Hepatoblastoma (HB) is the most common pediatric primary liver malignancy, accounting for approximately 90% of liver tumors in children (1). Radical surgery remains the cornerstone of treatment for hepatoblastoma (2). The PRE-treatment Tumor EXTension (PRETEXT) staging system designed by the International Childhood Liver Tumor Strategy Group (SIOPEL), combined with detailed annotation factors, can guide the indications for surgical resection and has become an internationally recognized standard (3–5). Previously, liver transplantation for hepatoblastoma included unifocal POST-TEXT IV (PRETEXT classification following chemotherapy) and/or POST-TEXT III or IV tumors with persistent widespread multifocality or major vessel involvement (6). Using neoadjuvant chemotherapy, studies have reported that extensive liver resection for POST-TEXT III and IV hepatoblastoma appears to have a comparable overall survival (OS) rate compared with liver transplantation when combined with chemotherapy (7, 8). Interestingly, Fuchs et al. performed extensive liver resection in 27 patients with POST-TEXT III and IV hepatoblastoma, and the 5-year OS rate was 80.7% (9). This aggressive surgical resection may mitigate the need for orthotopic liver transplantation in selective advanced cases.

Although chemotherapy is beneficial to children with HB, it can also cause toxic side effects and adverse outcomes (10). Furthermore, only a few studies have shown that the PRETEXT staging system tends to overstage patients (11), thus, delaying surgery and causing overtreatment. Additionally, a biopsy is required for a clear pathological diagnosis; however, chemotherapy is occasionally administered to patients whose pathological type cannot be determined (12) and studies have reported that preoperative chemotherapy can transform an originally resectable tumor into an unresectable type (13). Therefore, it is controversial whether preoperative chemotherapy is necessary for patients with HB (14, 15). For the 49 patients with low-risk HB in the AHEP0731 trial, chemotherapy was unnecessary when the tumor was completely resectable during diagnosis. It is worth noting that this included one patient with PRETEXT IV HB and two with PRETEXT III HB, although the authors did not specifically discuss the cases of these three patients (16). Only a few reports have focused on the feasibility of upfront surgical resection in children with PRETEXT stage III and IV HB exist.

Thus, this study retrospectively analyzed the clinical data of children with PRETEXT III and IV HB who underwent surgical resection in our hospital and explored whether preoperative chemotherapy is necessary for such children.

A total of 37 children with HB were included in this study. Patient characteristics are presented in Table 2. The RSG included 18 patients (13 male and five female patients), and the NCG included 19 patients (12 male and seven female patients), with a median age of 11.5 (IQR, 7.75–28.5) and 15.0 (IQR, 12.0–24.0) months, respectively. No statistical difference was observed between the two groups in age, sex, PRETEXT stage, involvement of the inferior vena cava, or AFP level during diagnosis (all p > 0.05). The RSG had a lower incidence of portal vein involvement than the NCG (p = 0.044).

Patients in the RSG received fewer courses of chemotherapy than those in the NCG [5 (IQR, 4.75–5.25) vs. 6 (IQR, 6–8), p < 0.001]. Furthermore, the RSG had lower incidences of febrile neutropenia (61.6% vs. 94.7%, p = 0.019), bleeding (5.6% vs. 84.2%, p < 0.001), and anemia (5.6% vs. 89.5%, p < 0.001) than the NCG. The incidence of gastrointestinal disorders was also lower in the RSG than in the NCG (66.7% vs. 94.7%, p = 0.042). However, no statistically significant differences were observed between the two groups in the incidences of other chemotherapy complications, including electrolyte disorders; ototoxicity; convulsions; and heart, liver, and kidney dysfunction (Table 3).

All patients with HB underwent surgery with R0 resection. No statistical difference between the two groups in the average operation time, intraoperative blood transfusion, severity of Clavien–Dindo complications, or Evans surgical stage was observed (all p > 0.05). A comparison of surgery-related parameters in the two groups is presented in Table 4.

The median follow-up period was 49 months. During the follow-up period, three of the 19 children in the NCG developed recurrence and one developed lung metastasis; no deaths were reported. Two of 18 children in the RSG died, one of whom died 6 months after the recurrence. The estimated 3-year EFS rate in the RSG was similar to that in the NCG (89 ± 0.7% and 79 ± 0.9%, respectively; p = 0.3923; Figure 3).

