Skip to main content

Researcher Interview: Elaine Mardis

, by Emma J. Spaulding, M.P.H.

Elaine Mardis, Ph.D.

Co-Director at the Genome Institute at Washington University School of Medicine and the Robert Lee and Louise F. Dunn Distinguished Professor at Washington University in St. Louis, Miss.

For this Researcher Interview, CCG Communications Director Emma J. Spaulding spoke with Elaine Mardis, Ph.D., Co-Director at the Genome Institute at Washington University School of Medicine and the Robert Lee and Louise F. Dunn Distinguished Professor at Washington University in St. Louis, Miss.

Emma Spaulding (ES): What are the top projects you’re working on right now?

Elaine Mardis (EM): There are many! Most of my work nowadays revolves around cancer and cancer genomics. One of the areas that’s very exciting for me is a project using biospecimens from a clinical trial of aromatase inhibitor response in breast cancer. This was conducted by one of the NCI’s Clinical Trials Cooperative Groups [Ed. Note: renamed the National Clinical Trials Network in 2014] which coordinated multi-site accrual of patient samples.

Patients with estrogen receptor positive (ER+) breast cancer were enrolled in the trial at different sites across the United States and received neoadjuvant aromatase inhibitor therapy. Patients donated a core biopsy sample from their tumor before they started taking the aromatase inhibitor. Then, after four months of treatment, a second core biopsy sample was collected.

This study compares the genomes of the same tumor before and after aromatase inhibitor therapy. We loosely stratified the patients into groups defined by responsiveness. In other words, one group showed tumor shrinkage and a decrease in the Ki-67 level of their tumors, which is a measure of aggressiveness, while the other group showed no tumor shrinkage or even some growth.

We’re trying to identify the differences in the genome and transcriptome of patients who responded well to therapy. This will help us develop predictive tests to evaluate patients’ potential for response to determine whether they start treatment or go to surgery. 

In 2012, we initially reported on the genomics of the pretreatment biopsies in Nature. We’re now finishing the comparison of the genomic profiles of the pretreatment to post-treatment biopsies. It’s a large and complex study that involves looking at both DNA and RNA.

In another project we are using genomics and specialized algorithms to identify which mutations in a patient’s tumor are likely to be most immunostimulatory.

ES: What are you hoping to learn from the immunology research?

EM: The end goal is to create personalized vaccine therapies for patients that allow the body to have a very highly specific immune response against the tumor cells. Patients would develop and sustain immunity to recurrent cancer. We have funding from the Susan G. Komen Foundation to investigate this immune therapy approach in breast cancer patients, and we have completed an early study in which metastatic melanoma patients were treated with personalized dendritic cell vaccines.

In this forthcoming  Komen-sponsored small clinical trial, we will sequence the genomes of triple negative breast cancer patients to identify the immunodominant epitopes expressed by the somatic mutations. These mutations would be a target for interaction between the T-cells and the cancer cells.

We’re hopeful that we will get the clinical trial approved and move forward with a small number of patients to show proof of principal for this type of work. So, I'm working in two different avenues: aromatase inhibitors and immunology.

ES: What do you see as the most promising technology right now?

EM: I think it’s only now that we have refined methods for using small tissue biopsies, next generation sequencing, and really advanced genomic analysis so that we can compare samples before and after treatment. More and more clinical trials are stipulating that biopsies are taken before and after the treatment if at all possible. By studying those before and after genomes, you get a huge amount of power to understand the mechanisms of resistance. If you understand that across multiple patients, you can design follow-up strategies.

For example, if a patient receives targeted monotherapy, the oncologist may know what signs to look for if the patient begins to develop acquired resistance to the therapy. That may mean that the patient can be rapidly switched from the therapy to which he's acquired resistance to a second therapy that helps battle that acquired resistance. If you’ve identified the "one-two punch," you can apply that clinically. This has already been done in melanoma

In melanoma, BRAF inhibitors have been very successful treatments, but patients can acquire resistance. In studying the before and after samples, researchers determined that MEK was activated in these patients to circumvent that pathway addiction. By combining BRAF and MEK inhibitors, patients show a longer, sustained response to the dual targeted therapies. However, patients do still acquire resistance!

ES: In the past, you’ve been involved in several big science and big data projects, like The Cancer Genome Atlas (TCGA), the Human Genome Project and 1000 Genomes. What are your thoughts on the future of big science and big data projects?

EM: I think that one of the benefits of doing big science projects like the Human Genome Project, TCGA, or the 1000 Genomes Project is that we provide the foundational knowledge that fuels the hypotheses that drive R01 proposals. We’re working on a grant fueled by mining large data sets, which you can’t just develop on your own dime. These data are already out there in the public domain so they’re available for use to generate hypotheses. That’s exactly what my collaborators and I have done in our grant application and I’m sure others have done as well.

I feel that there is value too in continuing to fund big science. Research at that level, while often not hypothesis-guided, is a hypothesis-generating vehicle. We still need to support that. But maybe now, it’s time to be a bit more strategic. Let’s consider moving towards capitalizing on what TCGA taught us by applying those technologies, study designs and analyses,  to smaller sample sets, potentially from clinical trial samples or locally banked patient samples that are more precisely able to address key clinical questions. Ultimately, this is the path to translating big science into advanced clinical cancer care.

I see it as a new integrated approach that expands on this translational element. We want to start making sure that the resources that went into TCGA begin to impact the lives of cancer patients.

To me, that is where the big science projects should be. We ultimately get more bang for our buck because we’re involving different parts of the oncology and cancer care community in applying what we’ve learned in TCGA, but in a much more focused and strategic fashion.

ES: Where do you see the future of cancer genomics going?

EM: Wow - that’s a really good question! A lot of what we are working on now is translation. TCGA has been a tremendous success in terms of defining what I call “the baseline”: What are the mutations in different cancer types? How prevalent are they? What combinations do they present in? What do you see not just in the DNA, but the RNA, methylation, and other data sets?

TCGA has really put together a beautiful fundamental story of the underpinnings of what makes cancer cancer. That’s an opinion shared by many people, not just me.

To me, this is the next step: To make sure translational research happens and starts to make a difference in cancer patients' lives. In the many cancer research conferences I attend, genomics figures prominently in every one of these meetings and TCGA data are mentioned in nearly every session. TCGA has provided critically important data for cancer genomics researchers. While providing and studying this foundational knowledge is important, you have to do something with it to make an impact. That’s the translational work being done now in many cancer centers and countries.

I know there’s still much to be done, but even in the short period of time since the advent of cancer genomics, we’ve accomplished a heck of a lot.

Featured Posts