Prognostic analysis for children with hepatoblastoma with lung metastasis: A single‐center analysis of 98 cases

To analyze the factors affecting the prognosis of hepatoblastoma (HB) with lung metastasis in children.

of HB, neoadjuvant chemotherapy combined with surgical resection increased the cure rate of about 70%, greatly increasing the cure rate of 30% than in the 1970s.
Lung is a common site for HB recurrence, about 10-20% of HB is associated with lung metastasis, and the overall survival rate of these cases is 25-50%. 4,5 Therefore, lung metastasis is still the main factor for poor prognosis of HB. [6][7][8] Currently, the widely accepted treatment scheme for HB with lung metastasis is first received chemotherapy, and lung tumor resection if patients still have tumor with lung metastasis after chemotherapy. 5,9,10 For example, Japan Children's Cancer Group liver tumor committee (JPLT) found that lung metastasectomy is clearly effective for cases of HB with lung metastasis at the time of diagnosis and complete resection of the primary liver lesion after chemotherapy. 11 The prognosis of HB depends on a number of factors, including serum α-fetoprotein (AFP) levels, age at diagnosis, integrity of tumor resection, and clinical stage of the disease. 12 In this study, the treatment outcome and prognosis characteristics of 98 HB patients with lung metastatic in our department were retrospectively analyzed. This study was to explore the effects of prognostic factors, such as the time of diagnosis of lung metastasis, extrapulmonary viscera involvement, standardized treatment, and lung metastasectomy on the prognosis of children with lung metastatic HB, which may provide the guidance for clinical treatment of HB with lung metastasis.

Clinical data
The HB patients with lung metastasis, who were hospitalized at

Evaluation of efficacy and prognosis
Curative effect according to the following criteria 9 : Complete remis-

Effect of standardized treatment on prognosis
Among the 98 patients, 23 patients give up standardized treatment, due to poor treatment resulted in recurrence or metastasis, or the  Figure 2, and there were statistically significant differences between the two groups (P < .001). patients with lung metastasectomy were higher than those without lung metastasectomy, but there was no statistical difference between the two groups (χ 2 = 2.728, P = .099, Figure 3).

Effect of different time to diagnose lung metastasis on prognosis
Among 75 patients treated with standard chemotherapy more than 6 cycles, 31 patients developed lung metastases while on treatment, the survival time of these patients was 71.06 ± 6.36 months, and the 3-year survival rate was 75.5%; while 44 patients had lung metastases at diagnosis, the survival time was 73.28±7.83 months, and the 3-year survival rate was 63.3%. The survival curve of patients with/without lung metastases of HB at diagnosis is shown in Figure 4, there was no statistical difference between the two groups (χ 2 = 0.805, P = .37), which might indicate the time of diagnosis of lung metastasis had no significant effect on prognosis.

Effect of extrapulmonary involvement on prognosis
Forty-six patients with lung metastases alone, of which nine patients died, the median survival time was 90.85±6.87 months, and the 3-year survival rate was 82.3%. Twenty-nine patients with extrapulmonary viscera involvement, of which 16 patients died, the median survival time was 57.99±7.72 months, and the 3-year survival rate was 55.2%.
The survival curves of HB patients with lung metastasis alone or with extrapulmonary involvement are shown in Figure 5, there had statistical difference between the two groups (χ 2 = 7.33, P = .007).

Effect of histology type and PRETEXT stage on prognosis
For 53 (54.08%) epithelial HB patients, the median survival time was 60.25±5.40 months, and the 3-year survival rate was 60%. For

Multivariate Cox proportional hazards model for patients' prognosis
Age, recurrence, and progression with lung metastasis, lung metastasectomy, and extrapulmonary involvement were drawn into multivariate Cox proportional-hazards regression model, which revealed that extrapulmonary involvement (HR = 0.460, 95% CI 0.239-0.888) could be considered as a risk factor contributing to poorer prognosis (Table 2).

DISCUSSION
HB is a malignant embryonal tumor that is the most frequent liver tumor among children, and most tumors develop in children younger than 2 years old. Arora et al 17 found that the age range of 157 patients was from 12 to 24 months. In this study, median age at diagnosis was 22.3 months, and the number of patients whose age was lower than 2 years old was 62 (63.27%), which are similar to the previous literature. 1,18 The lung is the most common site of metastasis, and 20% of children with HB present with lung metastases at the time of diagnosis. 19,20 Moreover, lung is also the most common recurrence site in children with HB. [6][7][8]10 In this study, all patients with lung metastases, 65.31% patients initially diagnosed with lung metastasis, which is far beyond the previous reports. 19,20 In the present study, 34 patients were diagnosed with lung metastasis during treatment (range, 2-20 months; median time 6.5 months). Therefore, in the first diagnosis of HB without distant metastasis, it is still necessary to closely monitor the nail fetus and imaging changes during the treatment process, so as to detect early whether there is lung metastasis, especially within half a year after diagnosis.
There was no consensus on the need for lung metastasectomy in patients with HB. For the treatment of HB lung metastasis, it is necessary to clarify the relationship between the primary and the metastasis. At present, the principle of treatment for HB is to first primary lesion and then metastatic lesion. For the treatment of lung metastases, a small part of them can indeed achieve CR by relying on chemotherapy alone, so most of them will reduce the tumor volume.
Due to most HB primary and metastatic lesions are sensitive to chemotherapy, the volume reduction of tumor mass in liver or lung years old), PRETEXT stage (III or IV), pathology subtypes, serum AFP level (AFP < 100 ng/mL), and metastasis were prognostic factors of HB in children. [29][30][31] In this study, the risk factors affecting the prognosis were also analyzed, such as age, recurrence, and progression after pulmonary metastasis, lung metastasectomy, and extrapulmonary involvement. The results demonstrated that extrapulmonary involvement was the risk factor affecting the prognosis.
However, this study has some limitations. As this study is a retrospective study, selective bias cannot be excluded, and patients whose AFP was lower than 100 ng/mL are few, which all have certain influence on the results. The survival differences between each chemotherapeutic regimen were not analyzed, due to the heterogeneity of tumors, and the sensitivity protocols of each child with tumors were not consistent.
After lung metastasectomy, a long-term follow-up through regular monitoring of lung function and pulmonary imaging reexamination was needed to determine the limitations of respiratory function in resection of metastatic lung tumors and effects on children's growth and development. In the future, a prospective cohort studies with a larger sample size will be conducted to verify the high-risk factors affecting the prognosis of HB children with lung metastasis and the effect of lung metastasectomy on their growth.

CONCLUSION
In general, for HB children with lung metastatic after standardized treatment, the 3-year survival rate of HB patients with lung metastatic alone was significantly longer than that with extrapulmonary involve-ment. Extrapulmonary involvement might be a high-risk factor affecting the prognosis. In children with lung metastases that do not completely disappear after adjuvant chemotherapy, lung metastasectomy might prolong survival and improve prognosis.