April 1, 2015 (Vol. 35, No. 7)

Leon Hall Ph.D. Senior Director Taconic Biosciences

Predictive Models Can Reduce Late-Stage Failures and Accelerate Time-to-Market for Effective Drug Candidates

Our ability to understand the complexity of human cancer progression was once limited by research models that did not recapitulate the disease with full accuracy. Those limitations are now behind us. Thanks to advances in immune-deficient and genetically engineered mouse models (GEMs), researchers are able to apply better tools to understand tumor biology, validate in vitro findings, and mimic cancer progression.

Though valuable in many respects, early models used for oncology research present challenges. Immortalized cell lines can have poor predictability in drug research because of their ability to grow under nonphysiological conditions. GEMs with an intact immune system cannot easily imitate the complex genetic profile of some cancers. Syngeneic models use a tumor and host that are of rodent origin, requiring mouse variants of the biologics being tested. Patient-derived tumor xenografts (PDXs) help overcome some of these challenges because the tumor is of human origin and maintains the genetic complexity of the original patient tumor. Yet, the models in which these PDXs have been used initially have limitations as well.

Early cell line and primary tumor xenografts were done in spontaneous mutant mice with compromised immune systems, including nude and scid mice that enabled introducing cancer cells, tracking disease progression, and performing high-throughput screening of molecular compounds. Yet engraftment difficulties for some cell lines and primary tumors, immune system “leakiness” that results in functional T and B cell development over time (confounding studies of slow-growing tumors), slow disease progression in complex tumor types, and radiation sensitivity limited the utility of these models.

Through genetic engineering, we’re now seeing more severe immune-deficient models that are more conducive to immuno-oncology studies. Take the Rag2 model, for example. Because of its targeted mutations of the recombinase activating gene 2 (Rag2) gene, the model does not have an adaptive immune system, does not develop functional cells over time, is tolerant of radiation, and is more accepting of tumor engraftments.

Models with targeted disruption of the interleukin 2 receptor common gamma chain (IL2rg) also have improved PDX use. Because the chain’s disruption helps cripple the most important cell subsets responsible for xenograft rejection (the NK cell), when this mutation is done in a NOD (non-obese diabetic) scid mouse the model is capable of highly reproducible tumor engraftment and growth rates. It also allows efficient engraftment of human hematopoietic stem cells and recapitulates many of the functional aspects of the human immune system. Studies have shown that PDX models can predict patients’ clinical response to treatment, and engraftment of PDXs in the humanized CIEA NOG mouse® has proven efficacious in immune checkpoint inhibition studies.

The future of oncology and immuno-oncology models will likely focus on two vital new developments. First, stable tumor cell lines developed from human patient primary tumors are an option for performing high-throughput ex vivo screens, and subsequent in vivo investigations, using tumors that maintain the original patient tumor’s heterogeneity. Second, primary patient stable cell lines are being developed for use in 3D cultures that can better mimic the differential regional proliferation in patient tumors, primary tumor morphology, tumor exposure to treatment and therefore therapeutic response, and differences in cell type and gene expression reported in solid tumors. 

Finally, as immuno-oncology aims to understand the immune system components that can detect and kill abnormally dividing cells before they develop into tumors, we’re seeing greater interest in finding ways to awaken a patient’s own immune system against the cancer. This can be accomplished using humanized immune system models. Engrafting a PDX or cell line onto a mouse with a humanized immune system yields an immensely powerful tool to measure the efficacy of drugs designed to activate human immune cell responses against the tumor. These predictive models can greatly reduce late-stage failures and accelerate time-to-market for effective drug candidates.

Regardless of the technology used, selection of the host for transplantation is critical, and tumor delivery method is one consideration. Subcutaneous injected cells are limited to forming tumors at the site of injection and provide only a simplified measurement of tumor growth. Orthotopic implantation—delivering cells or tissues directly to the organ of origin—may demonstrate clinically relevant metastatic proliferation and a more relevant tissue micro environment.

Whether the tumor cells are obtained from cell lines or sourced as live tissue from clinical biopsies is another factor. Cell lines tend to grow well in less expensive nude, scid, or Rag2 mice, although lines that are difficult to propagate in vivo may need more immune-deficient mice. Since live tissue can be technically complicated and the primary tissue can be difficult to scale, larger cohorts of mice engrafted with the same tumor burden may be best achieved using a super immune-deficient model with a better tumor uptake rate.

Advancements such as humanized immune system models, PDX models, stable primary patient tumor cells and 3D culturing have shown the potential for major improvements in oncology model predictability of patient response to treatment. These models reproduce variables that impact tumor response to treatment, including tumor phenotypic and genetic heterogeneity. With the development of biologics as viable therapeutics for modulating immune response and leveraging the patient’s immune system to target tumors, we’re also seeing a paradigm shift in cancer treatment, making humanized immune system mice even more valuable tools. Further transgenic modification of NOG mice – such as the expression of human cytokines IL-6 to support multiple myeloma engraftment; IL-2 to enhance NK cell development; or GM-CSF and IL-3 to support myeloid cell lineage development – will provide researchers with increasingly powerful tools to customize models to meet their study needs and greatly enhance oncology and immuno-oncology research. The future of translational oncology research looks promising.


Advancements in murine cell and tissue humanization are enhancing oncology research by improving model predictability of patient response to treatment.

Leon Hall, Ph.D. ([email protected]), is senior director, global scientific development and translational discovery services at Taconic Biosciences.

The GEN Exclusive article "Improved Tools for Improved Immuno-Oncology Research" is original content to the online version of GEN Magazine. For more GEN Exclusives click here.

Previous articleMerck KgaA and Intrexon Collaborate on CAR T-Cell Cancer Therapies
Next articleClimate Change