• Visit the Gerald P. Murphy Cancer Foundation web site for more info

We now have an approach to identify DNA copy number changes that allows us to zoom in on regions of the genome with ~75-fold greater resolution than was possible even just one year ago. Using this technology we will refine the genome regions of interest defined in CHF-760 and identify additional, smaller aberrations. Within these regions we will identify a series of candidate genes for functional analysis. We have developed an assay for use with archival canine tumor samples that allows us to rapidly determine the level of activity of multiple genes with higher sensitivity than was possible previously. Once we have identified the key genes of interest, we will use this assay to screen HM cases for the extent of gene deregulation. In this study we also will identify DNA copy number changes that are shared with human HMs.

Combining these approaches, we will narrow down the search for genes playing a key role in HMs and thus move a step closer to developing targeted therapies for canine patients diagnosed with this devastating cancer.

May, 2008 Update: Data thus far has validated that there is a segregation of distinct phenotypes and grades of tumors; and supports the notion that there are breed-specific differences in the age of lymphoma onset. We have identified preliminary regions in two chromosomes where gene expression signatures follow the same pattern of segregation for the gains and losses of DNA we identified previously. We predict that the results from this work will allow us to define how heritable factors influence the phenotypes and biological behavior of these tumors, setting the groundwork to develop better strategies for diagnosis, control, and treatment.

This project hopes to increase the sophistication of diagnosis and life expectancy for canine lymphoma.

For this study they used a population of archival tumor (lymph node) specimens from canine lymphoma patients that had been recruited as part of a clinical trial. The patient samples were of numerous breeds, both genders, from across the USA and had a wide range of ages. At the time of diagnosis, each patient has a lymph node biopsy taken that was fixed in formalin and embedded in paraffin blocks. These formed the archival sample population used for the present study. All patients were subsequently treated for their lymphoma with single agent doxorubicin and were clinically monitored during the course of their treatment and beyond. As part of the clinical follow-up the duration of first remission was noted.

Cells of lymphoma patients were extracted from the FFPE specimens and prepared on glass slides. A series of DNA probes was developed to assess copy number of four select regions of the genome using multicolor fluorescence in situ hybridization (FISH). Using this process they determined the mean copy number of each of the four regions in each patient sample. Statistical analysis of the data reveled that the mean copy number of one of the four regions evaluated was significantly correlated with the duration of disease free interval in the study population. This region may thus be considered to have prognostic significance. By combining these data with data from a parallel project (funded by Morris Animal Foundation), they identified two regions of the canine genome that together provided the means to predict duration of first remission of lymphoma patients treated with single agent doxorubicin with 95% confidence. Over the past year, during a no cost extension, they have developed a cytogenetic test that allows for these two regions to be evaluated routinely. The merged data from both projects (CHF-613 and MAFDO5CA019) were used to support a patent application and the resulting test has been licensed to a large biomedical company to bring to market. In addition they have evaluated the efficacy of this test to be able to predict duration of remission in canine lymphoma patients that were subsequently treated with multi-agent (CHOP) chemotherapy. These data are pending but early indications suggest that the region identified as part of CHF-613 is also significant in such cases.

In summary the outcome of this project is the development of the first cytogenetic test to predict treatment response in a canine cancer. The test should become widely available to the within the next year and will provide owners and veterinarians a new source of added data as they decide on the most appropriate treatment options for dogs diagnosed with lymphoma. This work will now move forward to investigate the genes within these two regions and determine their possible candidacy for new therapies.

It has been apparent for some time that certain dog breeds are prone to develop certain types of cancer. Specifically, studies completed between the late 1960s and the early 1980s defined relative risk of lymphoma for different dog breeds. Yet, there has been little progress since then to define factors that account for this risk. As part of ongoing programs supported by the AKC CHF in our laboratories, we showed recently that the breed-specific risk of lymphoma extends beyond the simple disease condition to a predisposition for specific forms of lymphoma. More importantly, we showed there are recurrent chromosomal abnormalities that segregate with specific forms of lymphoma and that are more common in Golden Retrievers than in other breeds, suggesting breed-specific profiles of genetic abnormalities will be found in canine lymphoma. To continue this work, we plan to use contemporary "array-based" technologies to identify genes that map to these regions and how they contribute to the disease. We anticipate that the results from this work will allow us to predict how heritable factors influence the occurrence of abnormalities in these genes, and will set the groundwork to identify specific genes associated with breed-dependent cancer risk.

