Successful nontransplant resection of POST-TEXT III and IV hepatoblastoma

Authors


Abstract

BACKGROUND:

Liver transplantation is increasingly advocated as primary surgical therapy for children with hepatoblastoma involving 3 or 4 sectors of the liver after neoadjuvant chemotherapy. This study evaluated the results of nontransplant hepatectomy in children who might otherwise have been considered for liver transplantation.

METHODS:

All children who underwent resection at a single institution from 1998 to 2009 for POST-TEXT IV or centrally located POST-TEXT III hepatoblastoma after neoadjuvant chemotherapy were reviewed.

RESULTS:

Fourteen children (7 boys) with a median age of 8 months at diagnosis met study criteria. Pulmonary metastases in 3 patients were resected in 2 and resolved with chemotherapy in 1 patient. Preoperative grouping after neoadjuvant chemotherapy was POST-TEXT IV in 3 patients and POST-TEXT III in 11 patients. Thirteen of 14 (93%) children who underwent aggressive resection despite being potential candidates for primary transplantation were alive and tumor-free with a median follow-up of 57 months. Observed survival rates at 1, 2, and 5 years were 93%, 91%, and 88% respectively. Event-free survival rates at 1, 2, and 5 years were 93%, 91%, and 75%, respectively.

CONCLUSIONS:

Excellent survival (93%) was obtained with aggressive resection in children with POST-TEXT III and IV hepatoblastoma meeting criteria for transplant referral. The 1 death occurred in a patient with unfavorable small cell histology. These children should be managed at institutions experienced in both advanced pediatric hepatobiliary surgery and transplantation. Operative exploration was frequently required to ultimately determine which tumors can be resected and which require transplantation. Cancer 2011. © 2010 American Cancer Society.

Hepatoblastoma is the most common primary liver neoplasm in children, accounting for 28% of all liver tumors and 66% of malignant liver neoplasms in the pediatric and adolescent populations.1, 2 More than two-thirds of cases occur in children less than 2 years of age, and hepatoblastoma accounts for >90% of malignant liver tumors in this age group.2 The incidence of hepatoblastoma has increased in recent years and a correlation with very low birth weight has been reported.3-5

The survival rate for children with hepatoblastoma has increased to greater than 75% in recent years, largely due to the introduction of cisplatin-containing neoadjuvant chemotherapy regimens.6-9 Preoperative chemotherapy allows many initially unresectable tumors to be safely removed.10, 11 For those children with disease contained to the liver that is deemed unresectable after a predetermined course of chemotherapy, liver transplantation is the only option. Long-term survival of 80% to 90% has been reported in children with unresectable hepatoblastoma who undergo primary transplantation.12-14

The excellent results of primary liver transplantation for unresectable hepatoblastoma, in contrast to the poor outcomes reported in children requiring rescue transplant for recurrence after a primary resection, have led some to advocate for the expanded use of primary transplantation.13, 15 In the current Children's Oncology Group (COG) hepatoblastoma protocol (AHEP0731), tumor extent is determined from computed tomography (CT) imaging using the PRETreatment EXTent of disease (PRETEXT) system, which is now referred to as POST-TEXT for studies obtained after treatment with neoadjuvant chemotherapy. Children with POST-TEXT III and IV tumors are referred to a surgical center with expertise in pediatric liver transplant and extreme liver resection at diagnosis if possible and no later than just after the second cycle of chemotherapy. This algorithm is designed to prevent protracted courses of chemotherapy and ensure expeditious access to transplant for truly unresectable tumors. However, the PRETEXT/POST-TEXT system is known to overstage in up to 37% of cases, and operative exploration is frequently needed to make an ultimate determination of resectability. We hypothesized that excellent oncologic cure, compatible with that reported for primary transplantation, is possible with nontransplant resection in appropriately selected children with POST-TEXT III and IV hepatoblastoma.

