IV lesions and a longer waiting list time and being older at the time of the transplant. Recurrence was found in 2 of 7 patients with HBL and pretransplant metastases, which were not found to be an independent risk factor for recurrence. Patients with HCC larger than the proposed Milan and University of California, San Francisco, criteria experienced good 5-year disease-free (82%) and overall (78%) survival after transplant. Being older at the time of transplant (18 vs 11 years; P = .04) and the presence of metastatic disease (1 patient vs none; P = .05) were associated with HCC tumor recurrence.
Conclusions and Relevance
Liver transplant combined with chemotherapy is an excellent treatment that provides long-term disease-free survival in children diagnosed with advanced HBL and HCC. Early addition to a waiting list and aggressive multimodal therapy provide excellent results. Transplant should still be considered in children with HCC larger than the Milan and University of California, San Francisco, criteria.
Hepatoblastoma (HBL) and hepatocellular carcinoma (HCC) are the most common primary liver cancers in infancy and childhood. Hepatoblastoma alone accounts for 91% of all liver tumors diagnosed in children younger than 5 years, and most HBLs arise in children by 3 years of age.1-3 Most cases are sporadic, but a genetic component has been recognized because HBL has been associated with congenital abnormalities, such as Beckwith-Wiedemann syndrome and familial adenomatous polyposis.
Risk stratification and staging for HBL have varied throughout the years as cooperative study groups have created several different staging criteria. Recently, the pretreatment evaluation system of tumor extension (pretreatment extent of disease [PRETEXT]) for risk assessment developed by the Société Internationale d’Oncologie Pédiatrique–Epithelial Liver Tumour Study Group (SIOPEL) (now the International Childhood Liver Tumors Strategy Group)4 has been adopted universally. The PRETEXT system divides patients into standard- and high-risk groups. High risk is characterized by PRETEXT stage IV tumors (all sections of the liver involved by tumor), an α1-fetoprotein (AFP) level less than 100 ng/mL (to convert to micrograms per liter, multiply by 1), or the presence of additional PRETEXT criteria suggestive of extrahepatic spread, including the portal vein, hepatic vein, or inferior vena cava; tumor rupture or extrahepatic extension; lymph node involvement; or distant metastases. High-risk patients are more likely to have tumor recurrence or treatment failure.
Hepatocellular carcinoma is rare and represents less than 1% of all tumors in the pediatric population but is the second most common primary liver tumor after HBL. Hepatocellular carcinoma may be associated with underlying liver disease, such as cirrhosis, metabolic disorders, chronic hepatitis, or chronic cholestasis. These tumors typically present in advanced stages, and long-term survival has been reported to be as low as 30% within the previous decade.5 Given the association of HCC with chronic liver disease, these tumors typically present in older children and account for 87% of primary hepatic tumors in children aged 15 to 19 years.2
Management of HBL and HCC has evolved during the past 3 decades. Disease-free survival for HBL has increased significantly from 30% in the prechemotherapy era to 70% through 80% in most recent studies.6,7 Despite this increase, disease-free survival hinges on complete surgical resection. The role of liver transplant as a treatment for unresectable HBL has grown over time. According to United Network for Organ Sharing (UNOS) data, 5 liver transplants were performed for HBL in 1990. That number has steadily increased to 43 in 2013.8 The number of pediatric transplants for HCC has remained sparse at less than 10 per year during the same time span.8 Despite this increasing reliance on liver transplant for the treatment of unresectable HBL and HCC, long-term outcomes remain unclear. However, advances in chemotherapy in combination with tumor removal have led to improved survival.3
Several groups5,9-11 have reported outcomes after transplant that included patients before the widespread use of the SIOPEL and Childhood Oncology Group (COG) protocols. Herein, to our knowledge, we report the outcomes of the largest single-center collection of patients with HBL treated with the SIOPEL or the COG chemotherapy protocols followed by liver transplant. By excluding patients who were not treated with these regimented chemotherapy protocols (pre-1997), our results should accurately represent current treatment for HBL.
