5/16/2017 August Schiess, CSL
Written by August Schiess, CSL
Big data analytics has endless possibilities in health care. Through data analysis, scientists can determine subtypes of diseases, as well as how drug treatments will influence individual patient prognoses. Arjun Athreya, a PhD student in electrical and computer engineering, advised by ECE ILLINOIS Professor Ravishankar K Iyer, works at the forefront of this precision medicine work with Mayo Clinic. Both Athreya and Iyer conduct research at the Coordinated Science Lab.
How metformin is being repurposed as a cancer drug
Metformin, an anti-diabetic drug, has shown promise in treating triple negative breast cancer, a molecular subtype of cancer for which there are no targeted treatments yet. Triple negative breast cancer (TNBC) is not driven by the three receptors known to fuel most breast cancers, so it doesn’t respond to current targeted receptor treatments.
Athreya’s data-driven analytics and machine learning research allowed Mayo Clinic scientists to more closely examine the molecular mechanisms of TNBC. Using single cell techniques, they found that metformin shows preliminary effectiveness in reducing or preventing further tumor growth in this type of cancer.
“Precision medicine is about identifying disease subtypes of a particular disease, and deciding which treatment is best for each subtype,” said Athreya. “We found that metformin, which normally treats diabetes, also had therapeutic effects on this certain type of breast cancer.”
Athreya worked with Dr. Liewei Wang and Dr. Richard Weinshiboum from Mayo Clinic to investigate the particular genes in TNBC that were affected by metformin. Athreya determined that a small cluster of cells showed a different distribution in gene expression after metformin treatment. With further help from bioinformatics expert Dr. Rani Kalari and her team, they identified one gene for laboratory experiments, which established that it inhibits cancer cell proliferation and migration.
“I took data that was more complex than what biologists could see under a microscope. You can’t functionally study 20,000 genes—that’s very expensive,” said Athreya. “So we used algorithms and data analytics to narrow down the hypothesis to test just a few genes, and we were able to find key genes.”
The Mayo Clinic team has conducted a clinical trial that included many TNBC patients to begin the process of developing a therapeutic pipeline. Because of the unique characteristics of TNBC, it does not always respond to treatment.
“Because we know triple negative is a whole other beast, success of any drug treatment is going to be very variable. We’re trying to understand: what is the baseline biological difference between women who have shown therapeutic effects and the ones who haven’t?” said Athreya.
The team is building a model using a game theory approach that may be able to predict—given genes as biomarkers—whether a patient will respond to drug treatment. They can determine with about 95% accuracy in predicting the women who will and will not respond, but it is still being explored. Such a model is now helping further probe the reasons why some patients may be predisposed to respond better to the drug.
Athreya receives top accolades
His success in this research encourages continuing partnerships with Mayo Clinic.
“Having the opportunity to unite the Illinois engineering training and expertise of Arjun with the biomedical expertise and data available at Mayo Clinic as a result of a special fellowship that he has received has undoubtedly helped us to develop novel insights into the causes and treatment of human disease,” said Dr. Weinshilboum.