Impact of treatment on the prognosis of childhood in hepatoblastoma: A SEER based analysis

Abstract

Background: The prognosis of patients with hepatoblastoma has been unsatisfactory. This study analyzed the effects of different treatment methods on cancer-specific survival (CSS) in children with hepatoblastoma.

Method: From 2000 to 2018, patients with hepatoblastoma were included in the Surveillance, Epidemiology, and End Results (SEER) database. CSS was estimated using the Kaplan-Meier method. Cox regression analysis assessed prognostic factors. The predictive models were validated using the concordance index (C-index), calibration curve and receiver operating characteristic (ROC) curve.

Result: Of the 785 included patients, 730 (93.0 %) underwent chemotherapy, 516 (65.7 %) underwent liver tumour resection and 129 (16.4 %) underwent liver transplantation. Both chemotherapy and surgery could significantly improve the CSS rate (all p < 0.001). However, there was no difference in CSS rate between the two surgical methods (liver tumour resection and liver transplantation) (p = 0.613). Further subgroup analysis revealed that children who underwent liver tumour resection or liver transplantation based on chemotherapy (all p > 0.05) had a similar prognosis. Multivariate analysis revealed that age (p = 0.003), race (p = 0.001), operative method (p < 0.001), chemotherapy (p < 0.001), distant metastasis (p < 0.001) and tumour size (p < 0.001) were independent factors related to CSS. The C-index of the new nomogram was 0.759, and its consistency was good. The ROC curves verified that the nomogram had a better prediction ability for 1-, 3- and 5-year CSS rates.

Conclusion: In children with hepatoblastoma, there was no statistically significant difference in CSS between chemotherapy combined with liver transplantation and liver tumour resection. The nomogram we constructed demonstrated satisfactory CSS prediction ability.

Keywords: Cancer-specific survival; Chemotherapy; Hepatoblastoma; SEER; Surgery.

Fig. 1

Flow chart of the overall study design.

Fig. 2

Survival curves for the CSS of patients with hepatoblastoma with different treatments. K–M curves for CSS with (A) chemotherapy vs. no chemotherapy groups and (B) surgery vs. no surgery groups. The number of at-risk cases in each group at 0, 50, 100, 150 and 200 months is indicated. Abbreviations: CSS: Cancer-specific survival; K–M: Kaplan–Meier.

Fig. 3

Subgroup analysis of CSS between patients with hepatoblastoma with chemotherapy combined with hepatectomy and liver transplantation, hazard ratio ±95 % confidence interval. Abbreviations: CSS: Cancer-specific survival; CI: Confidence interval; HR: Hazard ratio; AFP: Alpha-fetoprotein.

Fig. 4

The 1-, 3- and 5-year CSS probability is calculated by determining the sum of the risk points, which are based on Cox proportional hazards regression analysis. For each parameter, the number of associated risk points can be determined by drawing a vertical line from each variable axis upwards to the points axis (0–100). The total score projected on the bottom scale represents the probability of CSS rates of 1-, 3- and 5-years. Abbreviations: CSS: Cancer-specific survival.

Fig. 5

Development and performance of the nomogram. Calibration curves to assess the 1- (A), 3- (B) and 5- (C) year CSS rates. AUC values to predict CSS rates at 1- (D), 3- (E) and 5- (F) years. Abbreviations: CSS: Cancer-specific survival; AUC: Area under the curve.