Extramammary Paget’s Disease (EMPD) is rare and as such it typically gets little attention when it comes to research and new treatments. However, that is about to change with groundbreaking new research at Fred Hutchinson Cancer Research Center in Seattle, WA (USA). This research not only gives hope to patients with EMPD but it could also open up treatment for other rare cancers as well.

Dr. Taran Gujral, PhD is a systems biologist at Fred Hutchinson Cancer Research Center and he takes a big-picture, multidisciplinary approach to studies of cell-cell interactions. Taran investigates both tumor cells and their microenvironment — the noncancerous cells that surround them.

Dr. Gujral's lab has created a series of machine learning approaches, called KiDNN (Kinase inhibitor prediction using Deep Neural Networks), and KiR (Kinome Regularization) that utilize drug-target profiling and algorithms that mimic aspects of the human brain.

In the simplest of terms, Dr. Gujral's AI-based screening can be thought of as a game of 20 Questions. Many of us may remember the well-known game is played when one person asks a question and then a second person answers the question. As answers keep coming, the first person narrows their questions and within a short time they are able to determine the answer to the initial question. Dr. Gujral's project is similar to 20 Questions with his screening consisting of around 30 computationally-chosen, broadly specific compounds. The cancer tumor sample is tested against each compound and the answer to each test leads the AI to predict inhibitors and drug combinations with the highest likelihood of success.

As Taran explains, "Our approach does not require a biological understanding of why a drug candidate could work but instead employs massively powerful computation to predict which of millions of drug combinations is most likely to be effective in a rare tumor.”

Dr. Gujral's lab initially focused on three different types of rare cancers (Fibrolamellar cancer, Ependymomas, and Neuroblastoma). However, now they are including other rare cancers such as extramammary Paget’s disease (EMPD).

Taran plans to distribute copies of his AI-based chemical screening library to 100 labs across the world. The KiDNN screen is unique because it consists of multiple compounds and can be completed in a short amount of time. A cloud-based platform, using AI-based chemical screening, allows for machine learning algorithms to provide a ranked order list of FDA-approved drugs, or combinations of drugs. The analyses can investigate more than 13 million drug combinations.

The screening does require newly-obtained unadulterated tumor tissue for analyses. Dr. Gujral hopes to partner with other labs so that the testing can be done early while there is adequate uncontaminated tissue available.

If you are in the medical field, and actively treating or analyzing EMPD, you are encouraged to contact Dr. Gujral’s lab for additional information. If you are an EMPD patient, sharing this information with your physician could be useful. By advocating for this new testing with your doctor, you may be one of the keys to unlocking future EMPD treatment.

For a PDF summary of Leveraging Artificial Intelligence to Uncover Molecular Targets and Therapeutics for Rare Cancers, click this link.