MATERIALS AND METHODS

All patients treated for hepatoblastoma at our hospital from 1998 through 2009 were retrospectively reviewed. The study was approved by the hospital's Institutional Review Board. Pretreatment Extent of Disease (PRETEXT)16 grouping at the time of diagnosis was retrospectively determined from CT images by the senior author (R.S.) and by a pediatric radiologist (T.B.) who was blinded to the clinical outcomes of the patients. Due to the referral nature of our practice, computed tomography (CT) images from the time of initial diagnosis were not available for all patients. POST-TEXT grouping after neoadjuvant chemotherapy and before surgery was determined in all patients using the same protocol. Patients with POST-TEXT I and II tumors, and those with POST-TEXT III tumors not involving the main portal vein or its bifurcation, the hepatic vein confluence, or the inferior vena cava (IVC) were excluded because transplant evaluation is not generally warranted in these patients.

The medical records of these patients with these POST-TEXT III and IV tumors were retrospectively analyzed. Variables studied included age at diagnosis, age at operation, gender, neoadjuvant and adjuvant chemotherapy regimens, α-fetoprotein (AFP) levels, type of operation, histology, and clinical outcomes. The primary outcomes were overall and event-free survival after tumor resection.

Many of the patients were referred from another institution for consideration of transplantation versus extreme resection. All patients had CT imaging of the chest to exclude metastatic disease, and those with persistent pulmonary metastasis after neoadjuvant therapy underwent thoracotomy before liver resection. Most families were prepared for the possibility of living or deceased donor transplantation in the event that, upon exploration, the tumor proved unresectable without compromising the viability of the remaining liver, or if liver function became irrecoverably compromised after resection.

Standard hepatic trisectionectomy was performed when feasible, but nonanatomic resections were performed as needed. Nontypical operative techniques included mesohepatectomy (aka central hepatectomy), as well as shaving a margin or performing a separate wedge resection from the remaining sector of liver. In most cases, the tumor could be dissected free of the remaining portal vein branch or hepatic vein. If the abutting tumor was truly adherent, an ultrasonic dissection device was used to clear residual hepatocytes from the remaining vessel. All resections were performed by a single, experienced pediatric hepatobiliary surgeon who has performed well over 100 pediatric liver resections.

RESULTS

Twenty-three children with POST-TEXT III and IV tumors were managed at our institution between 1998 and 2009. Tumors in 8 children remained unresectable after neoadjuvant chemotherapy and they underwent primary liver transplantation.12 These 8 children all underwent operative exploration before being listed for transplant. One child had ongoing metastatic disease after chemotherapy and was not considered a transplant candidate. The remaining 14 children underwent aggressive resection despite being potential liver transplant candidates and constitute our series population (Table 1). Nine (64%) were referred from another institution for transplant evaluation or extreme resection. The median age at diagnosis was 8 months. Ten (71%) were less than 2 years of age. Seven (50%) were male. Mean serum AFP at diagnosis was 440,670 ± 373,309 ng/ml. All patients received cisplatin-based neoadjuvant chemotherapy. Mean AFP decreased to 9,969 ± 21,212 ng/ml after neoadjuvant chemotherapy.

Table 1. Management and Outcome of Children With POST-TEXT IV and Centrally Located POST-TEXT III Hepatoblastoma Managed With Aggressive Nontransplant Resection
Age (Months)/ GenderPRETEXT at DiagnosisPOST-TEXT at SurgeryMetastatic DiseaseNeoadjuvant Chemo (No. of Cycles)Adjuvant Chemo (No. of Cycles)OperationHistologyOutcome (Duration)
  • vinc indicates vincristine; cis, cisplatin; 5FU, 5-fluorouracil; ifos, ifosfamide; carbo, carboplatin; SCT, stem cell transplant.

  • a

    The patient had tumor extension into the right atrium but no distant metastatic disease.

  • b

    The patient was transplanted for ischemic cholangiopathy.

  • c

    The tumor involved all 4 sectors of the liver but was understaged by imaging.

