Independent Assessment of the Children's Hepatic Tumors International Collaboration Risk Stratification for Hepatoblastoma and the Association of Tumor Histological Characteristics With Prognosis.

Importance: Hepatoblastoma is the most common pediatric liver malignant neoplasm, and accurate risk stratification is essential for guiding treatment. Objective: To validate the Children's Hepatic Tumors International Collaboration-Hepatoblastoma Stratification (CHIC-HS) in an independent cohort of patients with hepatoblastoma and evaluate the association of pretreatment hepatoblastoma histological subtype with prognosis. Design, Setting, and Participants: This is a single-institution retrospective cohort study of 96 pediatric patients with hepatoblastoma diagnosed and treated between June 1, 2000, and December 31, 2016, with recent therapy and independent of the CHIC-HS discovery cohort. Each patient was assigned a risk group according to CHIC-HS. The histological characteristics of each tumor were assessed based on the International Pediatric Liver Tumor Consensus Classification. Data were analyzed from May 2018 to May 2019. Main Outcomes And Measures: The main outcomes were event-free survival (EFS) and overall survival (OS). Cox regression analysis was used to examine the associations of patient characteristics and tumor histological characteristics with survival. Results: A total of 96 patients (median [range] age, 1.9 [0.4-18] years; 36 [38%] girls and 60 [63%] boys) were assessed, including 15 with very low risk, 28 with low risk, 23 with intermediate risk, and 30 with high risk, according to CHIC-HS criteria. There were a total of 13 cancer-related deaths; median (range) follow-up was 3.5 (0.1-17.8) years for those alive at the last follow-up. The estimated 5-year OS rates were 100% in the very low-risk group, 94.7% (95% CI, 68.1%-99.2%) in the low-risk group, 89.2% (95% CI, 63.1%-97.2%) in the intermediate-risk group, and 57.9% (95% CI, 34.6%-75.5%) in the high-risk group. In a multivariable analysis, we confirmed that CHIC-HS significantly estimated EFS (high-risk group vs very low- and low-risk groups: hazard ratio [HR], 45.59; 95% CI, 9.39-209.5; P < .001) and OS (high-risk group vs very low- and low-risk groups: HR, 21.95; 95% CI, 2.76-174.29; P < .001). In the subcohort of 84 patients for whom pretreatment tumor histological data were available, tumor epithelial histological subtypes were found to be significantly associated with both EFS and OS. Patients in the CHIC-HS high-risk group and with embryonal-only histological subtype had the highest risk of relapse or disease progression (high-risk: HR, 42.62; 95% CI, 9.91-203.9; embryonal: HR, 3.28; 95% CI, 1.21-8.9) and death (high-risk: HR, 18.78; 95% CI, 2.31-152.84; embryonal: HR, 7.12; 95% CI, 1.51-33.52). Conclusions and Relevance: This cohort study found that CHIC-HS performed as expected in an independent cohort that was more recently treated. Incorporation of pretreatment tumor histological data into CHIC-HS may provide additional prognostic value.