The evolution of preclinical models for myelodysplastic neoplasms.

ABSTRACT

Myelodysplastic Neoplasms (MDS) are a group of clonal disorders characterized by ineffective hematopoiesis and morphologic dysplasia. Clinical manifestations of MDS vary widely and are dictated in large part by a range of genetic aberrations. The lack of robust in vitro models for MDS has limited the ability to conduct high throughput drug screens, which in turn has hampered the development of novel therapies for MDS. There are very few well-characterized MDS cell lines, and the available cell lines expand poorly in vitro. Conventional xenograft mouse models can provide an in vivo vessel to provide growth of cancer cells, but human MDS cells engraft poorly. Three-dimensional (3D) scaffold models that form human "ossicles" represent a promising new approach and can reproduce the intricate communication between hematopoietic stem and progenitor cells and their environment. Genetically engineered mice utilize specific mutations and may not represent the entire array of human MDS; however, genetically engineered mice provided in vivo proof of principle for novel agents such as luspatercept, demonstrating the clinical utility of this approach. This review offers an overview of available preclinical MDS models and potential approaches to accelerate accurate clinical translation.