PurposeWe aimed to assess the clinical utility of the mini patient-derived xenograft (MiniPDX) model in screening individualized chemotherapy regimens for pediatric hepatoblastoma.Materials and methodsWe included 31 children with hepatoblastoma who had unsatisfactory decreases in alpha-fetoprotein levels during neoadjuvant chemotherapy or poor clinical control of recurrence with or without metastasis. We established a MiniPDX model using surgically resected tumor tissue specimens. The sensitivities of five chemotherapeutic regimens were tested to determine the one with the lowest tumor proliferation rate, which was then set as the experimental group. We compared the clinical characteristics and efficacy with those of conventional chemotherapy regimens.ResultsThe median follow-up period for the experimental group was 27 months, with a complete remission (CR) rate of 80.64%. Among stage IV cases, there was a significant between-group difference in CR rate (experimental [73.68%] vs. control [37.5%]) and 3-year event-free survival rate (79.3% vs. 26.7%). The most effective individualized chemotherapy regimens were ifosfamide + pirarubicin + etoposide + carboplatin (54.84%), followed by pirubicin + cyclophosphamide + cisplatin (16.13%), ifosfamide + carboplatin + etoposide (12.90%), cisplatin + 5-fluorouracil + vincristine + adriamycin (12.90%), and vincristine + irinotecan + cyclophosphamide + cisplatin (3.23%).ConclusionUsing the MiniPDX model to screen individualized chemotherapy regimens for pediatric hepatoblastoma can significantly improve the CR rate.