Cancer researchers and drug developers continue to say that although gene expression profiling has led to significant advances in cancer diagnosis and prognosis, in vivo animal models that allow translation of therapeutic strategies to the clinic are sorely needed.
Used in research since the 1980s, xenograft models (PDX models) or avatars in which patient tumor samples are grown in immunocompromised mice have proven useful for drug screening and biomarker development. But investigators say that significant hurdles to adoption of the technology as a translational platform remain, including lack of tumor heterogeneity and genetic diversity, both hallmarks of human cancers. The models, they add, fail to replicate the natural tumor microenvironment critical to replicating tissue or organ-specific properties that contribute to tumor progression and modulate therapeutic response. Xenografts also may exclude the important interactions between immune and cancer cells during tumor initiation, maintenance, and metastasis.
Other liabilities to avatar models that may limit their practical use in a research setting include low tumor take rates, slow tumor growth rate, and the need for multiple passages in mice leading to transformation of the original tumors’ character, thereby reducing reliability and making it difficult to get enough mice with tumors to produce statistically reliable results. Production of a tumor model may typically require three to six months, and in the meantime, scientists and suppliers readily say, patients may die of their disease.
New Jersey and Maryland-based Champions Oncology, among several other companies, offer individualized testing and development of mouse avatars intended for specific patients, at a cost of $10,000 to $12,000. But insurance companies don’t yet pay for the technology, which remains experimental. Several companies, including Oncotest, Xentech, Crown Bio, and Champions Oncology, maintain xenograft mouse collections for research purposes to evaluate approaches to treat cancer generally.
And clinicians, even though several companies offer personalized avatars to patients who can afford them, remain divided as to these models’ utility in guiding therapy. An article in The Scientist noted while PDX mice used for basic research have helped identify drug candidates, personalized mouse avatars have yet to demonstrate clear changes in the course of a person’s disease, despite the occasional remarkable anecdote.
Edward Sausville, M.D., an oncologist at the University of Maryland School of Medicine, said because of that, “I would categorically never recommend it to my patients.” Commented Judy Boughey, M.D., an oncologist and researcher at the Mayo Clinic who is using PDX mice in a breast cancer study, “No one has really shown that [using PDX mice to evaluate drug responses] actually changes and improves outcome. The technique requires further evaluation and validation before it’s ready for everyone to just go and have their tumor xenografted.”
Such limitations were weighed in an article (“The promise and challenge of ovarian cancer models”) that appeared in Translational Cancer Research (February 2015): “It has now become clear that, although the multiple histological subtypes of ovarian (and other) cancers are being treated with similar surgical and therapeutic approaches, they are in fact characterized by distinct phenotypes, cells of origin, and underlying key genetic and genomic alterations.”
Multiple Animal Models Required
According to researchers, to distinguish among the subtypes and develop personalized treatment methodologies, animal models of all histotypes need to be generated to provide accurate in vivo platforms for research and the testing of targeted treatments and immune therapies.
Both genetically engineered mouse models (GEMMs) and xenograft models can, they maintain, further help researchers understanding key tumorigenesis mechanisms, and also offer insight into enhanced imaging and treatment modalities, and companies and institutions have recently focused on avatar mice as human cancer models, forming partnerships and commercializing xenograft mouse collections to further develop and use the models in drug development and individualization of patient cancer treatment.
The Jackson Laboratory announced an extensive agreement with Boston’s Beth Israel Deaconess Cancer Center (BIDMC) that will, the institutions said, encompass a broad range of activities to advance cancer research and patient care and accelerate personalized genomic medicine. As part of the initiative, Jackson Lab’s PDX JAX mouse models, that are immunodeficient mice bearing tumors from treatment-naive and -resistant patients, and BIDMC’s genetically altered mice (GEMM mice) engineered to replicate human cancers will be used to study individual cancers. By combining these approaches, scientists at The Jackson Laboratory and Beth Israel Deaconess believe they can gain a better understanding of how cancer progresses and how drugs can prevent its spread.
The Jackson Laboratory says the two models complement each other, xenografts representing all of the complexities present in the patient, and the engineered GEMMS providing a simpler system in which “untangle” the tumor’s genetic pathways. According to Edison Liu, M.D., head of The Jackson Laboratory, “It’s the direction in which a lot of research groups are going.”
“You put these two together, and you have a really powerful platform,” Dr. Liu said. “Not only can you empirically test if something works or not, you can figure out why it doesn’t work.”