Hepatocellular carcinoma in children remains resistant to chemotherapy, with complete resection as the mainstay treatment of choice. The role of liver transplant for HCC in adults has been studied extensively with the adoption of the Milan criteria to determine transplant candidacy.12 Whether this criteria and other expanded criteria, such as the University of California, San Francisco (UCSF), criteria, are valid in children remains unclear because HCC appears to behave differently in this population.13,14 Herein we review our most recent experience treating the advanced forms of these tumors with liver transplant.
We performed a retrospective review of electronic medical records and the UNOS UNet database for patients younger than 18 years who underwent liver transplant for HBL (n = 31) or HCC (n = 10) from January 1, 1997, through September 17, 2014. This study was approved by the institutional review board of the Stanford University School of Medicine. Informed consent was waived through the institutional review board, and the data were deidentified. We reviewed patients’ medical records for demographics, underlying disease, radiographic findings, pathology reports, type of graft, pretransplant and posttransplant adjuvant therapy, patient survival, graft survival, and recurrence of tumor. For HCC, the Milan criteria are defined as a single tumor measuring 5 cm or less or up to 3 nodules measuring 3 cm or less, and the UCSF criteria are defined as a single tumor measuring 6.5 cm or less or up to 3 nodules measuring 4.5 cm or less and a total diameter of 8 cm or less. Other primary liver tumors, such as focal nodular hyperplasia, adenoma, and sarcoma, were not included. One patient with HBL was excluded because of a lack of medical records. National survival statistics were obtained from the Organ Procurement and Transplantation Network database. Mean follow-up was 8.0 years for patients with HBL and 7.4 years for patients with HCC. Follow-up was complete by September 17, 2014. Data analysis was performed from January 1, 1997, through the end of follow-up on September 17, 2014.
Statistical differences were determined by the unpaired 1-tail t test and Kaplan-Meier survival analysis with significance at P ≤ .05. We used the log-rank test to determine significance. Statistical and graphing software used included Excel 2010 (Microsoft Corp) and JMP Pro, version 11 (SAS Institute Inc).
The characteristics of 40 patients who underwent transplant for HBL (n = 30) and HCC (n = 10) from 1997 through 2014 are depicted in Table 1. The mean age at the time of transplant was significantly younger for patients with HBL compared with those with HCC (4 vs 12 years; P < .01). Male patients predominated in the HBL group (20 of 30 [67%] vs 5 of 10 [50%]). Most patients were white or Hispanic (33 of 40 patients [83%]). Every patient in the HBL group presented with an abdominal mass or abdominal pain with or without a mass. The AFP level was significantly higher in those diagnosed with HBL compared with HCC (Table 2 and Table 3).
All patients with HBL underwent computed tomography or magnetic resonance imaging before surgery. All hepatic masses underwent biopsy to confirm the diagnosis of HBL. Levels of AFP were elevated in all patients to a varying degree and ranged from 10 000 to 2 900 000 ng/mL. Tumors were staged and risk stratified according to the SIOPEL criteria using the PRETEXT stages. The COG and SIOPEL criteria recommend transplant referral for all PRETEXT stage III (3 sections involved) and IV (all 4 sections involved) lesions.7,16 Thus, most patients who underwent transplant had PRETEXT stage III or IV tumors with or without hepatic vein or vena cava invasion, portal vein involvement, metastases, or tumor rupture (Table 2). The only exception was patient 26, who had PRETEXT stage II disease but whose resection was complicated by transection of the porta hepatitis obviating immediate transplant. Given the advanced disease in these patients, all of them received preoperative and postoperative chemotherapy according to the COG or SIOPEL study group guidelines.3,15
Twenty-six of 30 allografts (87%) were from deceased donors, and 14 of 30 grafts (47%) were reduced in size and provided the left lobe or left lateral segment for transplant. Two patients underwent resection before transplant. Patient 26 underwent right hepatic lobectomy that was complicated by complete transection of the porta hepatis, as described above. Patient 11 underwent a trisegmentectomy but had recurrence in the remaining lobe of the liver. Given the prior resection, the transplant was technically difficult and the patient developed a postoperative hepatic artery thrombosis requiring a second transplant several days after the initial transplant (Table 2).
