Anticancer Immunotherapy by CTLA-4 Blockade Relies on the Gut Microbiota


Pictured above: B. Routy, M. Vetizou, N. Jacquelot, JM Pitt, T. Yamazakie, R. Daillère, S. Rusackiewicz, L. Zitvogel, C. Flament Not Pictured: MP Roberti, CMP Duong, S Becharef, V Poirier-Colame

Marie Vétizou, Jonathan M. Pitt, Romain Daillère, Patricia Lepage, Nadine Waldschmitt, Caroline Flament, Sylvie Rusakiewicz, Bertrand Routy, Maria P. Roberti, Connie P. M. Duong, Vichnou Poirier-Colame, Antoine Roux, Sonia Becharef, Silvia Formenti, Encouse Golden, Sascha Cording, Gerard Eber, Andreas Schlitzer, Florent Ginhoux, Sridhar Mani, Takahiro Yamazaki, Nicolas Jacquelot, David P. Enot, Marion Bérard, Jérôme Nigou, Paule Opolon, Alexander Eggermont, Paul-Louis Woerther, Elisabeth Chachaty, Nathalie Chaput, Caroline Robert, Christina Mateus, Guido Kroemer, Didier Raoult, Ivo Gomperts Boneca, Franck Carbonnel, Mathias Chamaillard, Laurence Zitvogel

Immune checkpoint blockers (ICB) have been successfully used as cancer immunotherapy with unprecedented success in patients with advanced malignancies. However, the majority of patients fail to benefit from this revolution and can develop severe immune-related adverse events (irAEs) entailing the end of their ICB therapy. A partial explanation for these failures could stem from the gut microbiome.

In our latest research published in Science, we investigated whether commensal bacteria could determine or boost ICB therapy by anti-CTLA4 antibody. We found that antibiotics or germ free (GF) conditions compromised tumor control. With a focus on the gut epithelial barrier where the irAEs mainly occur, we identified signs of subclinical colitis in mice. These modifications are associated with an alteration of the microbiome, leading to an underrepresentation of Bacteroidales and Burkholderiales families as well as a concomitant increase in Clostridiales.

The tumor growth control is restorated in antibiotics-treated mice or in GF conditions by oral gavage with certain species of Bacteroides (B. fragilis, B. thetaïotaomicron) or Burkholderiales (B. cepacia). In addition, adoptive transfer of B. fragilis–specific T cells or dendritic cells loaded with B. fragilis polysaccharides increased anti-CTLA4 efficacy in the absence of gut microbiota.

Lastly, we analyzed the bacterial composition in stool samples from 25 patients with metastatic melanoma before and after ipilimumab treatment. These patients clustered into three groups (A, B and C): cluster A driven by Alloprevotela / Prevotella whereas clusters B and C are populated by different Bacteroides species. Ipilimumab treatment appeared to change repartition into each cluster in favour of cluster C. Fecal microbial transplantation of feces from patients into GF mice significantly affected anti-tumor responses to CTLA4 blockade. Cluster C transplants (rich in B. fragilis) restored anti-CTLA4 efficacy in GF mice whereas cluster B failed to do so. We found an anti-correlation between the relative quantities of B. fragilis in GF mice receiving cluster C feces and the tumor size, illustrating the capacity of intestinal bacteria to influence anti-tumor immunity by ICB.

In an attempt to uncouple efficacy from toxicity following anti-CTLA4, we evaluated signs of subclinical colitis in mice receiving B. fragilis alone or in combination with Burkholderia cepacia. This combination was able to protect against intestinal lesions associated to anti-CTLA4 therapy.

Our study demonstrates consideration of our microbiote will be a significant step towards potentiating cancer immunotherapy with ICB.

To read the full paper click here.