Background
As a rare cancer with an incidence of only 1.2–1.5/1000,000 in children, hepatoblastoma (HB) is usually diagnosed in children aged under 3 years [
1]. However, HB is the most common malignant tumor in hepatic cancer in children, accounting for 60% of the primary hepatic cancer in children [
2,
3]. With the continuous improvement in modern medical technology, the disease-free survival rate in low-risk children with HB can reach 80–90% [
4]; however, the prognosis in high-risk pediatric patients, especially those with distant metastasis, remains poor. Therefore, it is vital to understand the risk factors affecting the prognosis of children with HB. In the present study, the clinical data of 316 children aged under 3 years who were diagnosed with HB by pathology between May 2005 and May 2019 in our single center were analyzed, and the curative effect of multidisciplinary combined therapy and the risk factors affecting the prognosis were investigated, with the aim of providing the corresponding intensive treatment for children with HB with different risk stratification and improving the prognosis.
Discussion
In recent years, international cooperation in the treatment of HB has made great progress through cooperative multicenter research [
9]. However, distant metastasis occurs in most cases at the time of diagnosis of HB, which seriously reduces the post-therapeutic survival and prognosis in pediatric patients with HB. Therefore, it is essential to understand the relevant risk factors affecting the prognosis of HB in children, as this can guide the corresponding stratified treatment in clinical practice and achieve better clinical efficacy.
It has been reported that the 5-year OS of pediatric patients with HB can reach 75%, and the 5-year EFS has reached approximately 65% [
10]. In the present study, children aged under 3 years (median age of 1.45 years old) with HB were studied with a median follow-up duration of 62 months. The results showed that the clinical therapeutic efficacy was 81.0%, and the 5-year OS and EFS were 79.8 and 75.1%, respectively, which were higher than the percentages reported internationally. In addition, this study also analyzed the survival rate of older HB children over 3 years old in the same period, and found that HB children under 3 years old had a better prognosis. Maibach et al. [
11] believed that the older the diagnosis age of children with HB, the more obvious the adverse trend that affects their prognosis might be. The recent study confirmed this (
p = 0.039). However, multivariate analysis indicated that age was not an independent risk factor for the prognosis of HB, which might be due to the fact that the present study focused on children aged under 3 years. This result can be verified in the future by expanding the age range of samples in the database. It is worth noting that in the present study, there were five cases of fetal onset of HB during pregnancy, and three cases were born with low birth weight. Maruyama et al. [
12] investigated 15 children with low birth weight who were found to have fetal onset of HB during pregnancy and diagnosed with HB after birth, and suggested that children with low birth weight, especially those with very low birth weight, were prone to develop HB in the future, which might be another risk factor for HB.
In the present study, statistical analysis of the epidemiology of HB in China was conducted. Although there was no statistical difference in the survival of HB among different regions (
p = 0.869), the analysis revealed that the incidence of HB in the coastal areas was higher than that in the inland areas. A reason for this might be that the industrial development in coastal areas was more advanced than that in the mainland, the air and water pollution were relatively serious, and the diet structure was mainly composed of seafood and bacon. In addition, low birth weight and parental tobacco using prior to or during pregnancy may increase risk of hepatoblastoma [
13]. However, the results of the present study were limited to the statistical results of our single center. The incidence of HB might also be affected by many factors, such as medical resources and economic development in different regions. As our single center is located in North China, most of the patients came from the local region. It is expected that our center will cooperate with other centers in the future to conduct multi-center, large-scale epidemiological investigations.
As a tumor marker of HB, serum AFP not only has important prognostic significance in the initial diagnosis, but is also one of the important therapeutic indicators during the treatment [
14]. In the present study, the prognosis of children with an AFP level < 100 ng/mL was worse than that of other groups (
p < 0.05), which was an independent risk factor affecting the prognosis of HB. The results of Piotr Czaudernaa and Rebecka L Meyers et al. [
15,
16] were consistent with those of the present study. In 2015, a study [
17] showed that the increase in platelet count in the peripheral blood was correlated with infection, inflammatory disease, malignant tumor, and some chronic myeloid diseases. During the clinical practice, we found that some cases, especially those with stage IV, were prone to demonstrate an abnormal increase in platelet count at the initial diagnosis. The present study also confirmed that the increase in platelet count was another independent factor affecting the prognosis of HB.
It has been reported [
18] that the prognosis of patients with HB of the mixed type is worse than that of patients with HB of the epithelial type, with the fetal type having the best prognosis, and that of the small-cell undifferentiated type being relatively poor. In the case analysis in the present study, it was found that different pathological types were correlated with the survival of pediatric patients. All four children whose pathologic type were small cell undifferentiated type eventually died with poor prognosis, which is consistent with the literature. However, the pathological type was not an independent risk factor for the prognosis. At present, SIOPEL believes that the PRETEXT in children with HB has an important prognostic value [
11], and the advantage of the PRETEXT over postoperative staging such as the COG stage is that it is applicable to all children with HB, especially those without surgery, and it can predict the resectability of the tumor to a certain extent, while complete resection of the liver tumor is the key to the treatment of HB in children. The present study demonstrated that the OS of children with PRETEXT IV was significantly lower than that of pediatric patients with other stages. It also confirmed that the PRETEXT was significantly correlated with the prognosis.
Distant metastasis is prone to occur in HB and is an independent risk factor for the prognosis. The lung is the most common site of metastasis, most of which is from blood transmission. The tumor cells will stay to form metastasis as they reach the end of pulmonary vessels. Therefore, pulmonary metastasis usually occurs at the edge of the supply area at the end of the blood vessels of the bilateral lungs [
19]. Of the 132 cases with distant metastasis in the present study, lung metastasis accounted for 80.3%. Moreover, single lung metastasis was more common than bilateral lung metastasis, and single marginal lung metastasis was more common than central lung metastasis, which is consistent with the literature.
Conclusion
According to the results of multivariate analysis of the COX regression model, in addition to AFP level, platelet count, PRETEXT and distant metastasis, vascular invasion was an independent risk factor for the prognosis of HB in children. However, it should be noted that whether there was tumor rupture, infiltration of the extrahepatic adjacent tissues and organs, and multiple lesions in the liver, and whether the tumor can be completely resected had no clear effect on the prognosis (
p > 0.05), which was not completely consistent with the relevant reports in the literature [
20,
21]. It was speculated that the main reason for these differences might be that most of the children treated for HB in our center were in a late stage or refractory stages, which to some extent covers up the influence of these risk factors on the survival. Other studies have suggested that maternal hypertension during pregnancy, excessive amniotic fluid, smoking history, and birth weight < 1500 g might increase the incidence of HB [
22]. Although these factors were not investigated in the present study, it is expected that clinical HB data of more centers will be collected in the future to establish a more scientific and perfect risk stratification system. Thus, the individualized and standardized treatment may be realized, and the survival and long-term prognosis of HB in children may be further improved, which will help in the rehabilitation of HB.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.