The major goal of this study was to investigate the molecular characteristics of canine follicular thyroid carcinoma (FTC). This is a rapidly growing and highly aggressive tumor in dogs, and many patients present with evidence of invasion or metastasis. Some smaller independent studies have attempted to evaluate the role of single molecules such as p53 and thyroid transcription factor-1 in tumor development, often with inconclusive results. In the present study, a genome-wide approach was employed to achieve the first objective of determining the gene expression profile of FTC compared to normal thyroid tissue. Microarray analysis was performed in a pilot study using five FTC tissues and four normal thyroid gland tissues, and this showed 489 transcripts as differentially expressed between the two groups; 242 were down-regulated and 247 were up-regulated. Some important biological functions that were affected include regulation of cell shape, cell adhesion, regulation of MAP kinase activity, angiogenesis, and regulation of cell migration.
Osteopontin was a gene of interest as tumors consistently expressed it at high levels while it was expressed at low levels in all of the healthy samples. One of its up-stream regulators, VEGFA, was also differentially expressed but with a smaller fold change. The expression of osteopontin was validated by quantitative PCR using three groups: non-invasive FTC (tumors with capsular invasion only), invasive FTC (tumors with capsular and vascular invasion), and normal thyroid tissue. Both non-invasive FTC and invasive FTC had higher osteopontin gene expression than normal thyroid tissue but the two tumor groups were not different from each other. The second objective was to determine the protein expression of osteopontin and VEGFA in the same cases using semi-quantitative scoring of tissues stained by immunohistochemistry. The results were similar, with non-invasive and invasive FTC having higher osteopontin protein expression than normal thyroid tissue, but showing no difference from each other. With respect to VEGFA, there was no difference in gene or protein expression among the three groups.
The final objective was to determine the plasma concentration of VEGFA and osteopontin in dogs diagnosed with FTC compared to clinically healthy dogs using a commercially available canine-specific ELISA. In this case, both VEGFA and osteopontin had higher plasma concentrations in dogs with FTC compared to healthy dogs. A small number of FTC cases were also measured two weeks after surgical removal of the tumor. Some cases showed a post-surgical decrease in VEGFA and osteopontin while others either remained the same or increased; however, the sample size for this comparison was small. The consistent expression of osteopontin in both tissues and blood suggest that it is a promising marker for identification of canine FTC. As in human studies of osteopontin in aggressive carcinomas, it is also possible to investigate it as a means of monitoring response to therapy, recurrence, and clinical outcome.