Spontaneous cancers in dogs and cats are an underused group of naturally occurring malignancies that share many features with human cancers such as osteosarcoma, prostate and breast cancers, non-Hodgkin's lymphoma, melanoma, soft tissue sarcoma, head and neck carcinoma, and virally induced lymphomas. Treatment of pet animals - primarily dogs - with naturally occurring cancer helps researchers better understand the biology of cancer and to improve the assessment of novel treatments for humans. The National Cancer Institute (NCI) Comparative Oncology Program has a collection of cancer tissue samples from dogs that span five different histologies (mast cell tumor, hemangiosarcoma, soft tissue sarcoma, pulmonary tumor, osteosarcoma, lymphoma, and melanoma) with anywhere from 100-500 patients (dogs) per histology. Each patient has tumor and normal tissue, serum, plasma, whole blood, and urine. All patients are enrolled while treatment-nave and have clinical outcome data. RNA-Seq experiments were completed on selected canine tissue samples from five tumor histologies (melanoma, pulmonary neoplasms, osteosarcoma, and B and T cell lymphoma). Bioinformatics analysis of the data identified five gene co-expression modules associated with the cancers. Each cancer type has a unique activation pattern across these modules, representing distinct biochemical pathway sets. This disease model was not previously known, and it has been used to generate biomarker and drug repurposing hypotheses associated with the various cancers. With respect to the expression data: 1) 386 biomarker genes were evident at the transcript level; (2) 271 of these were present as proteins in tumor tissues; and (3) 45 of these were also detected in blood. Using literature mining, 62 existing drugs were identified that are known to affect expression of at least one of the 386 biomarker genes. When tested as single agents against a canine cancer cell line panel, 44 of these drugs were active in at least one cell line. Of these 44 drugs, 35 synergistic dual drug combination hypotheses were assembled. Experimental testing of the combinations was completed, although results of these tests were inconsistent with the single agent tests. The inconsistencies were due to issues regarding drug stability under the conditions of the tests and significant contamination of the cell lines with mycoplasma. These issues have been addressed and both single and dual drug combination tests are being repeated. Comparative genome hybridization and resequencing experiments have been completed. Bioinformatic analysis to determine the copy number and mutational status of the biomarker genes is ongoing. Analysis of publicly available human RNA-Seq data for melanoma is complete. It was found that many of the genes upregulated in canine melanoma are also upregulated in human melanoma. Analysis of publicly available human data for the other cancer types is ongoing. Large-scale proteomics studies in canine will be performed to gather statistics confirming biomarker specificity among cancer types. Drug combinations active in canine cell lines will be confirmed by in vivo by testing in canines, informing potential translation to human therapeutic development. Biomarkers, drug combinations, and important genes that are found to be consistent across both human and canine analyses will then have the highest chance of being successful in a potential clinical trial.
|Grimm, Fabian A; Iwata, Yasuhiro; Sirenko, Oksana et al. (2016) A chemical-biological similarity-based grouping of complex substances as a prototype approach for evaluating chemical alternatives. Green Chem 18:4407-4419|