Abstract 2884: Establishment of patient-derived xenograft models for high-risk neuroblastoma

Background: High-risk neuroblastoma (HR-NB) accounts for 15% of all pediatric oncology deaths, with long-term survival approximately 50%. Predictive preclinical models play an important role in the assessment of new treatment strategies and as avatar models for personalized medicine. We investigated the feasibility of establishing patient-derived xenograft (PDX) models of HR-NB from a range of tumor-bearing patient materials at diagnosis and relapse, and assessed techniques that may improve engraftment times and success rates. Methods and Results: Tumor materials were obtained from 12 HR-NB patients, including primary and metastatic solid tumor biopsies, bone marrow mononuclear cells, pleural fluid and residual cells from cytogenetic analysis. Samples were processed and implanted subcutaneously into immunocompromised mice. The engraftment success rates were 33% (4/12) for diagnosis samples and 100% (6/6) for relapse samples. Median engraftment time to 1000mm3 tumor was 93.5 days (range 33–210 days). Engraftment at diagnosis was associated with poor outcome. PDX models were validated against primary material by histology, short-tandem repeat analysis, and single nucleotide polymorphism (SNP) array. PDXs matched the donor patient specimen in all but one case, where attempted engraftment from pleural fluid resulted in EBV-associated lymphoid proliferation. Attempted engraftment of tumor derived from a lymph node metastasis was associated with human T-lymphocyte infiltration of the mouse liver and spleen. Matched PDX models established directly from tumor biopsy and via short-term cytogenetics culture were equally similar to the donor tumor by SNP array and histological analysis. We directly compared engraftment rates for three different patient tumors at subcutaneous, intramuscular and orthotopic implantation sites. For two patient tumors, orthotopic engraftment was significantly faster than other sites, while the third model is ongoing. Conclusion: HR-NB PDX models can be established from diverse sample types, allowing us to develop a preclinical testing platform for new therapies and the capacity to model personalized therapy. Engraftment of relapse samples can be readily achieved, while engraftment at diagnosis may portend poor prognosis. The engraftment time of neuroblastoma patient samples might be accelerated by orthotopic implantation. Citation Format: Alvin Kamili, Andrew J. Gifford, Nancy Li, Chelsea Mayoh, Georgina L. Eden, Jinhan Xie, Robyn E. Lukeis, Murray D. Norris, Michelle Haber, Geoffrey McCowage, Toby N. Trahair, Jamie I. Fletcher. Establishment of patient-derived xenograft models for high-risk neuroblastoma [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2019; 2019 Mar 29-Apr 3; Atlanta, GA. Philadelphia (PA): AACR; Cancer Res 2019;79(13 Suppl):Abstract nr 2884.