While fresh tissue specimens are ideal for cytogenetic analysis, we also have developed a means to obtain data from archival specimens, represented by tumor samples that have been fixed in formalin and stored in paraffin blocks. These are the specimens that are evaluated by a pathologist during diagnosis. A major advantage to the use of archival samples is that we can select cases associated with detailed clinical and pathological information from the outset. In this study we analyzed cytogenetic data from 102 archival cases of canine lymphoma and demonstrated that the distribution of DNA copy numbers was comparable to that we had identified in 150 cases of fresh tissue. These data suggest that we are able to make use of large pathology archives as a source of case material for cytogenetic analysis. In this particular project the net results were that we were able to assess DNA copy number changes in all 252 cases and add substantially to the overall cohort.

With this large data set we identified several high frequency DNA copy number aberrations, some of which are associated recurrently with either B cell or T cell phenotype and some of which may be associated with further subtypes. In addition many of the cases used in this study were treated with standard of care chemotherapy and so as we follow the outcome of these patients we will be in a position to determine if any of the recurrent chromosome changes we have identified are associated with prognosis. As we proceed we now are using much higher 3 resolution technology, allowing us to narrow down the regions of interest and ultimately lead us to more accurate gene discovery.

MicroRNAs (miRNAs) are small non-coding RNAs that have been implicated in humans as having a fundamental role in cancer initiation and progression. Canine diffuse large B cell lymphoma (DLBCL) represents one of the most frequently encountered canine neoplasms. We hypothesized that canine DLBCL possess a unique miRNA expression signature and that miRNA dysregulation contributes to chemoresistance and prognosis. In this study we confirmed that miRNAs in the dog are very similar to those in humans, i.e. are highly evolutionarily conserved, and that different miRNA analysis technologies that are applied to human cancers can be similarly applied to canine cancers. Furthermore, miRNA expression in healthy normal tissues is similar in dogs and humans. When we used "next generation" sequencing of canine DLBCL tumor samples to determine their miRNA expression profiles we identified a miRNA that was significantly overexpressed in chemotherapy naïve tumors compared to chemotherapy relapse tumors from the same patients. Significantly, this was a miRNA with known importance in human cancers. Additionally, we identified a unique miRNA expression signature (multiple coordinately expressed miRNAs) that was associated with overall survival in dogs with DLBCL and treated with a standard chemotherapy regimen. These findings provide insights into mechanisms of lymphomagenesis and potential targets for future therapies.

Osteosarcoma (OSA), or bone cancer, affects 8,000 - 10,000 dogs in the United States annually. Large and giant breeds are at a much higher risk for this disease, suggesting that inherited risk factors are involved. Roughly 10 - 15 percent of Rottweilers, a mastiff-type breed and 15 - 20 percent of Greyhounds, a long-limbed hound-type breed, get the disease. Recently, we have identified several regions of the canine genome that are associated with an increased risk for OSA in Rottweilers. A similar study is underway for Greyhounds and is expected to identify additional regions of importance. The purpose of this study is to identify the actual genes and mutations causing the increased risk for bone cancer in Rottweilers and Greyhounds. We will then determine the frequency of mutations in these genes in OSA in other breeds as well as in other tumors. This work should allow the development of specific genetic tests for carriers of OSA and suggest further studies leading to improved treatments for bone cancer.

The completion of the canine genome has shown the remarkable similarities to that of humans. Similarly, many of the malignancies that occur in dogs are also like their human counterparts especially for the tumors of the lymphoid system. Research based on the DNA from cases accessed for the main study of this grant proposal are now being used to define the specific chromosomal and DNA transcript changes that are unique to the major type of lymphoma seen in dogs and in humans. Because the genetic record of inbred pure bred dogs is so much more consistent than in outbred humans it is possible to detect changes related to specific lymphomas in dogs much more easily than in humans. Thus instead of the dog being compared to humans for the study of lymphomas, humans are now being compared to genetic changes related to lymphomas detected in dogs.

