Naturally occurring canine cancers have remarkable similarities to their human counterparts. To better understand these similarities, we investigated 671 client-owned dogs from 96 breeds with 23 common tumour types, including those whose mutation profile are unknown (anal sac carcinoma and neuroendocrine carcinoma) or understudied (thyroid carcinoma, soft tissue sarcoma and hepatocellular carcinoma). We discovered mutations in 50 well-established oncogenes and tumour suppressors, and compared them to those reported in human cancers. As in human cancer, TP53 is the most commonly mutated gene, detected in 22.5% of canine tumours overall. Canine tumours share mutational hotspots with human tumours in oncogenes including PIK3CA, KRAS, NRAS, BRAF, KIT and EGFR. Hotspot mutations with significant association to tumour type include NRAS G61R and PIK3CA H1047R in Hemangiosarcoma, ERBB2 V659E in pulmonary carcinoma, and BRAF V588E (equivalent of V600E in humans) in urothelial carcinoma. Our findings better position canines as a translational model of human cancer to investigate a wide spectrum of targeted therapies.

Canine tumours provide a powerful platform for translational investigation. Over the past decade, genomic characterization of canine cancers has highlighted the marked biological and molecular similarities between several canine and human cancers, including lymphoma, osteosarcoma, Hemangiosarcoma, glioma, melanoma, mammary tumours, and urothelial carcinoma. Some of the somatic mutations identified in these canine cancers occur at the orthologous position to known mutational hotspots found in human cancers, including PIK3CA H1047, BRAF V588(V600 for human), and FBXW7 R470 (R465 for human). These somatic mutations do not always occur in the same cancer type across species. More canine studies are needed to better characterize this association.

Although studies have shown genomic concordance between canine and human cancers, the number of canine tumours that have undergone genomic sequencing lags behind human tumours by an order of magnitude (fewer than 2000 canine tumours have been sequenced, compared to more than 20,000 human tumours). Consequently, the landscape of actionable tumour mutations in canine cancers is not fully understood. We sought to address this issue in order to assess the feasibility of matching dogs with spontaneous cancers to targeted therapy, thereby providing treatment opportunities to canine patients while developing a platform that could accelerate a more global understanding of the clinical as well as translational potential from dogs to humans.

To do this, we developed a next-generation sequencing (NGS) panel targeting coding exons of 59 genes frequently mutated in human cancers. Using this panel, we performed the largest sequencing study of canine cancers to date, including 671 tumours of 23 histologic types from dogs representing more than 96 breeds. Importantly, our study revealed 18 canine mutational hotspots, 8 of which were orthologous to hotspots reported in human cancers and clinically actionable. These results demonstrate significant overlap in somatic hotspot mutations between human and canine cancers, further highlighting spontaneous canine cancers as an excellent model for the investigation of targeted therapies.