NCI-funded Cancer Model Development Centers (CMDCs) and the Human Cancer Models Initiative (HCMI) consortium members are creating as many as 1000 new cancer models.  The models are being derived from many tumor subtypes, including rare and understudied cancers. Normal tissue, originating tumor, and the tumor-derived models will be sequenced and this information, as well as patient-pertinent clinical data, will be available to researchers through the Genomic Data Commons (NCI) and European Bioinformatics Institute (European consortium members) in a way that protects patient privacy (Figure). The clinical information will allow correlation of genetic features and clinical characteristics.  The models will be available to researchers through the non-profit American Type Culture Collection (ATCC) repository.

HCMI-developed models and related clinical and molecular data will be available to cancer researchers as a community resource in an effort to advance cancer research and more fully understand how in vitro findings are related to clinical biology.

               flow chart depicting the human cancer model development process   

Figure: HCMI Process.  HCMI validated models and their associated genomic and clinical data will be available to the research community. The models will be generated by the Development Centers and validated by molecular characterization data that verify case matching by genotype and that the model originated from the primary tumor. Clinical data (collected by the Development Centers) and molecular characterization data (generated by the Genome Characterization Centers) will be included with all models. The models which pass the validation criteria will be available through a Distributor. Molecular and clinical data will be available through the Genomic Data Commons (GDC) for US-generated models, the European Bioinformatics Institute (EBI) for UK- and Dutch- generated models, and the HCMI websites (Databases). The multiple databases are required in accordance with the regulations of the countries participating in HCMI. All materials will be shared in a way that protects patient privacy. The shapes in the flowchart indicate different steps. Rectangle = Process, Trapezoid = Input/Output, Diamond = Decision, Line with arrow = Flow Line. Download PDF iconHCMI_Process.pdf 


Many of the cancer lines that are commonly used in cancer research were established decades ago and have not always been propagated under standardized culture conditions. These cell lines have been useful for in vitro experiments to understand cancer biology, biochemistry, and drug targets. However, drawing conclusions about how in vitro observations may relate to clinical biology is challenging for the following reasons:

  • Cancer cell lines often lack the cellular complexity and architecture of human tumors.
  • Cancer cell lines are not associated with clinical information from patients.
  • The cell lines’ genetic relatedness to the original tumors are unknown, and molecular characterization including assessment of genomes and transcriptomes of the model, until recently, were mostly not available.
  • Cancer cell lines have been propagated with limited knowledge of the genetic information from the originating tumor. This introduces the possibility that genetic drift may have occurred after the model was established.
  • The process of immortalization may contribute unexpected genetic and biological alterations to cell lines. 
  • Rare cancers, some common cancers, and cancer subtypes that are common to various ethnic groups, are rarely represented in currently available cell lines.

HCMI will be addressing the deficiencies in current models by collecting patient clinical data and assessing the genomes and transcriptomes of the original tumors, case-matched normal tissue, and the resulting model and providing this information to the end-user. In order to address the issue of tumor complexity, the HCMI will use next-generation culture techniques including organoids, conditionally reprogrammed cells, or others, to develop models that more closely mirror the architecture and cellular complexity of human tumors.

Next-generation Models

Next-generation cell culture methods, such as organoid and conditionally reprogrammed cell (CRC) models introduce the opportunity to propagate primary normal and cancer cells so that they retain their original genotypes and phenotypes for many passages. This characteristic makes them preferable to traditional immortalized or alternatively short-lived primary cell lines.

The term organoid has been used previously; here it refers to a three-dimensional structure grown from stem cells in vitro that comprises organ-specific cell types. Organoid culture methods were first developed using stem cells from the mouse small intestine. Organoids often consist of two or more cell types and develop structures that reflect the structural and functional properties of the organs from which they originate, but without innervation, blood supply, or stroma. These three-dimensional structures are grown from epithelial stem cells and are amenable to expansion when cultured under the appropriate conditions. Cultures are grown in extracellular matrix using media that contain components that drive cell proliferation and differentiation.

CRC methodology was pioneered using human keratinocytes and subsequently expanded to other epithelial types. Propagation of primary human cells using the CRC methodology requires Rho kinase inhibitor and irradiated mouse fibroblast ‘feeder’ cells. These additions to the culture result in immortalization through cytoskeletal alterations and the activation of endogenous telomerase. CRCs proliferate as stem-like cells but maintain the ability to differentiate when presented with the correct conditions. Although they are typically cultured as monolayers, CRCs can also develop into organ-like three-dimensional structures when cultured with a matrix.

The organoid and CRC methods have been used to successfully grow human normal tissues from multiple different organs and have been applied, with some modifications, to grow tumor tissues.  Established tumor models propagate for many months, are representative of the original tumor (i.e. have shared genetic alterations), and are amenable to high-throughput screening and other experimental manipulations. In addition, methodologies have been developed for the growth of organoids using the air-liquid interface culturing method that enable the co-culture of tumor and associated stroma.

Culture models are constantly evolving. In an effort to reduce dependence on ‘feeder’ cells for growth of primary tumors, methodologies have been developed that use either specialized and highly defined media or alternative culturing procedures which include the use of collagen and keratinocyte serum-free medium. The goal of the HCMI is not only to contribute models to the research community, but also further the burgeoning understanding of the development of next-generation cancer cell models and provide this information to the public.

Next-generation cancer model systems present a unique opportunity for the scientific community to study individual human tumors in vitro to further advance knowledge of subjects such as cancer biology, biochemistry, and exposure to perturbagens. The next-generation cancer models that result from the efforts of the HCMI will serve as excellent tools for the support of personalized medicine research and give insight into the pathways that influence tumor progression.

Tumor Types

Through this initiative, we are planning to generate models from many of the human cancer subtypes. The list of available tumor types will be updated as models are successfully completed. 

Last updated: May 08, 2018