The final objective of this research is to obtain follow-up information on as many of the 1000 cases studied as possible, to determine the impact of specific factors including age, stage of disease and general health at diagnosis, treatment protocol and specific type of lymphoma on survival. The application of this new information will permit tailoring of treatment for canine lymphoma according to tumor type and increase survival in our animal companions that share our lives and environments.

These finding were published in the Sept 2011 issue of Journal 'Genes, Chromosomes and Cancer'. Subsequent to that component of the study, we analyzed 23 of the canine cases using a CGH platform with considerably higher resolution. This new array comprised 180,000 probes (compared to the 2,100 probes on the earlier version) that provided 76 times the amount of data. Analysis of these 23 canine osteosarcoma cases concluded that a pattern of high genomic instability is maintained when increasing genomic resolution. As expected, the higher resolution analysis, however, identified several micro-regions (<500kb) of copy number changes that were shown to be comparable to what we observed in our parallel study of human osteosarcoma.

Comparison of canine and human osteosarcoma aCGH data identified that the most frequent regional copy number gains and losses in our dog and human cases were shared, with remarkably similar frequencies. The incidence of these aberrations in both in human and dog OS suggest that these specific genes may be directly contributing to the cancer phenotype. Nine genes of interest were analyzed for altered expression levels; MYC, MEN1, CDKN2A/B, TERT, PTEN, RHOC, RUNX2, TSC2 and TUSC3. From a comparative perspective these genes have now been implicated to be associated with human osteosarcoma. Functional analysis of these genes identified several cellular pathways deregulated in osteosarcoma. A manuscript reporting the finding of our high resolution, comparative assessment of canine and human osteosarcoma is in the final stages of preparation and is scheduled for submission in early 2012.

It is very clear from the data generated in thus study that human and dog OS share conserved DNA copy number aberrations. We identified that orthologous regions in humans and canines have similar types (gain and loss) and frequencies of aberrations in osteosarcoma. These data strongly support the conserved nature of osteosarcoma. These data are now being used to advance new areas of research that ultimately we hope will lead to improved therapies for our canine patients, and maybe even our human patients.

(also see Grant 947A above)

Objective: The researchers aim to use the most modern genetic techniques to identify the genetic location of this trait in the Rottweiler and eventually identify the actual genetic mutation associated with this trait.

Through statistical analysis a region of chromosome 21 was indicated as the most likely to contain a causative mutation for SAS. Through an independently funded study these results were paralleled in Golden Retrievers with SAS. This implicates the possibility for a shared causative mutation between Rottweilers and Golden Retrievers with SAS.

Evaluation of genes within this region on chromosome 21 is the next step toward discovery of the genetic mutation responsible for SAS in Rottweilers.

Entire final report as presented to AKC.

Our results show that AK inhibitors are capable of promoting cell death of dog bone cancer cells when used at relatively high concentrations in laboratory culture. In other words, these cells are rather resistant to these compounds, especially when compared to a variety of human cancers that are targets for this treatment. It is likely that the resistance is a characteristic of bone cancer itself. We documented mechanisms that potentially mediate resistance, and so conclude that these compounds should be possibly evaluated as part of combined approaches that target resistance (protocols could include combination with conventional chemotherapy agents that are currently used to treat dogs with bone cancer).

