Abstract A02: Patient-derived xenograft models of neuroblastoma as improvement for the prediction of targeted therapies for childhood cancer

Abstract

Abstract Introduction: Highly variable clinical outcomes including immediate fatal progression to spontaneous regression are characteristically observed for neuroblastoma (NB), a pediatric cancer arising from progenitors of the sympathetic nervous system. Pronounced rates of relapse and fast-evolving resistance against standard-of-care (SOC) treatments, together with a high degree of mutational mobility and heterogeneity, indicate the need for more precise and molecular targeted strategies. Aim: Patient-derived xenograft (PDX) models closely resemble human tumor biology from primary tissue and could be utilized for predictive preclinical testing. For this aim we established a cohort of 10 NB PDX models from 33 primary human tissue samples obtained from the Charité University Clinic. Specimen originated from proliferating areas while possible necrotic lesions had been removed prior to subcutaneous transplantation to immunodeficient NOG mice. After documenting stable growth over 3-4 initial passages, NB PDX models were screened for sensitivity to SOC drugs including actinomycin D, dacarbazine, docetaxel, doxorubicin, etoposide, ifosfamide, gemcitabine, and vincristine. Results: Characteristic NB pathology of all PDX was verified using HE and IHC staining of FFPE tumor sections comparing individual passages. Post-transplant lymphoproliferative disorder (PTLD) was excluded analyzing hCD45 specific marker. Depending on traceable somatic mutations, NB PDX doubling times dramatically varied between 4 days in aggressive, 8 days in intermediate, and up to 18 days in slowly growing tumors. Not surprisingly, high-risk NB tumors with confirmed MYCN amplification and ALK mutation exhibited the fastest growth in our cohort. In contrast, doubling rates of tumors with single ALK mutation or TERT rearrangement were not significantly different compared to the slowest PDX models without driver mutations. Most effective treatment response was observed for vincristine, leading to stable disease or partial regression in 5 out of 8 models (63%) achieving a tumor growth inhibition (TGI) of 70%, whereas all models were insensitive against dacarbazine treatment with a TGI of 33%, respectively. Conclusion: Currently, whole-exome sequencing of all models is performed. Molecular analysis will enable prediction of more precise therapeutic strategies. We will use our NB PDX to validate targeted treatment effects of recent pipeline drugs. As part of the European Innovative Therapies for Children with Cancer-Paediatric Preclinical Proof of Concept Platform (ITCC-P4) consortium, these findings will be correlated with predictions of the NB PDX molecular profile to achieve valuable benefit for pediatric patients. Citation Format: Dennis Gürgen, Jana Rolff, Johannes H. Schulte, Hedwig E. Deubzer, Karin Schmelz, Anton G. Henssen, Patrick Hundsdörfer, Georg Seifert, Angelika Eggert, Wolfgang Walther, Jens Hoffmann. Patient-derived xenograft models of neuroblastoma as improvement for the prediction of targeted therapies for childhood cancer [abstract]. In: Proceedings of the AACR Special Conference on the Evolving Landscape of Cancer Modeling; 2020 Mar 2-5; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2020;80(11 Suppl):Abstract nr A02.