In patients with melanoma and in in vivo models, gut colonization by Bacteroides intestinalis is associated with increased immune-related adverse events, such as colitis, resulting from treatment with combined immune checkpoint blockade.

Immune-related adverse events (AE), such as colitis, are common and sometimes severe in patients receiving combination CTLA4 and PD-1 blockade. However, what causes these AEs has not been fully established, making it challenging to predict which patients are most at risk for them and to determine optimal solutions. Now, evidence from a recent study suggests that certain types of gut microbes in patients with melanoma may mediate these side effects—and may do so in part via the cytokine IL1β (Nat Med 2021 Jul 8 [Epub ahead of print]).


In 77 patients with advanced melanoma who received combination CTLA4–PD-1 blockade, 93.5% experienced immune-related AEs; 49% of all patients developed grade 3 or higher immune-related AEs. Pretreatment microbiome profiling performed using stool samples from 54 patients for whom specimens were available revealed that severe immune-related AEs were linked to the presence of certain gut bacteria, particularly Bacteroides intestinalis and, to a lesser extent, Intestinibacter bartlettii.

These results weren't a complete surprise. “Multiple reports by independent groups have linked outcomes of immune checkpoint inhibitor therapy with intestinal microorganisms,” says Diwakar Davar, MD, of the University of Pittsburgh in Pennsylvania, who was not involved in the study.

However, investigators were struck by the extent of the relationship. “It was a little bit surprising that the gut microbiota was the strongest association with toxicity in this cohort—stronger than any other factors that we could identify,” says Jennifer Wargo, MD, of The University of Texas MD Anderson Cancer Center in Houston, co–senior author of the paper with Laurence Zitvogel, MD, PhD, of the Gustave Roussy Comprehensive Cancer Institute in Villejuif, France.

This finding was validated in a second cohort of 45 patients treated for advanced melanoma. In this group,B. intestinalis was exclusively identified in pretreatment stool samples from patients who developed immune-related side effects.

In vivo experiments provided additional credence to the correlation between B. intestinalis colonization and immune-related AEs. Researchers treated mouse models of sarcoma and melanoma with antibiotics to ablate their normal microbiota prior to inducing recolonization with B. intestinalis and initiating combined immunotherapy. They found that the mice developed intestinal toxicity. Previous research suggested that commensal gut microbes can cause colitis through a mechanism driven by the inflammation-linked cytokine IL1β. Thus, the group gave the mice the FDA-approved IL1 receptor antagonist anakinra (Kineret; Sobi), which substantially diminished gut inflammation.

If validated, the finding that B. intestinalis mediates immune-related AEs in patients treated with combined immune checkpoint blockade could inform treatment. For example, clinicians could choose other therapies for patients with significant B. intestinalis colonization—or attempt to decrease colonization, for example, with selective bacteriophage treatment—before starting immunotherapy. “Our hypothesis is that by therapeutically targeting and depleting B. intestinalis, we could actually abrogate toxicity and preserve response to treatment,” Wargo says.

Another potential technique may be to target the proposed mechanism for immune-related AEs by blocking IL1β signaling. However, this method should be approached with caution because IL1β is also involved in pathways that mediate antitumor T-cell function, so preventing IL1β signaling may have the unintended consequence of suppressing therapeutic response.

A further option is fecal microbiota transplant, a procedure that has gained traction for multiple conditions despite varying levels of evidence for its efficacy. All of these approaches will require investigation in randomized clinical trials, yet Wargo is hopeful: “We're going to get tractable strategies to modulate either gut microbes or metabolites that will actually make a meaningful difference.” –Nicole Haloupek