Seven patients with HBL (23%) had documented metastatic disease before transplant. All patients had surgical resection of metastases or radiographic resolution after chemotherapy and before transplant. Pathologic examination was not routinely performed for metastatic lesions responsive to chemotherapy. In particular, patient 7 underwent successful treatment for enlarged mediastinal nodes.
Risk factors for tumor recurrence included having PRETEXT stage IV disease (P ≤ .01) or tumor rupture (P = .01) and being older at the time of the transplant. Mean age at the time of the transplant for those who developed recurrence was 78 months, whereas the mean age for those who did not develop recurrence was 36 months (P = .01) (Table 4). PRETEXT stage IV disease (P ≤ .01) and recurrence (P ≤ .01) after transplant were associated with higher mortality (Table 4). Those who developed recurrence spent significantly more time on the transplant waiting list, with a mean waiting list time of 31 vs 15 days for those without recurrence (P ≤ .01). Long-term disease-free survival at 10 years was 82%, overall patient survival was 84%, and graft survival was 83% (Figure). Recurrence of HBL after transplant was noted in the lung, transplanted liver, or both. Two of the lung recurrences occurred at the same site as documented preoperative metastases.
Hepatic artery and/or portal vein thrombosis occurred in 4 patients (13%) undergoing transplant for HBL. Mean age for patients who developed thrombosis was 18 months whereas mean age for those who did not develop thrombosis was 47 months. The difference in age appears drastic but was not significant (P = .08). Vessel thrombosis was found to be a significant risk factor for graft failure because all 4 patients required retransplant (P ≤ .01).
Ten transplants were performed for HCC. The mean age at transplant was 12 years, which was significantly higher than the age for HBL (P ≤ .01) (Table 1). All patients received a deceased donor graft, with 2 requiring size reduction of the graft at the time of transplant. Half of the patients had underlying liver disease with or without cirrhosis. Only 1 patient developed HCC without underlying disease. The AFP level was elevated in some patients but was not a consistent marker for tumor diagnosis or burden. Patient 3 had undergone prior resection for HCC but developed a recurrence in the remaining liver. At the time of transplant, the recurrent tumor was directly invading the adjacent stomach and required en bloc partial gastrectomy during the hepatectomy. No other patients had known metastases at the time of transplant (Table 3).
Neoadjuvant/adjuvant chemotherapy for HCC has been difficult to standardize because of the relative chemoresistance of HCC and disease rarity in children. Five patients with HCC (50%) in our series received neoadjuvant systemic chemotherapy with only limited tumor response. Transarterial chemoembolization was performed more frequently in patients who waited longer on the transplant list. Postoperative chemotherapy was given to those patients whose tumors responded to preoperative chemotherapy.
Two of the 10 patients had a tumor that met the definition of the Milan criteria. The remaining 8 had tumors either that were beyond established expanded criteria or for which not enough was known to characterize them. There was no difference in recurrence within the 2 groups (P = .24). Survival was not assessed because of the limited number of patients in this group. None of the Milan criteria patients had metastatic disease at the time of transplant, whereas 1 patient with an 8-cm fibrolamellar tumor did.
Two patients had recurrences that presented as a multisite recurrence in one and resectable lung nodules in the other. Tumor size, vascular invasion, and time to transplant were not associated with recurrence. Being older at the time of transplant (P = .04) and having metastatic disease (P = .05) were risk factors for tumor recurrence. Three patients died after transplant, including one who died less than 1 year after transplant owing to tumor recurrence. The other 2 deaths occurred more than 8 years after surgery secondary to reasons not related to liver transplant or HCC (Table 3). Children with the fibrolamellar variant of HCC appeared to have a worse prognosis, but the incidence was not found to be significantly higher in those who died after transplant (P = .06). Long-term disease-free survival at 10 years was 78%, with overall patient survival of 72% and graft survival of 85% (Figure).
