Gut Microbiome Can Influence Response to PD-1 Blockade in Melanoma

Jennifer A. Wargo, MD, MMSc

For patients with metastatic melanoma in particular, immune checkpoint inhibitors were a great advancement in therapeutics. Since then, investigators have focused on ways to improve on treatment with these agents, and one potential method is through manipulation of the gut microbiome.

In research led by investigator Jennifer A. Wargo, MD, MMSc, an assistant professor in the Department of Surgical Oncology and the Department of Genomic Medicine at The University of Texas MD Anderson Cancer Center, she and her colleagues demonstrated that patients who responded better to anti—PD-1 therapy had more diverse microbiome in their gut. Additionally, these patients had an abundance of the Ruminococcaceae family of bacteria within the Clostridiales order.

“There is a relationship between the gut, the tumor microenvironment, and the bacteria within our intestine that can shape our immune response and even shape how patients respond to therapy,” stated Wargo in an interview.

As part of the “Workshop on Single Cell Techniques in Immunology and Cancer Immunotherapy” sessions, Wargo will explain the role of the gut microbiome in patients’ responses to treatment with immune checkpoint inhibitors.

Although her research team is continuing to try to better understand the biologic mechanisms by which the gut microbiome enhances responses to immunotherapy and how these findings can be used, this is clearly an interesting subject to follow because of the importance of checkpoint inhibitor therapy, with further findings to come.

TARGETED ONCOLOGY:Can you please provide some background to your talk for those who will be attending the meeting?

Wargo:We have made advancements with cancer therapy with immune checkpoint inhibitors, first with CTLA-4 blockade and then with PD-1 blockade. Many patients respond to these treatments, but there are also many patients who do not respond to therapy. Our group has focused on mechanisms of therapeutic response and resistance to develop strategies to allow more patients to have durable responses.

During my surgical residency at Harvard Medical School, I looked at responses to different types of targeted therapies for cancer. We found that the stroma was responsible for mediating resistance to therapy with targeted agents, such as BRAF inhibitors. This research was published in Nature in 2012.

Hepatocyte growth factor was a factor that was being secreted by stroma cells in this therapeutic resistance. However, we wanted to investigate other cancer types to determine whether the same factors were mediating resistance to chemotherapy and additional forms of therapy. What we found with pancreatic and colon cancer cell lines was that there was one stroma cell line that would rescue all the cancer cells from treatment with chemotherapy.

We discovered that the cell line was infected with bacteria. This led us to wonder whether the bacteria were contributing to the cell line’s ability to mediate resistance to the cancer cells. What was happening was that the bacteria within these stroma cell lines were able to break down chemotherapy. This is how we began studying the microbiome. We launched a big effort to look at bacteria within human tumors. We became familiar with the literature of pioneers in the field like Dr Thomas Gajewski and Dr Laurence Zitvogel, who have shown that in addition to the tumor microbiome, the gut microbiome could influence response to cancer therapy, specifically immunotherapy, as well. However, most of the studies had been done in mice.

We decided to collect microbiome samples in patients with cancer. There was a year and a half spent collecting cancer samples from over 200 patients with metastatic melanoma who were going on to systemic therapy. These patients were profiled using different techniques. We profiled the oral and gut microbiome samples with sequencing. We then used whole-genome sequencing in a subset of exceptional responders, either exceptionally good or exceptionally bad.

We also profiled tumors in the blood of patients to look at immune responses. We first studied the largest subset within this cohort of patients who were treated with PD-1 blockade. When we looked at the signature of bacteria in the gut of tissues that responded to PD-1, there was a distinct difference in the patients who responded to PD-1 blockade versus those who did not. Patients who responded had a much higher diversity of bacteria within their gut and a different composition of the gut bacteria, with more Faecalibacterium, Ruminococcaceae, and Clostridiales. Patients who failed to respond to PD-1 blockade had a low diversity of the gut microbiome and a high abundance of Bacteroides. We investigated the tumor microenvironment to correlate what we were seeing in the gut. We saw that there was no strong difference in the oral microbiome between responders and non-responders. We looked at the tumor itself for patients between the bacteria and gut and saw that more patients had favorable microbiome in their gut, such as a higher diversity and more Faecalibacterium, Ruminococcaceae, and Clostridiales. We also found more CD8-positive T cells within the tumor microenvironment.

After investigating the blood of these patients, [we confirmed that] those with a more favorable microenvironment responded to checkpoint inhibitors. The next step was to take equal samples from responders and nonresponding patients and put those into germ-free mice for systemic immunity testing. We saw a difference between them. We implanted tumors and discovered that the tumors grew slowly in the mice with the transplants from the responders. In the mice that received transplants from the nonresponders, the tumors grew quickly. We treated the mice with PD-L1 blockade and found that the mice that received the transplant from responders had great responses to PD-L1, whereas those who received a transplant from nonresponders did not. This suggests that it is not just an association. There is a relationship between the gut, the tumor microenvironment, and the bacteria within our intestine that can shape our immune response and even how patients respond to therapy. We’re planning a clinical trial to change the microbiome in patients with cancer who are going to receive immune checkpoint blockade with the aim to improve responses to therapy.

TARGETED ONCOLOGY:Can you discuss that future research and what questions you hope to answer?

Wargo:A compelling area that is being investigated is to better understand how the microbiome may modulate immunity overall. There are data suggesting that patients who receive antibiotics either 1 month before or 2 months after the initiation of checkpoint blockade have a much worse response to therapy. That has more implications as far as curing patients with cancer who are on these checkpoint inhibitors. We will be closely monitoring and potentially limiting antibiotic use in those patients.

Should we be profiling the gut microbiome in patients going on to therapy? Again, that is something to consider, but my hunch is yes. The other question is, should we be considering the microbiome in looking at preclinical models, because then surgeons can use the modulation of the gut microbiome to enhance therapeutic responses.

TARGETED ONCOLOGY:Can you discuss the gut microbiome’s influence on other forms of cancer?

Wargo:There are many patients with renal cell carcinoma and lung cancer who are going to receive PD-1. Research has been conducted to investigate the gut microbiome and saw very similar signatures and patterns that had a higher diversity. Those who demonstrated a higher diversity did respond better to this form of treatment. Dr Gajewski has done work in a cohort of patients with melanoma with similar bacteria to those in our study, although not identical, since there are geographic differences and other factors, such as diet, that play a role. There is a definite consistency from data sets and different cancer sites. This is not something that is specific to melanoma. It can be applied to multiple cancer types as well as to other therapies, although that needs to be tested further.

TARGETED ONCOLOGY:Is there anything else oncologists should understand regarding the mechanism of action of the gut microbiome?

Wargo: