Cholangiocarcinoma Risk Factors Open the Floodgates for Gut Microbes and Immunosuppressive Myeloid Cells

Cholangiocarcinoma is an aggressive malignancy of the bile ducts with high mortality rates, discouraging prognoses, and limited therapeutic options. Its etiology is highly multifactorial, and the contributions of different predisposing conditions remain incompletely understood. Primary sclerosing cholangitis (PSC), cirrhosis, and inflammatory bowel disease (IBD) including colitis are known factors that contribute to the development of cholangiocarcinoma (1). Interestingly, gut dysbiosis, associated with intestinal barrier dysfunction, is commonly observed in patients with these types of chronic gastrointestinal diseases. Whether such gut dysbiosis is also directly involved in cholangiocarcinoma development is unclear.

Moreover, much remains unknown about the immune milieu dynamics in cholangiocarcinoma, particularly the factors that control antitumor or immunosuppressive states. This is particularly critical in light of the relatively low cholangiocarcinoma response rates to immune-checkpoint therapies (ICT) in early clinical trials (2, 3). Cholangiocarcinomas have previously been found to harbor potentially immunosuppressive cell types such as tumor-associated macrophages and myeloid-derived suppressor cells (MDSC), but their relative contributions to cholangiocarcinoma etiology are unclear. MDSCs are a heterogeneous population of immature myeloid cells capable of suppressing cytotoxic T-cell and natural killer cell activity. Recently, multiple studies demonstrated that targeting MDSCs in other cancers can activate an antitumor immune response and enhance ICT efficacy (4, 5). Furthermore, recent studies have identified a significant role of the gut microbiome in modulating responses to anti–PD-1 ICT. Indeed, the patient microbiota composition in patients with melanoma can be used as a predictive biomarker of response to immunotherapy (6, 7).

In this issue of Cancer Discovery, Zhang and colleagues unite intestinal barrier dysfunction, the gut microbiome, and MDSC regulation to describe a novel process by which IBD and PSC can induce an immunosuppressive, cholangiocarcinoma-promoting hepatic microenvironment (8). Because the intestinal barrier is the first line of defense that inhibits the microbes in the intestinal lumen from entering the portal vein and eventually the liver, the intestinal barrier was studied in mouse models of colitis [via dextran sodium sulfate (DSS) administration] and PSC [via bile duct ligation (BDL) or MDR2 knockout], risk factors for the development of cholangiocarcinoma. In these models, a decrease in gut barrier function allowed gut-derived bacteria and lipopolysaccharide (LPS) to accumulate in the portal vein, in turn leading to a TLR4-dependent increase in immunosuppressive MDSCs in the liver. Antibody-mediated ablation of the MDSCs enhanced the frequency of liver-associated activated cytotoxic T cells. Supporting this connection, fecal microbiota transplantation (FMT) using stool samples derived from mice with gut dysbiosis promoted MDSC accumulation in the liver. This was blocked by pretreatment of the donor mice with neomycin to deplete Gram-negative bacteria, but not by vancomycin to deplete Gram-positive bacteria.

Mechanistically, BDL and DSS treatment caused a significant increase of CXCL1 in the liver, leading to an accumulation of hepatic PMN-MDSCs positive for the CXCL1 cognate receptor CXCR2. This accumulation could be blocked by either CXCL1 neutralization or pharmacologic CXCR2 inhibition. Moreover, CXCL1 overexpression and MDSC accumulation in response to LPS or DSS could be blocked by hepatocyte-specific TLR4 knockout. Most critically, four different mouse models of PSC or colitis led to accelerated cholangiocarcinoma growth in an AKT/YAP-hydrodynamic tail-vein injection model. Cholangiocarcinoma growth was also promoted by CXCL1 overexpression. Demonstrating the critical role of Gram-negative bacteria-induced MDSCs inthis cholangiocarcinoma enhancement, antibody-mediated MDSC depletion, CXCL1 neutralization, CXCR2 inhibition, neomycin treatment, or hepatic TLR4 knockout, but not NK-cell depletion, all reduced DSS-induced cholangiocarcinoma enhancement.

Finally, human patient samples were studied. LPS was detected in patients with liver cirrhosis, another inflammatory risk factor for cholangiocarcinoma. LPS induced CXCL1 in human hepatic cell lines, and FMT from cirrhosis patients into mice induced myeloid cell accumulation and a trend toward increased CXCL1. Moreover, PSC patients with active colitis had higher levels of CD15⁺ myeloid cells compared with PSC patients with inactive or absent colitis. Finally, a Tlr4 gene-expression signature was associated with worse survival in a published cohort of patients with cholangiocarcinoma. These preliminary studies suggested a conserved LPS–TLR4–CXCL1–MDSC axis between humans and mice.

This is the first study to delineate an immunosuppressive, tumor-promoting mechanism of cholangiocarcinoma risk factors that induce chronic inflammatory states, by allowing gut microbiota to spill into the portal vein and induce hepatic MDSC accumulation. The results suggest multiple novel nodes for pharmacologic intervention against cholangiocarcinoma growth, including Gram-negative gut bacteria, TLR4, CXCL1, CXCR2, and MDSCs themselves. Most critically, such steps toward reversing the immunosuppressive microenvironment would be predicted to enhance immunotherapeutic approaches including ICT. Given the clinical development of CXCR2 inhibitors, natural next steps could include combination therapy studies in immunocompetent cholangiocarcinoma mouse models, mirroring similar studies in colorectal cancer, rhabdomyosarcoma, and other models (4, 5). This is especially urgent, given that the response rate to PD-1 blockade monotherapy is low in cholangiocarcinoma, 6% to 17% in initial trials (2, 3). An interesting but more difficult question is whether prophylactic approaches in patients with PSC/IBD/cirrhosis to mitigate liver cancer induction might carry a sufficient risk–reward balance, as only a minority go on to develop cholangiocarcinoma or hepatocellular carcinoma.

Overall, this and other studies should motivate a deeper exploration of microbiome composition in patients with cholangiocarcinoma as well as searches for additional therapeutic modalities that can safely alter the microbiome—including selective FMT that has recently been shown to improve ICT in melanoma (9)—to make headway against this destructive disease.

No disclosures were reported.