Several studies have shown improved long-term survival during the past 3 decades for children with primary liver tumors.10,17 This benefit has been observed especially in HBL as collaborative study groups have determined optimal chemotherapy regimens that are timed to coincide with surgical treatment.3,18 To avoid tumor progression and metastases, patients with HBL are now considered to have 1b status (UNOS designation for transplant) at the initial listing as opposed to waiting 30 days at a Model for End-Stage Liver Disease/Pediatric End-Stage Liver Disease model score of 30. Advanced and unresectable HBL has a historic 5-year survival of 69%,6 but with aggressive transplant listing and improved chemotherapy, we demonstrate a 10-year survival of 84%. This survival is similar to if not better than the current Organ Procurement and Transplantation Network data showing 3-year survival after liver transplant for HBL in patients younger than 18 years to be 77.97%.8
Compared with PRETEXT stage III tumors, PRETEXT stage IV tumors were associated with a higher risk for tumor recurrence and patient death after transplant. Vascular invasion, metastatic disease, AFP level at the time of diagnosis, postoperative chemotherapy, and tumor histologic characteristics were not independently associated with tumor recurrence or patient death. Tumor rupture occurred in a single patient, who developed recurrence 1 year after the transplant. Being older at transplant was associated with a higher risk for recurrence. Because the number of older patients with HBL is quite small, determination of whether age is an independent risk factor for tumor recurrence or simply a marker for different tumor biology is difficult. Those who develop recurrence are on the transplant waiting list more than twice as many days as those who do not develop recurrence (Table 4). This difference may reflect lapses in chemotherapy or the biological aggressiveness inherent to HBL. Regardless, the recent modification of UNOS policy eliminating the 30-day waiting period before institution of 1b status should reduce wait times further. Rescue or salvage transplant for patients with HBL has been associated with an increased risk for recurrence in previous reports.11,19 However, our series includes only 1 patient who underwent rescue transplant after resection because of recurrence. This patient did not develop recurrence but did have surgical complications requiring retransplant (Table 2).
Pretreatment metastatic disease was found not to be associated with recurrence or death after transplant for HBL. In fact, several of our patients with metastases at the time of diagnosis were successfully treated with chemotherapy with and without surgical resection and experienced long-term disease-free survival after transplant. This finding is consistent with previous reports describing HBL as a chemosensitive tumor in which pulmonary metastases can be treated without detriment to long-term survival.9,20 In our study, 2 patients with lung metastases developed recurrence at the site of the treated metastases. These metastases were treated with chemotherapy alone, given their complete regression on imaging. Although our number of patients is too low to draw substantial conclusions, it may be prudent to combine resection of metastases with chemotherapy to ensure the removal of metastatic disease irrespective of radiographic resolution.
Liver transplant is not without complication. Hepatic artery thrombosis is a devastating complication after pediatric liver transplant that is a significant cause of graft loss and morbidity.21,22 In this series, 3 patients (11%) developed hepatic artery thrombosis requiring retransplant within 1 month of the initial transplant. One patient died after retransplant secondary to primary nonfunction of the graft. Among the remaining patients, one developed portal vein thrombosis requiring retransplant, and the other lost the graft owing to chronic rejection. These patients are at higher risk for thrombosis because of technical challenges associated with anastomoses of smaller vessels, use of segmental grafts, and the increased risk for thrombosis associated with hepatic malignant neoplasms.23 This finding is similar to that of a previous series with incidence of hepatic artery thrombosis as high as 28%.11
Advanced unresectable HCC is often found to be chemoresistant. Sorafenib is the only chemotherapeutic agent approved for advanced treatment of HCC in adults.24 Sorafenib has not been studied in children and its efficacy in this population is unclear. A standard chemotherapy regimen does not exist for children with HCC, and the regimen before and after chemotherapy has varied greatly in our patients (Table 3). Debate is ongoing on whether HCC behaves differently in children than in adults. Most HCC in adults is related to cirrhosis, whereas most cases in children occur without underlying liver disease.13 Our findings are consistent with this conclusion, because 6 of our 10 patients with HCC (60%) did not have cirrhosis and 5 (50%) had no underlying liver disease. Adults who undergo transplant for HCC larger than what is defined in the Milan and UCSF criteria are at higher risk for recurrence and death.12,25,26 Only 2 patients undergoing transplant for HCC in this study met the Milan criteria; the rest were well outside any established criteria (Table 3). Recurrence occurred in 2 patients, one of whom had undergone a prior resection and had extrahepatic disease extension at the time of the transplant. Historically, patients undergoing salvage transplant have overall poor outcomes.27 The risk for tumor recurrence in children with HCC was not related to tumor size, vascular invasion, multiple tumors, or fibrolamellar variant, in contrast to the well-known risks these factors hold in adults (Table 4).28 Patients who developed recurrence after transplant were significantly older, with a mean age of 18 compared with 11 years (P = .04). The HCC disease process in older children may mimic that in adults and place them at high risk for recurrence when undergoing transplant outside the Milan and UCSF criteria because older children are more likely to have a chronic inflammatory process of adequate duration to induce tumor genesis in a similar fashion to the adult disease process.
