As part of our coverage of the 2017 American Society of Clinical Oncology (ASCO) Annual Meeting, held June 2–6 in Chicago, we are joined by Jennifer Wargo, MD, an associate professor in the department of surgical oncology and genomic medicine at the University of Texas MD Anderson Cancer Center in Houston. Dr. Wargo’s laboratory studies the spectrum of cancer patient responses to targeted and immunotherapies to understand factors that facilitate responses to treatment and also to track and understand resistance. At the ASCO meeting, Dr. Wargo discussed her recent work on the effect of the gut microbiome on responses to anti–PD-1 antibodies.
—Interviewed by Anna Azvolinsky
Cancer Network: Dr. Wargo, your study examined the role of the gut microbiome in responses to anti–PD-1 antibodies. Can you define the gut microbiome for us and talk about what is known about how it may modulate responses to immune checkpoint blockade?
Dr. Wargo: In our bodies, we have trillions of bacteria that outnumber normal cells by up to 10 to 1, and the largest numbers of these bacteria are actually in the gut. There is certainly a growing appreciation of the role of the gut microbiome in health and disease, and it’s implicated in a number of different diseases, and more recently it’s been implicated in responses to immune checkpoint blockade. Most of these studies to date have been in mice, including some very elegant investigations by Dr. Tom Gajewski and Dr. Laurence Zitvogel, which were published in Science in 2015, showing that if you modulate the gut microbiome in mice, you can actually enhance the response to immunotherapy.
In our ASCO presentation, we reported on findings where we basically show this in patients—what we wanted to do is to study the role of the gut microbiome in response to immune checkpoint blockade in patients with melanoma. We wrote a protocol and enrolled a large number of patients with metastatic melanoma who were going onto systemic therapy and collected both oral and gut microbiome samples. In the course of about a year and a half we collected samples from over 200 patients. The majority of patients were treated with checkpoint blockade, with a significant number of patients treated with anti–PD-1 therapies (around 115 patients). In those patients we collected both oral and gut microbiome samples just before they started therapy, and we also did a tumor biopsy when feasible. We then started them on anti–PD-1 therapy and measured responses after about 12 weeks.
What we found when we did this—and we actually did sequencing on the gut and oral microbiomes—was that there were no differences in the oral microbiome between responders and non-responders, but when we looked at the gut microbiome, we found significant differences between the gut microbiome of responders and non-responders. Mainly, the responders to anti–PD-1 therapy had a more diverse gut microbiome, and on top of that had different bacteria within the gut, which were associated with a response.
We looked at the immune infiltrates—so the infiltrating immune cells of these patients—and found that the responders to anti–PD-1 therapies certainly had more infiltrating CD8-positive T cells, which is no surprise. However, when we actually looked at the bacteria that were in the intestine and compared that to various markers of cytotoxic T cells within the tumor microenvironment, we found that there were several bacteria in responders that were highly associated with the presence of cytotoxic T cells—mainly Faecalibacterium, Ruminococcaceae, and Clostridiales. Conversely, we found that non-responders had a high abundance of Bacteroides within their gut and had a relatively cold tumor microenvironment.
Cancer Network: Were any of these results particularly surprising or unexpected compared to your hypotheses going into the study?
Dr. Wargo: No, I think we—and others who have been following these types of studies—are seeing interesting correlations or associations between the gut microbiome and immune responses, particularly in response to checkpoint blockade, but we’re not sure what came first. We and others have been doing studies where we take fecal samples from patients, either responders or non-responders, put them into germ-free mice, and look at what the immune system is doing in the mouse before we plant the tumor. Then we can implant the melanoma tumor and see how the tumor responds to immunotherapy or checkpoint blockade with anti–PD-1 or anti–PD-L1 therapies. What we see when we do this is that mice that get fecal transplants from patients responding to anti–PD-1 therapies have delayed tumor outgrowth and also have a markedly better response to anti–PD-1 therapy compared to non-responders.
I think the biggest implications of this are several. First and foremost, I think the gut microbiome does seem to have a major response to immune checkpoint blockade, and this brings up a couple of questions. Should we be profiling the gut microbiome of patients going onto these forms of therapy? I think the answer is probably going to be yes. And if so, how often should we profile them?
This also has other implications. There have been other investigations that have shown that if you give antibiotics before giving someone a checkpoint inhibitor, it actually impairs their ability to respond to that immunotherapy agent. Should we be limiting or at least recording the use of antibiotics in these patients? And furthermore, should we also be looking at diet and other factors, which potentially will influence the gut microbiome? And finally, is there a way to change the microbiome in order to make patients respond better to therapy?
We have a clinical trial under development that will start later this year to test that concept—can we actually modulate the gut microbiome by a fecal transplant or by some other strategy, give it along with a checkpoint inhibitor, and potentially enhance responses to therapy?
Cancer Network: Are there particular hurdles or difficulties in undertaking a study like this or even just doing the analyses of the gut bacteria?
Dr. Wargo: I think that the analysis of the gut microbiome is high yield, but how you interpret the data is important. There are people who have the expertise to know how to do this well, and I think that we need to both go to those people and to train more people in bioinformatics and microbiome analyses.
The other thing is that patients have contacted me and asked, Should I be taking a probiotic? What I tell them is that they should not be doing this outside of a clinical trial, and I think it’s important to study this in clinical trials. We are not sure that one single type of bacteria will actually be the answer, maybe it's a cocktail that we have to give, and it may not be the same bacteria for different treatments. There may even be regional differences where there are patients in Houston that may not be the same as patients in Chicago, Los Angeles, or anywhere else in the country. I think that it is still very unclear.
Nonetheless, I think it is very exciting and very provocative, and it opens a whole new field of opportunities not only to study the microbiome in patients but also to use the microbiome to potentially cure patients of cancer. Ultimately it may help us prevent cancer altogether in certain people.
Cancer Network: Thank you so much for joining us today, Dr. Wargo.
Dr. Wargo: Thank you.