SCDA’s Olga Troyanskaya Works to Understand Genetic Basis of Malignancy

When Olga Troyanskaya, a consultant at the Simons Center for Data Analysis, began studying gene function and regulation in cancer, she didn’t expect that her own pet would lead her to explore new ways of studying the disease. But when her 13-year-old German shepherd, Jessy, developed cancer in 2006, Troyanskaya met Karin Sorenmo, chief of medical oncology at University of Pennsylvania School of Veterinary Medicine’s Ryan Veterinary Hospital in Philadelphia, and realized that the animals treated there could provide valuable insights into the development of canine — and maybe even human — tumors.


Troyanskaya points Jessy to visualizations of big data.

During Jessy’s treatment, Troyanskaya and Sorenmo quickly saw the opportunity for an advantageous collaboration. Although Sorenmo’s Penn Vet Canine Mammary Tumor Program was founded to treat dogs, the program’s data collection allows Troyanskaya to study the development of cancers on a molecular level. And with those data, she hopes to one day deepen the understanding of what makes tumors become malignant in humans.

When a dog, usually from a nearby shelter, is brought to Sorenmo for treatment, two parallel tumor samples are taken from each tumor: one for diagnosis and treatment of the animal, and one for Troyanskaya’s research into the genetics of cancer.

“Dogs are more analogous to humans than many other animals we study,” says Troyanskaya. “So it’s very useful to study them as patients who develop cancers naturally and also live in a similar environment to ours.

Because dogs have multiple mammary glands, and because there is a lack of early tumor diagnosis among shelter dogs, Sorenmo’s patients tends to have multiple mammary tumors at different stages of development. Such data allow the researchers to study the signals associated with cancer’s progression from normal tissue to benign growths to malignant tumors. This array of information is particularly rare, as it is nearly impossible to find tumors at different stages of development in humans that are not simply metastases of one another.

These samples together provide Troyanskaya with an ever-larger dataset that one day may permit identification of the genes or groups of genes responsible for making tumors in dogs malignant. In this way, Troyanskaya, Sorenmo and their collaborators aim to eventually improve diagnosis and treatments for humans, and to develop targeted gene therapies to combat pre-malignant tumors.

“Cancer is this fundamental and incredibly important problem. We should be able to make a dent in it, and these large datasets can help us both understand the molecular basis of cancer and hopefully lead to development of better therapies,” says Troyanskaya.

Read also recent New York Times coverage of Troyanskaya’s work, ‘From Dogs, Answers About Breast Cancer.’