The limiting factor of this study is its retrospective nature and the inherent selection bias. The small number of patients in both groups weakens the statistical power and makes it difficult to reach concrete conclusions, especially for HCC. Regardless, this sample is the largest, to our knowledge, of patients with HBL undergoing transplant at a single center within the past 2 decades and is an accurate reflection of outcomes with our current treatment regimens. Prospective multicenter databases, such as the Pediatric Liver Unresectable Tumor Observatory (PLUTO registry), will be vital to establishing evidence-based clinical practice guidelines.19
Liver transplant combined with chemotherapy provides the best curative chance in children with advanced, unresectable HBL. We show that PRETEXT stage IV tumors are strongly associated with tumor recurrence and death, whereas metastatic disease in HBL is treatable and is a relative but not absolute contraindication to transplant. More time spent on the transplant waiting list is associated with an increased risk for recurrence in HBL. Hepatocellular carcinoma in children behaves differently than in adults because transplant for lesions well outside the Milan and UCSF criteria results in excellent long-term survival. More investigation is needed to further establish the role of transplant in children with HCC.
Accepted for Publication: April 16, 2015.
Corresponding Author: C. Andrew Bonham, MD, Division of Abdominal Transplantation, Department of Surgery, Stanford University School of Medicine, 750 Welch Rd, Ste 200, Palo Alto, CA 94304 (firstname.lastname@example.org).
Published Online: August 26, 2015. doi:10.1001/jamasurg.2015.1847.
Author Contributions: Drs Pham and Bonham had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design: Pham, Gallo, Concepcion, Bonham.
Acquisition, analysis, or interpretation of data: Pham, Gallo, Esquivel, Bonham.
Drafting of the manuscript: Pham, Gallo, Bonham.
Critical revision of the manuscript for important intellectual content: Pham, Concepcion, Esquivel, Bonham.
Statistical analysis: Pham, Gallo, Bonham.
Administrative, technical, or material support: Bonham.
Study supervision: Gallo, Esquivel, Bonham.
Conflict of Interest Disclosures: None reported.
Funding/Support: This study was supported in part by contract 234-2005-37011C from the Health Resources and Services Administration.
Role of the Funder/Sponsor: The funding source had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.
Disclaimer: The content is the responsibility of the authors alone and does not necessarily reflect the views or policies of the US Department of Health and Human Services, nor does mention of trade names, commercial products, or organizations imply endorsement by the US government.
Previous Presentation: This study was presented at the 86th Annual Meeting of the Pacific Coast Surgical Association; February 21, 2015; Monterey, California, and is published after peer review and revision.
Additional Contributions: Sarah Conlon, PA-C, Lucille Packard Children’s Hospital, Pediatric Intestinal Translantation, assisted with data collection. Linda Wong, MD, Department of Surgery, Liver Transplant Program, University of Hawaii School of Medicine, reviewed the manuscript. Neither received financial compensation for their role.