Patient-derived Human Xenografts: Their Pros and Cons as Cancer Models

Apr 21, 2016 | Disease Models

A xenograft (or heterograft) is a piece of living tissue taken from a donor of one species and grafted into a recipient of different species. Cancer refers to a group of diseases characterized by uncontrolled division of abnormal cells in a part of the body. According to WHO’s (World Health Organization) site, cancer is one of the leading causes of morbidity (rate of illness) and mortality (rate of death) worldwide, with approximately 14 million new cases and 8.2 million cancer-related deaths in 2012. With the number of annual cases set to rise to 22 million within the next 2 decades, the search for new drugs is both vital and urgent.

According to the FDA, the rate of success for a medicinal compound entering phase I testing actually reaching the market is only 8%. These drugs would have shown signs of efficacy in in vitro and/or in animal models and yet the culmination of perhaps a decade of pre-screening and evaluation results in another failure to translate into success in human studies. While the reasons for this are complex, one major factor is the need for cancer models that better represent the native tumor microenvironment and thus respond to potential cancer drugs in a manner more representative of a human response.

In the search for the best cancer drug models, one type of model of particular note are patient-derived tumor xenograft (or PDTX) mice. Here, we’ll discuss the pros and cons of PDTX mice as cancer models.

Patient-Derived Human Xenografts

PDTX models or “xeno-patients” are made by implanting cancerous tissue from a human primary tumor into an immunodeficient mouse. Human tumor cells may be transplanted into a mouse model by a number of methods including under the skin (subcutaneously), into the abdominal cavity (intraperitoneally), and into the organ of origin (orthotopically). Once the tumor has grown, it can be removed, divided and implanted into more mice allowing researchers to create a colony. Two commonly used immunocompromised mouse strains are nude and NOD/SCID mice.

Immunocompromised mice are ideal for implanting xenogeneic tumors into as they are less able to reject them. They also allow you to simulate the complex microenvironment a tumor would normally grow in including a nutrient- and oxygen-rich blood supply capable of removing toxins, the extracellular matrix, the presence of other cell types and of growth factors. It also gives the tumor the ability to promote angiogenesis and metastasize, a feat that cannot yet be adequately models outside of a living organism.

Jargon Alert!

Nude mice: a hairless mouse strain that is athymic resulting in a lack of T cells. They also only have a partial B cell response. These mice must be homozygotic for the nu gene meaning two copies of the gene must be present to be useful as PDTX models.

NOD/SCID mice: severe combined immunodeficient mice lack functional B and T cells if they are homozygous for the SCID. It is common for SCID mice to be crossed with non-obese diabetic (NOD) mice, which lack natural killer or NK cells to further reduce their immune capabilities. They may also be crossed with a mouse strain with a defective interleukin-2 receptor γ-chain gene (IL2rg), also called the common γ-chain gene (γc) to create NOD-scid-γc or an NSG mouse. IL2RƔnull mice can also be used to create a NOD-scid-IL-2RƔ or NOG mice. NSG and NOG mice not only have defective B, T and NK cells but other innate immune cells are also dysfunctional.

These mice are more extensively reviewed by Goyama et al..

Let’s go over some of the pros and cons of PDTX models.

  • Intraperitoneal and orthotopic injections allow for the study of metastatic dissemination and subcutaneously injected cells create ideal mice for the study of early localised disease at the site of injection.
  • With the formation of biobanks of patient tumor samples, the problem of lack of tissue is becoming less of an issue.
  • Missing elements of the immune system can be added to the mice to create humanized PDTX models allowing researchers to study how each cell type plays a role in the cancer subtype their studying.  This is carried out by injection of peripheral blood or bone marrow cells, allow for an almost complete reconstitution of the immune response to the tumor. One popular combination of immune system-creating tissues is called BLT or human bone marrow (BM), liver, and thymus tissues that can be engrafted.
  • This model allows you to recreate the heterogeneity of tumor cells seen in tumors in humans, including the genetic and epigenetics variations.
  • In nude or SCID mice without a reconstituted immune system, tumorigenetic formation is fairly rapid. It is usually detectable within weeks and full tumor development usually occurs within 1-4 months.
  • By placing the tumor in various parts of the mouse, the tumor microenvironment can be manipulated to create the closest representation of the true tumor microenvironment when in humans.
  • The stroma can also be added again facilitating the creations of a truer tumor microenvironment.
  • Multiple therapies can be tested on the same tumor biopsy.
  • Aside from the advantages created by using this model in drug discovery, it can also be used to aid in the development of individualized molecular therapeutic approaches in a clinical setting.
  • You can now purchase immunodeficient mice with human tumors.


  • Overall in this area of research, there is a lack of standardization when it comes to which mouse and which engraftment procedure is the most optimal each cancer subtype. There are several reasons for this including:
    • The variable rates of successful engraftment mean each lab uses the technique they’ve find works best for them.
    • Different mice are used, from different degrees of immunodeficiency to different degrees of immune reconstitution and humanization.
    • The technique is relatively new so standardized protocols are thus inherently rarer.
    • There are so many subtypes of cancer and thus far only a few have even be studied with the aim of creating PDTX models.
  • .The consequence of this is a lack of continuity across protocols and this again leads to a problem of with standardizing protocols.
  • The lack of a competent immune system in normal PDTX mice means these models may not not accurately represent disease progression and the therapeutic response that would be observed in immune-competent humans. Additionally, it makes them less suitable for immunotherapeutic cancer vaccine research.
  • Creating immunocompetent mice is both Labor-intensive and technically-challenging.
  • PDTX models are unsuitable for the study of early tumorigenesis including the genetic events that led to tumor formation.
  • While humanized mice can be create, to populate these immunocompromised mice with human immune cells further increases the finance and labor required and also extends the timeline for creating a viable mouse model on which to test compounds.
  • Creating a humanized PDTX mouse is a feat contingent on a lot of variables coming together and syncing up correctly. To create humanized mice and then use them as PDTX models, newborn mice must be irradiated and then engrafted with human CD34+ hematopoietic stem cells from human umbilical cord blood. The timing of obtaining cord blood, irradiating newborn mice and verifying the humanized phenotype of the NOD/SCID mice after engraftment, makes this an inconvenient and very involved technique.
  • Every 10-12 passages of tumors from mouse-to-mouse, researchers must start from scratch to avoid using tumors that have acquired mutations.
  • It is very technically difficult, expensive and time-consuming to arthroscopically place a tumor.
  • Full restoration of the immune system in a humanized mouse is not possible, as restoring HLA class I- and class II-selecting elements in T-cell populations remains a challenge.


Despite the disadvantages around using the various PDTX models available, they’ve already improved the success rate of several drugs tested in mice translating into efficacious treatments for humans. This field is evolving so rapidly and involves so many different PDTX models, it’s very difficult to pin down which approach is the closest and most accurate model of the human tumor and its microenvironment. It’s likely that this depends largely on the questions the researchers are trying to answer, the type of cancer their studying and the specifics of the mice with which they’re working. As the research protocols used becomes more streamlined, more widely used and more routine, the real value of these models will become more quantifiable but in the meantime it’s still a very promising area that’s sure to produce more pertinent and helpful data as they already have.

Article by Olwen Reina. Contact Olwen at