Turicibacter and Acidaminococcus predict immune-related adverse events and efficacy of immune checkpoint inhibitor

Abstract

Introduction: Immune checkpoint inhibitors have had a major impact on cancer treatment. Gut microbiota plays a major role in the cancer microenvironment, affecting treatment response. The gut microbiota is highly individual, and varies with factors, such as age and race. Gut microbiota composition in Japanese cancer patients and the efficacy of immunotherapy remain unknown.

Methods: We investigated the gut microbiota of 26 patients with solid tumors prior to immune checkpoint inhibitor monotherapy to identify bacteria involved in the efficacy of these drugs and immune-related adverse events (irAEs).

Results: The genera Prevotella and Parabacteroides were relatively common in the group showing efficacy towards the anti-PD-1 antibody treatment (effective group). The proportions of Catenibacterium (P = 0.022) and Turicibacter (P = 0.049) were significantly higher in the effective group than in the ineffective group. In addition, the proportion of Desulfovibrion (P = 0.033) was significantly higher in the ineffective group. Next, they were divided into irAE and non-irAE groups. The proportions of Turicibacter (P = 0.001) and Acidaminococcus (P = 0.001) were significantly higher in the group with irAEs than in those without, while the proportions of Blautia (P = 0.013) and the unclassified Clostridiales (P = 0.027) were significantly higher in the group without irAEs than those with. Furthermore, within the Effective group, Acidaminococcus and Turicibacter (both P = 0.001) were more abundant in the subgroup with irAEs than in those without them. In contrast, Blautia (P = 0.021) and Bilophila (P= 0.033) were statistically significantly more common in those without irAEs.

Discussion: Our Study suggests that the analysis of the gut microbiota may provide future predictive markers for the efficacy of cancer immunotherapy or the selection of candidates for fecal transplantation for cancer immunotherapy.

Keywords: Acidaminococcus; PD-1 inhibitor; Turicibacter; clinical efficacy; gut microbiota; immune checkpoint inhibitors; immune-related adverse events.

Figure 1

Relative abundance of intestinal bacteria in each patient before initiation of anti-PD-1 antibody therapy. (A) Percentage of bacteria at discernible genus level in the total stool of each patient. (B) Names of the bacteria represented in the bar graph in (A). (C) Bar graph showing the proportions of the bacteria in (A) that were found in 0.1% or more of the stools, summed to 100%. (D) Names of bacteria shown in (C).

Figure 2

Percentage composition of microbiota in groups based on the therapeutic efficacy of anti-PD-1 antibody treatment in cancer patients. (A) Relative abundance (%, composition) of bacteria at the genus level in the Effective and Ineffective treatment groups. (B) Names of bacteria shown in (A). (C) Bacterial tree diagram, with the dark gray and light gray lines indicating the bacteria found in the Effective Ineffective groups, respectively. (D) Bar graph showing the bacterial composition of the microbiota in the Effective and Ineffective groups. Bacteria that were found in more than 0.1% of the cases were summed to 100%. (E) Names of bacteria shown in (C).

Figure 3

Statistically significant differences in intestinal bacteria. We compared the statistical significance of differences in bacteria in the presence or absence of treatment effect, presence or absence of irAE, and presence or absence of irAE within the effective treatment group, using the Mann–Whitney Utest. The red line indicates a P value of 0.05. (A) Top-10 bacteria by treatment effect at the genus level. (B) Top-10 bacteria by irAE at genus level (C) Top-10 bacteria by genus level according to the presence/absence of irAE in cases showing effective treatment response to anti-PD-1 antibody.

Figure 4

Microbiota composition according to the presence or absence of immune-related adverse events (ir-AEs). (A) Relative abundance (%, composition) of bacteria at the genus level in the irAE- and no-irAE groups. (B) Names of bacteria shown in (A). (C) Bacterial tree, with dark gray lines indicating bacteria found in the no-irAE group and light gray lines indicating bacteria found in the irAE group. (D) Bar graph showing the microbiota composition in each group, where the sum of all the bacteria found in more than 0.1% of the cases in each group were summed to 100%. (E) Names of bacteria shown in (D).

Figure 5

Bacterial proportions in the microbiota in the group showing an effective response to anti-PD-1 antibody, with and without irAE. (A) Relative abundance (%, composition) of bacteria at discriminable genus level in patients with and without irAE who responded to anti-PD-1 antibody treatment. (B) Names of bacteria shown in (A). (C) Bacterial tree, with dark gray lines indicating bacteria found in the no-irAE group and light gray lines indicating bacteria found in the irAE group. (D) Bar graph showing the proportion of bacteria in each group, where the sum of all the bacteria found in more than 0.1% of the cases in each group were summed to 100%. (E) Names of bacteria shown in (D).

Figure 6

Alpha diversity of the intestinal microbiota. (A) Comparison of Simpson diversity index between effective and ineffective groups. (B) Comparison of Simpson diversity index between responders and non-responders in terms of immune-related adverse effects. (C) Comparison of Simpson diversity index between responders and non-responders in terms of immune-related adverse effects in the effective group.