8/FIV P2V3 (central)III P1V3 (central)Navinc, cis, 5FU (4)vinc, cis, 5FU (3)Nonanatomic hepatectomy +atrial thrombectomyMixedAlive, NED after transplantb (11 months)
3/MIV P0V2IV P0V0 (central)Nvinc, cis, 5FU (3)ifos, carbo, VP-16 (4)MesohepatectomySmall cell/ rhabdoidDeceased
7/MIV P1V3 (central)IV P0V3 (central)Nvinc, cis, 5FU, ifos, carbo (2)ifos, etop, carbo, vinc, (4)TrisectionectomySmall cellAlive, NED (15 months)
5/FIII P2 (central)III P0V1 (central)Nvinc, cis, 5FU (4)vinc, cis, 5FU (1)TrisectionectomyMixedAlive, NED (36 months)
8/MNot availableIII P0V2 (central)Nvinc, cis, 5FU (6)NoneNonanatomic extended hepatectomyNot availableAlive, NED (23 months)
32/FNot availableIII P1 (central)Nvinc, cis, 5FU (4)vinc, cis, 5FU (2)TrisectionectomyMixedAlive, NED (115 months)
4/FIII P1 (central)III P1 (central)Ncis, carbo (2)cis, carbo (4)TrisectionectomyMixedAlive, NED (104 months)
11/FIII P0 (central)III P0V1 (central)Nvinc, cis, 5FU (5)Not availableTrisectionectomyMixedAlive, NED (52 months)
4/MIV P0V3III P1 (central)Nvinc, cis, 5FU (4)vinc, cis, 5FU (2)TrisectionectomyMixedAlive, NED (55 months)
20/MNot availableIII P0 (central)Nvinc, cis, 5FU, doxorubicin (7)NoneTrisectionectomy+ wedge resectionFetalAlive, NED (102 months)
1/FNot availableIII P1V1 (central)Netop, ifos (5)NoneTrisectionectomyMixedAlive, NED (60 months)
45/MNot availableIII P1V1 (central)Pulmonary (resected)vinc, cis, 5FU, doxorubicin (4)NoneTrisectionectomyMixedAlive, NED (70 months)
28/FIV P1V2 (central)III P0V1Pulmonary (resected)cis, amifostine (6)Autologous SCTNonanatomic extended hepatectomy +wedgeMixedAlive, NED (94 months)
26/MNot availableIII P0V1cPulmonary (resolved)vinc, cis, 5FU (5)carbo, ifos, vinc (6)Trisectionectomy + wedge. Liver recurrence after 48 months resectedMixedAlive, NED (78 months)

Three patients were COG stage IV and 11 were COG stage III. All 3 patients with stage IV disease had pulmonary metastasis at the time of diagnosis. Metastatic disease resolved with chemotherapy in 1 patient and was resected in 2 patients. One additional patient had direct extension into the right atrium that was resected under cardiopulmonary bypass at the time of partial hepatectomy.

Primary liver transplantation may have been considered as an alternative to aggressive resection in these patients for the following reasons: 1) ongoing involvement of all 4 liver sectors (POST-TEXT IV) in 3 patients (Fig. 1), 2) concern for residual disease in the remaining sector of liver in 3 patients with initially PRETEXT IV tumors that were downsized to POST-TEXT III after neoadjuvant chemotherapy (Fig. 2), and/or 3) encasement, invasion, or close proximity of main hilar structures (Fig. 3) or all 3 hepatic veins after neoadjuvant chemotherapy in centrally located POST-TEXT III tumors in 10 patients. Two patients with POST-TEXT III tumors met both criteria 2 and 3; they involved all 4 sectors of the liver before chemotherapy and remained centrally located at the time of operation. One patient was categorized as noncentral POST-TEXT III but was found to have a small focus of multifocal disease in the remaining sector of liver at the time of exploration. He might have been considered for transplantation at that time, but underwent successful right trisectionectomy plus wedge resection of the segment III nodule.

Figure 1.

Computed tomography axial images of representative patients with POST-TEXT IV hepatoblastoma amenable to surgical resection. The tumor appeared to involve all 4 sections of the liver after neoadjuvant chemotherapy in (A) a 14-month-old boy and (B) a 3-month-old boy.