Preoperative chemotherapy results in good outcomes in children with HB; however, it also increases the risk of chemotherapy complications. Studies have reported that the prognosis of children with PRETEXT I and II HB undergoing surgical resection without preoperative treatment is not worse than that of those who receive preoperative chemotherapy (23). Additionally, upfront surgical resection in patients with PRETEXT I and II HB can reduce the number of courses and complications of chemotherapy (16). However, only a few studies on upfront resections in patients with PRETEXT III or IV HB have been reported. We retrospectively analyzed the treatment outcomes of children with PRETEXT III and IV HB at our center and observed that upfront surgical resection is safe and feasible in these patients, bringing new insights into the timing of surgery for children with PRETEXT III and IV HB.

A previous study observed that the 3-year EFS rate for patients with PRETEXT III and IV HB after neoadjuvant chemotherapy and resection was 75.7% (24). Another study reported that the 3-year EFS rate of patients with PRETEXT IV HB who underwent resection after neoadjuvant chemotherapy was 76% (25). These results are consistent with the 3-year EFS rate of 79% in the NCG of our study. However, the 3-year EFS rate in the RSG was 89%, indicating that upfront surgical resection in patients with PRETEXT III and IV HB may result in improved outcomes.

We observed that upfront surgical resection in patients with PRETEXT III and IV HB significantly reduces the total number of chemotherapy sessions. Multiple toxic effects of chemotherapeutic drugs administered for HB are observed, including myelosuppression, infection, gastrointestinal reactions, liver and kidney toxicity, and ototoxicity (26). Myelosuppression with infection is the most common complication of chemotherapy (16), and it can cause septic shock and death in severe cases. In our study, the incidences of febrile neutropenia, anemia, and bleeding were significantly lower in the RSG than in the NCG. Platinum-based chemotherapy is essential for HB treatment (27), and its most common side effect is ototoxicity (28), which can cause permanent damage in children (29) and is related to the platinum accumulation. In our study, the incidence of ototoxicity was 11.1% and 26.3% in the RSG and NCG, respectively. However, no significant difference in the incidence of ototoxicity between the two groups was observed, probably because of the small sample size. In addition, the incidence of gastrointestinal reactions was also significantly lower in the RSG than in the NCG. Therefore, the advantage of this treatment strategy is that it reduces acute and long-term drug toxicity and medical expenses by lowering the course of chemotherapy and the accumulation of these drugs.

PRETEXT III and IV HB tumors tend to be larger and compress or invade the surrounding blood vessels, causing the tumor margins to be <1 cm away from important blood vessels and making resection more difficult. In this study, most of the tumors in the RSG were large and were <1 cm from the major blood vessels. However, the incidences of intraoperative and postoperative complications were not higher in the RSG than in the NCG, indicating that upfront surgical resection is safe. In a previous study by our group, an obvious fibrous capsule around the HB was observed, which blocked the local tumor invasion; thus, even if the surgical margin is narrow (<1 cm), the tumor can be surgically removed safely and completely (20). In addition, studies have reported that the recurrence rates of tumors with surgical margins <1 cm and those with surgical margins >1 cm were the same (30). When the surgical margin is <1 cm, the patient may have microscopic positive surgical margins; however, our research reported no significant differences in the microscopic positive surgical margin rates between the two groups. Therefore, even if the surgical margin is microscopically positive, it does not affect the prognosis (31, 32).

Our study results demonstrated that upfront resection was associated with a lower incidence of portal vein involvement. Vascular invasion is considered a risk factor affecting the prognosis of children with HB (33–36) and is also an important indicator of resectability (37). Although the latest PRETEXT system considers vascular invasion (4), it tends to overstage tumors (11). Distinguishing between PRETEXT II and PRETEXT III can be especially difficult because the question of tumor compression vs. tumor invasion of the adjacent liver parenchyma exists (4). Hisense CAS 3D reconstruction indicated PRETEXT overstaging in three patients in the RSG, which was confirmed during surgery. Therefore, we suggest that the degree of vascular involvement should be the most important reference factor for assessing the surgical resectability of PRETEXT III and IV HB. However, this does require a computer-assisted surgery system for 3D reconstruction and an experienced surgeon to conduct the preoperative evaluation.

However, our study had a few limitations. First, this study was a retrospective analysis of a small sample from a single medical institution; therefore, a potential selection bias may have been present. Second, our study lacked external verification. Finally, because of the short follow-up period, we could only analyze the EFS at 3 years, although we will continue to follow these patients. We look forward to conducting a large-sample, multicenter, prospective collaborative study in the future to support the conclusions of our research. In summary, this study confirmed that upfront surgical resection is safe and feasible for children with PRETEXT III and IV HB, and it can reduce the course and side effects of chemotherapy. The degree of vascular involvement is the most important consideration when assessing resectability. Our study provides new insights into the timing of surgery for children with PRETEXT III and IV HB.