Treatment outcome of Paediatric Hepatoblastoma in Hong Kong

Abstract

Background and aims: The Hong Kong Paediatric Haematology and Oncology Study Group had formulated a consensus protocol (HB/HCC 1996) based on results published by the Children’s Oncology Group (COG) for treatment of paediatric hepatoblastoma (HB). We review the treatment outcome of HB in Hong Kong over the past 2 decades and investigate the role of PRETEXT staging and risk stratification. Methods: Review of population-based paediatric oncology database for patients (<18 years) diagnosed with HB and treated with the HB/HCC 1996 protocol between July 1996 and June 2014. Central radiological review and retrospective assignment of PRETEXT staging and SIOPEL risk group (2005 version) were performed, whenever possible. Overall survival (OS), event-free survival (EFS) and predictors of outcome were determined as of 31 August 2016 or last follow-up date. Results: Sixty patients were enrolled during the period, representing an annual incidence of 0.29 per 100,000. Thirty-four (57%) were male. Median age at diagnosis was 1.1 year (range, 0.08-14.82) and median follow-up time was 6.8 years (range, 0.002-19.6). Abdominal distension was the commonest presenting symptom. Eight (13%) patients had tumour rupture. Alpha-fetoprotein (AFP) was raised (>100ng/ml) in 58 (97%) patients. Epithelial type was the commonest subtype occurring in 39 (65%) patients. Five (8%) had lung metastases at presentation. Treatment details were not available in 3 patients and 1 patient died of tumor rupture before treatment. Twenty-nine patients (48%) received treatment with first-line chemotherapy; 23 (38%) required alternative agents. Three patients did not require chemotherapy after surgery; 1 died of liver failure post-operatively before initiation of adjuvant treatment. Surgical resection could be performed in 48 (80%) patients whilst 8 patients had unresectable (+/- metastatic) disease among which 3 (5%) underwent upfront liver transplantation. There were 14 deaths (disease relapse in 6; disease progression in 5 and tumor rupture in 3) and 11 relapses (native liver in 6; transplanted liver in 1; lung in 3 and bone in 1). The 5-year OS and EFS rates were 77.6% (±5.5%) and 69.2% (± 6.1%), respectively. Predictors of inferior outcome included advanced COG staging, unresectable or bilobar disease, tumour rupture, low AFP and suboptimal response to first-line chemotherapy. Long term sequalae included hearing loss (Grade 3 in 4) and renal tubular dysfunction (Grade 2 in 2). Radiological review for PRETEXT staging could be performed in 29 of 60 subjects yielding PRETEXT I in 3, II in 8, III in 14 and IV in 4. Eighteen patients belonged to standard risk (SR) and 11 to high risk (HR) according to the SIOPEL risk criteria. 5y OS and EFS rates were significantly better in the SR patients than the HR patients (89.9%±6.8% vs 42.2%±15.6% [p=0.017]; 84.7%±8.2% vs 13.6%±11.7% [p<0.001], respectively). Conclusions: Although treatment with the HB/HCC 1996 protocol resulted in cure of about three-quarters of HB patients, it is clear that we need to identify high risk patients upfront by a robust risk stratification system and improve their treatment outcome by intensifying chemotherapy regimen and consolidating surgical treatment plan (including selection criteria for total hepatectomy and liver transplantation).