Figure 2.

. (A) Pre- and (B) postchemotherapy computed tomography axial images of a 4-month-old boy with a PRETEXT IV tumor that became a resectable (POST-TEXT III) tumor after neoadjuvant chemotherapy.

Figure 3.

. Computed tomography images of representative patients with centrally located POST-TEXT III hepatoblastoma. The tumor truly encases main hilar structures and invades the inferior vena cava in (A) an 8-month-old girl with atrial extension; but only abuts these structures in (B) a 5-month-old girl.

Resections included trisectionectomy or nonanatomic extended lobectomy in 10, extended lobectomy plus separate wedge resection in 3, and mesohepatectomy in 1 patient. Ten patients had mixed (fetal and embryonal) histology, 1 had fetal histology, and 2 had small cell undifferentiated histology (one of which had aspects of rhabdoid morphology). One patient had a microscopically positive margin and the patient with atrial extension had a positive atrial margin. Further adjuvant chemotherapy was given to 9 of 13 patients (unknown in 1).

Surgical complications occurred in 2 patients. The patient who required cardiopulmonary bypass for removal of tumor from the right atrium developed ischemic cholangiopathy and required a living donor liver transplant. Another child developed portal vein thrombosis and required a distal splenorenal shunt.

Thirteen of 14 patients (93%) are alive and tumor-free with a median follow-up of 57 months (range, 11 to 115 months). The 1 patient who died had unfavorable small cell undifferentiated histology with aspects of rhabdoid morphology. All 6 patients who responded to neoadjuvant chemotherapy with a >99% decrease in serum AFP are alive and event-free. Among survivors, 10 (77%) have at least 2 years and 7 (46%) have at least 5 years follow-up. One patient had a hepatic recurrence 48 months after his initial surgery and underwent repeat resection. He required no additional chemotherapy and is doing well 30 months after his second operation. Observed survival rates at 1, 2, and 5 years were 93%, 91%, and 88% respectively. Event-free survival rates at 1, 2, and 5 years were 93%, 91%, and 75%, respectively.

DISCUSSION

We report 93% survival after nontransplant resection in children with POST-TEXT IV and centrally located POST-TEXT III tumors who might have otherwise been considered for primary transplantation. The 1 child who died had unfavorable small cell undifferentiated histology and may have done equally poorly with transplantation. Our results compare favorably with the 79% event-free survival reported for COG stage III and IV tumors that are completely resected17 and 82% overall survival for patients undergoing primary liver transplant for POST-TEXT III and IV tumors.18

Liver transplantation plays an essential role in the management algorithm of children with large and multifocal hepatoblastoma. During this study period, 8 children at our institution underwent primary liver transplant for unresectable hepatoblastoma with 88% survival.12 Long-term survival of 80% to 90% in children with POST-TEXT III and IV hepatoblastoma undergoing primary transplant has been similarly reported from several institutions around the world.13, 14, 19 Three-year survival in patients with bilobar disease undergoing transplant approaches that achieved in patients with standard risk tumors managed by straightforward resection.13 Accordingly, protracted courses of chemotherapy, with the intention of making unresectable tumors resectable, as well as hazardous attempts at resection are now avoided in patients who are candidates for transplantation.13, 20