In commenting on patient individualized avatar mice, Dr. Liu told Clinical OMICs that “I have no doubt there’s some benefit in these models but it won’t be perfect. The more you try to isolate the cancer components of a tumor the more you get away from the tumor environment. Every time you harvest cells there’s a selection process. What these models do is provide an approximation of the truncal mutations that were tumor drivers; there is no such thing as a perfect representation of tumor.”
Dr. Liu stresses that the collaboration between JAX and BDMIC is based on complementarities in technology. “We want to work toward the cure of cancer, defining over time what we think we can control. PDX models offer the best representation of patient tumors we have today, despite complexities and heterogeneity, which we are attacking using genomics.”
He pointed out that BDMC has expertise in cell biology and clinical sciences that complement current genomic technologies and genetic approaches. “The idea that you can know everything about a tumor from its genomic configuration is just not correct. You need downstream biology.”
He said the GEMMS models allow researchers to determine the specificity of a drug, while the PDX models provide a reproducible surrogate for a primary tumor. Each supports cell biological findings.
“By engineering specific mutations in different genetic backgrounds, which we can more easily do now, we have another discovery tool for genetic modifiers in cancer,” he further notes. “These tools can capture the complexity of cancer in a systematic way.”
Cheryl Marks, Ph.D., associate director of the division of cancer biology, at the National Cancer Institute, believes that “If we’re going to improve the reliability of information from animal models that we’re going to have to use a variety of models.” Other than genotyping a cancer patient’s individual tumor researchers have increasingly turned to these models, as, she said, “We seem to need every tool in the shed.”
Some clinical trial results have supported the use of the personalized tumor graft model as an investigational platform for therapeutic decision making and to guide cancer treatment in the clinic. Hidalgo et al., reported in a 2011 paper in Molecular Cancer Therapeutics that propagation of tumors resected from 14 patients with refractory advanced cancers in immunodeficient mice identified effective treatment regimens.
The authors, one of whom was David Sidransky, M.D., Ph.D., director of the head and neck cancer research program at the Sidney Kimmel Comprehensive Cancer Center at Johns Hopkins University and founder of Champions Oncology, say the treatments selected for each individual patient were not obvious and would not have been the first choice for a conventional second- or third-line treatment. In the study, tumors resected from 14 patients with refractory advanced cancers were propagated in immunodeficient mice and treated with 63 drugs in 232 treatment regimens.
Xenograft models identified an effective treatment regimen for 12 patients. One patient died before receiving treatment, and the remaining 11 patients received 17 prospectively guided treatments. Fifteen of these treatments resulted in durable partial remissions. Overall, the authors noted, there was a remarkable correlation between drug activity in the model and clinical outcome, both in terms of resistance and sensitivity.
In a July 2014 paper in the journal Cancer, Justin Stebbing, Ph.D., of the department of oncology, Imperial College and Imperial Healthcare National Health Service Trust, Hammersmith Hospital and colleagues at several institutions, including Johns Hopkins, and Champions Oncology reported, use of its TumorGrafts technology in 29 individuals who had advanced sarcoma.
The investigators implanted tumors resected from 29 sarcoma patients into immunodeficient mice to identify drug targets and drugs for clinical use. The results of drug sensitivity testing in the TumorGrafts were then used to personalize cancer treatment. Of 29 implanted tumors, 22 (76%) successfully engrafted, permitting the identification of treatment regimens for these patients. Although six patients died before the completion of TumorGraft testing, a correlation between TumorGraft results and clinical outcome was observed in 13 of 16 (81%) of the remaining individuals. No patients progressed during the TumorGraft-predicted therapy.
The authors note that the xenograft approach has its challenges, including the need for a sufficient amount of fresh tissue, which they note can be difficult considering that many of these patients are not otherwise scheduled for surgery. And, they acknowledge, although the development and propagation of a TumorGraft model can take as little as six weeks, it is more often up to six months before study results are available.
Obstacles aside, they say, the TumorGraft models used in their study included the tumor microenvironment present in the human host and maintained the features of the transplanted tumor, including gene expression profiles, copy number variants, and, most important, treatment response. The current data, they concluded, support the use of the personalized TumorGraft model as an investigational platform for therapeutic decision-making that can guide treatment for rare tumors such as sarcomas and note that the results warrant a randomized Phase III trial versus physician’s choice.
But according to Len Lichtenfeld, M.D., deputy chief medical officer of the American Cancer Society, “There’s not a lot of science to say how well this works, and it should be considered highly experimental.”