A 3D pelvic limb computer model, focusing primarily on the stifle, was developed and verified in our study. Our model was developed using CT imaging to depict the pelvic limb anatomy and gait analysis to characterize the kinematics and kinetics of a single canine subject during walking. We used our model to evaluate stifle ligament loads and tibial translation and rotation in the intact and CrCL-deficient stifle during simulated walking. The CrCL was found to be the primary intact stifle load-bearing ligament (peak load of 41% body weight occurring at 10% stance), and the caudal cruciate ligament was the primary CrCL-deficient stifle load-bearing ligament (peak load of 183% body weight occurring at 50% stance). In the CrCL-deficient stifle, peak tibial translation was 20.1 mm and peak tibial internal rotation was 8.8°, both occurring at 50% stance. Our model predicted tibial cranial translation and tibial internal rotation in reasonable agreement with previous in vitro and in vivo canine studies. This model provides a noninvasive visualization and quantitative analysis of stifle biomechanics and was used to investigate anatomical characteristics and biomechanical parameters such as body mass, ligament stiffness, etc. to gain an understanding of their role in producing conditions that predispose dogs to CrCL deficiency. Factors identified as having an influence on ligament loading and stifle stability are key to an improved understanding of the etiology of CrCL deficiency. Although beyond the scope of this study, our long-term research goal is to utilize our model towards an evidence-based assessment of surgical procedures used to stabilize the stifle following CrCL rupture.

The research conducted in this study will provide the basis for future research that may, ultimately, lead to scientists being able to provide a better assessment of individuals¿ risks for cancer (or for cancer in progeny), as well as determine whether a given dog is a good candidate for a given therapy. This project has helped to broaden the understanding of why tumors happen, so that the abnormalities can be targeted and better therapies devised. Researchers developed and tested gene therapy for melanoma. In a clinical trial involving five dogs with facial or oral melanoma, they found that the gene therapy, in which tumors were injected with modified genes, was both free of adverse effects and effective.

Canine Hip Dysplasia is the number one genetic health problem in dogs, with a significant frequency in over 100 breeds. The mode of inheritance may be polygenic (due to more than one gene) which has discouraged some studies of the genetic cause. However, with the proper approach, a polygenic disease can also be solved, and a DNA test or tests can be offered to help get rid of the disease. We propose a comprehensive approach studying 12 breeds and 11 additional candidate genes (beyond 2 already ruled out). Using multiple breeds increases the chance a given candidate gene will score a "hit." When a hit occurs, we will develop a DNA test to detect affected and carriers. In those breeds where the candidate gene search fails, we will do a genome-wide scan to establish linkage, which may allow us to offer a linked marker test, but in the longer run, will lead to gene discovery through homologous mapping. In the end, we expect to have a DNA test for hip dysplasia in most breeds.

Hemangiosarcoma (HAS) is a common cancer in dogs that originates from cells lining the blood vessels. HAS can affect any dog, but is seen more often in German Shepherds, Skye Terriers, and Golden Retrievers. This suggests that this disease has a heritable component. Tumors arise when cells respond inappropriately to growth factors, allowing them to divide continuously in an uncontrolled fashion. Tumor suppressor genes contain or eliminate these rapidly dividing cells, but mutations in these genes can disable their ability to function correctly. Our laboratory is examining the idea that the loss of function of one of these tumor suppressor genes, PTEN, leads to the increased production of tumor growth factors. In our studies, we will examine the frequency of the mutations in the PTEN gene from dogs with HAS, and the relationship of these mutations to increased production of a specific tumor growth factor, VEGF. The results of our research could lead to tests for screening dogs for mutations in PTEN, and information could have an immediate and long-lasting impact on canine health when used judiciously for breeding decisions. We will also test the ability of a novel therapeutic approach to restore normal function within these cells as a treatment for HAS. Such work may lead the way for the further development of novel therapies for the treatment of canine hemangiosarcoma.

Subvalvular aortic stenosis (SAS) is a congenital heart disease characterized by a fibrous ridge located below the aortic valve. Affected dogs are at risk of developing heart valve infections, congestive heart failure or sudden death. This trend has been reported with increasing frequency in the Rottweiler. The defect has been shown to be inherited in the Newfoundland breed, however the inherited nature of the disease in other breeds of dogs, including the Rottweiler, is unknown. The objectives of this study are to define the clinical presentation of SAS in the Rottweiler, compile pedigrees with reference to defined clinical cardiovascular status and evaluate for specific modes of inheritance, and accumulate a databank of clinical information, pedigrees and DNA from Rottweilers affected with SAS and all surviving family members. This study will help define both the clinical attributes and inherited nature of the disease. Information obtained in this study will provide the background for developing both screening and treatment programs and provide the initial materials for molecular studies to be performed in the future. Final Report: Read findings/results; read more on SAS