Despite the oncologic success of liver transplantation for POST-TEXT III and IV tumors, transplantation is not without incumbent morbidity and mortality. If equivalent long-term disease-free survival were achievable with partial hepatectomy for a subset of these patients, this latter approach would be preferable. Recommendations for primary transplant in all patients with POST-TEXT IV or centrally located POST-TEXT III disease are based on indirect and hypothetical evidence only.13, 15, 18, 21 Proponents of this approach cite 1) the universally poor outcomes that have been reported in patients requiring rescue transplant for disease recurrence after resection, 2) the risk of macroscopic residual disease after radical nonanatomic resections, 3) the potential hazards of microscopic residual disease when centrally located tumors are dissected off main hilar structures, and 4) the possibility of viable tumor cells in sections of the liver that appear to have been cleared of disease by preoperative chemotherapy.13-15, 18, 21 Although the outcomes of rescue transplant for recurrent disease after resection have been poor, with most series reporting 20% to 30% survival,12, 13 it is unknown how often and in what patients this scenario occurs. Although inferior outcomes in patients who received nonanatomic liver resections were reported in 1 study, 17% of patients in this study had incomplete tumor resection.8 Complete macroscopic tumor resection must be achieved.22, 23 Whether it is accomplished with total hepatectomy and transplant, anatomic partial hepatectomy, or nonanatomic resection may not matter. The need for microscopically negative margins may be less rigid. Although inconclusive, existing evidence does not link microscopic residual disease with tumor recurrence or death.22, 24 For patients with central tumors that are dissected off main hilar structures, postoperative chemotherapy may be sufficient to destroy any remaining tumor cells adherent to these central vessels. Finally, the management of patients with multifocal tumors where chemotherapy clears 1 or more sections of the liver of macroscopic tumor is challenging. Given the success of cisplatin-based chemotherapy at abolishing small pulmonary metastases and downsizing large multifocal tumors to render them resectable, it is possible that these patients do not all require transplant.10, 25, 26 Our experience would suggest that this course of action is successful and therefore warranted in many cases, although it must be carefully considered on a case by case basis both by direct inspection of the liver and possible biopsy at the time of primary resection.

A structured surgical management algorithm for children with POST-TEXT III and IV hepatoblastoma is needed to help maximize survival while minimizing unnecessary long-term morbidity. Expeditious referral to a center with expertise in both pediatric liver transplantation and extreme resection is of paramount importance. Evaluation by a surgeon experienced in both operative options is key to ensuring that 1) transplantation is performed early, when indicated, to avoid the long-term toxicities related to extra cycles of chemotherapy and the poor prognosis inherent in rescue transplant for recurrent tumors; while 2) avoiding the long-term morbidity of transplant in patients who can be safely resected. Within these specialized centers, management should be guided by postchemotherapy imaging while recognizing that the PRETEXT system overstages up to 40% of patients.27 POST-TEXT III tumors that clearly spare at least 1 branch of the portal vein and 1 hepatic vein will almost always prove resectable. At the other extreme, when imaging clearly demonstrates significant multifocal disease in all 4 sectors of the liver, or when there is definite venous obstruction, encasement, and /or invasion of the main portal vein or bifurcation (P2) or the IVC or all 3 hepatic veins (V3), then the tumor is truly unresectable and requires transplantation.16 More commonly, however, there will be uncertainty as to whether the tumor truly involves, or merely abuts, the remaining portal venous branch, hepatic vein, or liver sector. Abdominal exploration is essential for children falling in this gray zone to ultimately determine if the tumor is resectable or requires transplantation.

Our study is limited by the retrospective design and the referral nature of our surgical practice. Preoperative and postoperative care was not standardized, and frequently managed at other institutions. Nonetheless, this series demonstrates that excellent results are possible with extreme resections for hepatoblastoma when performed by an experienced pediatric hepatobiliary surgeon. The findings justify further investigation into the optimal surgical management for children with large and centrally located tumors.

In conclusion, both liver transplantation and aggressive nontransplant resection have important roles in the surgical management of children with hepatoblastoma that involves all 4 sectors of the liver or main vascular structures after neoadjuvant chemotherapy. Referral to institutions with expertise in both pediatric transplantation and hepatobiliary surgery is essential. Preoperative imaging will clearly identify some tumors as unresectable, but the PRETEXT/POST-TEXT system tends to overstage and there is no substitute for operative exploration to make an ultimate determination of resectability. The ongoing challenge for surgeons treating these children is to provide the best oncologic therapy for each individual child while limiting unnecessary long-term morbidity.

CONFLICT OF INTEREST DISCLOSURES

The authors made no disclosures.

Ancillary