Molecular Characterization of Patient-Derived Human Pancreatic Tumor Xenograft Models for Preclinical and Translational Development of Cancer Therapeutics

Preclinical evaluation of novel cancer agents requires models that accurately reflect the biology and molecular characteristics of human tumors. Molecular profiles of eight pancreatic ductal adenocarcinoma patient tumors were compared to corresponding passages of xenografts obtained by grafting tumor fragments into immunocompromised mice. Molecular characterization was performed by copy number analysis, gene expression and microRNA microarrays, mutation analysis, short tandem repeat (STR) profiling, and immunohistochemistry. Xenografts were found to be highly representative of their respective tumors, with a high degree of genetic stability observed by STR profiling and mutation analysis. Copy number variation (CNV) profiles of early and late xenograft passages were similar, with recurrent losses on chromosomes 1p, 3p, 4q, 6, 8p, 9, 10, 11q, 12p, 15q, 17, 18, 20p, and 21 and gains on 1q, 5p, 8q, 11q, 12q, 13q, 19q, and 20q. Pearson correlations of gene expression profiles of tumors and xenograft passages were above 0.88 for all models. Gene expression patterns between early and late passage xenografts were highly stable for each individual model. Changes observed in xenograft passages largely corresponded to human stromal compartment genes and inflammatory processes. While some differences exist between the primary tumors and corresponding xenografts, the molecular profiles remain stable after extensive passaging. Evidence for stability in molecular characteristics after several rounds of passaging lends confidence to clinical relevance and allows for expansion of models to generate the requisite number of animals required for cohorts used in drug screening and development studies.

Abstract licence: CC BY-NC

• Application of Highly Immunocompromised Mice for the Establishment of Patient-Derived Xenograft (PDX) Models · 2019

• Application of Cancer Organoid Model for Drug Screening and Personalized Therapy · 2019

A person counts once a day per paper. We keep no cookie and no address: a reader is a hash that changes every day. Any month under five reads shows no number.