Of Mice and Men:
Why we Need to Embrace
Mouse Models of Human Sarcomas
Like most of the people who read this newsletter (ESUN), I am hell-bent on curing sarcoma in my lifetime. I participate in the process by focusing my clinical practice on the pathologic diagnosis of sarcomas and my laboratory is trying to understand the pathogenesis of sarcomas. We’re trying to unravel the molecular circuitry that tells sarcoma cells to divide endlessly and invade and destroy vital organs. The overall goal is to identify proteins that can be targeted by drugs to kill tumor cells selectively over non-tumor cells. We take a variety of approaches to accomplish this in my laboratory, but the approach that I am most excited about is mouse models of human sarcomas.
Traditionally, researchers have studied frozen sarcoma samples from individual patients to find out what genetic mutations and unusual proteins are present in sarcomas. Alternatively, they have either developed or obtained sarcoma cell lines and manipulated these in culture in plastic dishes or less often, engrafted these cell lines onto immunocompromised mice (xenografts). This has been quite productive, resulting in the discovery of many oncogenes and tumor suppressors and, in some cases, drugable targets such as KIT receptor tyrosine kinase in gastrointestinal stromal tumor and platelet-derived growth factor receptor beta (PDGFRB) in dermatofibrosarcoma protuberans have been identified. These findings have revolutionized the treatment of these two neoplasms as imatinib mesylate (Gleevec) targets KIT and PDGFRB and is a highly effective, relatively non-toxic therapy. Yet, these examples are the exceptions and we are faced with the daunting task of finding drugable protein targets and drugs for the other 70 or so sarcomas for which there are no good systemic therapies.
While tissue samples are helpful in gene/protein discovery, and the presence of different potential targets can be validated in these samples, it is difficult to study functional proteins and impossible to know if a given drug will work using only frozen tissue. Cell lines and xenografts have the advantage that they are alive and so they can be drugged. However, cancer cells change dramatically in culture and lack the microenvironment found in the whole organism. This is crucial in understanding how cancer cells function in the context of the immune system and surrounding tissues/organs. This has left us with a system where new drugs/therapies that have been developed are tested only modestly, if at all in these suboptimal cell line models which fail to recapitulate so many of the important aspects of sarcoma biology, prior to testing them in humans. Their utility has to be regarded as questionable, especially when so many drugs that looked promising in cell lines and xenografts failed miserably in clinical trials.
In my opinion, most drugs are brought to human clinical sarcoma trials before adequate preclinical evaluation in suitable models. They are generally given to the sickest, most hopeless patients. Drug companies have a lot invested in getting their drugs to market as soon as possible so even the most meager response, in very few patients is considered a success and often serves as the basis to develop larger clinical trials. This system is more or less hopeless, and success has been minimal at best.
Since it is difficult to find sarcoma patients and clinical trials are extremely expensive, very few clinical trials in sarcoma can be conducted. However, new drugs with huge potential for clinical benefit are being developed at an extraordinary rate. There are literally thousands of potentially useful compounds for the treatment of sarcoma. If we are given so few attempts at putting together good human sarcoma trials, and there are so many drugs, we need to develop a good system of well-validated preclinical models to identify the best candidate drugs. A solution, I believe, lies in the development and use of mouse models of human sarcomas.
Mice have been used for decades to understand various aspects of mammalian biology. They are closely related to humans and essentially have all of the same genes, proteins and organs. Over the last couple of decades, scientists have been perfecting techniques known collectively as transgenic stem cell technology to alter the mouse genome with extraordinary success. Indeed, the 2007 Nobel prize in Physiology or Medicine was given to a trio of stellar scientists for "their discoveries of principles for introducing specific gene modifications in mice by the use of embryonic stem cells". They laid the groundwork for the development of mouse models of human sarcomas. Fortunately for the sarcoma field, one of these Nobel laureates, Dr. Mario Capecchi (University of Utah) is interested in sarcomas and created the first mouse models of synovial sarcoma and alveolar rhabdomyosarcoma. His mouse models recapitulate all of the salient features of both of these sarcomas.
Other mouse sarcoma models that have also been created include malignant peripheral nerve sheath tumor, osteosarcoma, leiomyosarcoma, embryonal rhabdomyosarcoma, gastrointestinal stromal tumor and malignant fibrous histiocytoma (sarcoma, not otherwise specified). The development of still more mouse models of human sarcomas are under way. Refinements to these models which allows real time imaging of the sarcomas using fluorescent proteins and many other useful features are ongoing. Mouse models are getting better all of the time, and some of them are ready for prime time in preclinical drug screening scenarios.
Since these are genetically engineered models, and it only takes 18-20 days to produce a litter of mice, we can produce virtually limitless numbers of mice with sarcomas. These mice can be evaluated using novel treatment regimens at a fraction of the time and cost it takes to conduct such trials in humans, even if you could find enough patients with a given cancer and get all of them to enroll in protocols, which often times is impossible. It is obvious that mice offer many advantages as a pre-clinical pipeline for evaluation of the most promising drugs. Once efficacy has been established, mouse models can be used to establish optimal dosages, modes of administration, etc.. The possibilities for their use are endless. Of course, we are still learning a lot about making better mouse models, and the next generations of mouse models will continue to improve. The National Cancer Institute fully supports these models and has created the Mouse Models of Human Cancer (MMHCC) Consortium to facilitate their development, proliferation and use. Mice that are part of the MMHCC are supplied free of charge to interested researchers.
I can easily envision a not-too-distant future where legions of mice with sarcomas are used to evaluate promising drug candidates, and only after considerable validation in these models, will the most well-validated drugs be studied in large-scale human sarcoma trials. It seems to me that there will be a day when many sarcoma patients owe their lives to the lowly laboratory mouse.