Acute radiation-induced esophagitis (ARIE) is one of the most common and debilitating tissue toxicities associated with radiotherapy in patients who received thoracic radiation, especially for esophageal cancer (EC) (Bradley and Movsas, 2004). ARIE is characterized by inflammation of the esophageal mucosa, with dysphagia, odynophagia, retrosternal pain, and other symptoms. The development of ARIE severely affects the quality-of-life of these patients. It also increases the need for narcotic analgesics and parenteral nutrition, and may even lead to treatment interruption or early treatment termination (Murro and Jakate, 2015). However, there are currently no effective non-invasive biomarkers or mucosal protective strategies for the management of ARIE.

The gut microbiota is the largest bank of microorganisms in the human body, the collective genes of which are 150-times larger than the human genome (Qin et al., 2010). Previous studies reported that gut microbiota can influence the occurrence and development of diseases in other organs via the gut-liver axis (Albillos et al., 2020), gut-brain axis (Cryan et al., 2019), gut-lung axis (Budden et al., 2017), and gut-bone axis (Zaiss et al., 2019). Recent research also demonstrated that patients with EC have altered gut microbiota (Deng et al., 2021).

Some studies suggested that the pathogenesis of radiotherapy- or chemotherapy-induced mucositis can be characterized by five major phases (Sonis, 2004) — initiation, upregulation, primary damage response, signal amplification, and ulceration — followed by wound healing. This five-phase model does not consider the role of gut microbiota in the pathophysiology of mucositis. However, with the development of high-throughput sequencing technology, increasing evidence has shown that commensal gut microbes play roles in several local and systemic inflammatory diseases, such as inflammatory bowel disease, radiation-induced enteritis, autoimmune disorders, allergic diseases, obesity, and diabetes (Clemente et al., 2018; Gerassy-Vainberg et al., 2018; Peroni et al., 2020; Scheithauer et al., 2020). Nevertheless, no study has yet reported the association of ARIE with the characteristics of gut microbiota.

Hence, we performed a longitudinal evaluation of the severity of ARIE in a cohort of patients with EC who received radiotherapy with or without chemotherapy, and examined fecal samples of these patients at the start, middle, and end of the radiotherapy regimen. Based on high-throughput 16s rRNA gene sequencing and bioinformatics analysis, we studied changes in the gut bacterial profile during the course of treatment, and examined the association between the severity of ARIE and the characteristics of the gut microbial community.

Analysis of the baseline demographic characteristics of all 49 patients indicated that the groups with different ARIE severity were similar in terms of age, sex, BMI, tobacco use, alcohol use, comorbidities, pathology, radiation dose, clinical classification, or receipt of combined chemotherapy (Supplementary Table 1). The patients were 48–87 years old (average: 64.8), 79.59% were male, and 91.84% had squamous cell carcinoma. According to the eighth edition of the American Joint Committee on Cancer staging system for EC, the most common T classification was T3/T4 (77.55%) and the most common N classification was N0/N1 (69.39%). The median radiation dose was 6000 cGy and 42 patients (85.71%) received concurrent chemoradiotherapy. Analysis of the intervention treatments (use of analgesics, anti-inflammatory agents, and parenteral nutrition) showed significant differences among the groups.

In total, we generated 10,036,517 paired-end reads of high-quality sequences (average 68,276 per sample). There were 549 OTUs at a 97% similarity level. Analyses of the rarefaction and rank abundance curves showed that these curves tended flatten as the number of sample sequences increased, indicating the sample had sufficient richness and uniformity, and the sequencing depth was satisfactory (Supplementary Figures 1A,B). The species accumulation boxplot showed that the gene richness approached saturation as the sample size increased, indicating the number of samples was sufficient to resolve most of the genera present (Supplementary Figure 1C). These results thus indicated the sample size and sequencing depth were sufficient.

Esophageal cancer is the sixth leading cause of cancer-related deaths and has the seventh highest of incidence of all cancers worldwide (Sung et al., 2021). Most patients have locally advanced disease at diagnosis, and radiotherapy is one of the most important therapeutic modalities for these patients (Ajani et al., 2019). The esophagus consists of rapidly proliferating squamous epithelial cells that course longitudinally through the entire extent of the mediastinum, and is highly vulnerable to radiation-induced injury. ARIE remains the main dose-limiting factor for radiotherapy of patients with EC. Most studies of ARIE have focused on radiotherapy dose-volume parameters (Caglar et al., 2010; Werner-Wasik et al., 2010). However, no previous research investigated the possible role of gut microbiota on the development of ARIE.

In addition to the widely studied effect of the gut microbiota on radiation-induced enteritis (Gerassy-Vainberg et al., 2018), recent studies confirmed that gut microbiota also regulates radiation-induced injury of extraintestinal organs, such as cardiopulmonary and hematopoietic damage (Chen et al., 2021; Kordahi and Chassaing, 2021). Analysis of samples collected at the “middle” of the radiotherapy regimen, that is, when ARIE occurred, showed that the RTOG 0 and RTOG 1 groups both had greater gut microbial richness than the RTOG 2 and RTOG 3 groups, and a PCoA plot also showed that the samples were clustered by groups. This indicated that the severity of ARIE was associated with the gut microbial community structure. Further analysis indicated that all genera which increased significantly as mucositis severity increased were in the phylum Proteobacteria, and that several members of the phyla Firmicutes, Cyanobacteria, Chloroflexi, and Actinobacteria decreased significantly as mucositis severity increased. A systematic literature review reported that an imbalanced gut microbiota often arises from a sustained increase in the abundance of bacteria in the phylum Proteobacteria, and the natural human gut microbiota normally contains only a minor proportion bacteria of this phylum (Shin et al., 2015). Although most of the bacteria that had high correlations with mucositis severity had relatively low abundances, they could still play major roles in causing ARIE. In particular, the “keystone pathogen” hypothesis (Hajishengallis et al., 2012) proposes that certain low-abundance bacteria, such as Porphyromonas gingivalis in the human periodontium or enterotoxigenic Bacteroides fragilis in the human colon, can provoke inflammation by remodeling the entire microbiota via direct effects on different microbes or indirect effects due to host modulation.

Previous research showed that pelvic radiation induced gut dysbiosis, and that the altered gut microbiota increased the susceptibility to inflammation (Gerassy-Vainberg et al., 2018). This raises the question of whether chest radiation can also alter gut microbiota. We therefore collected patients’ feces longitudinally during the radiotherapy regimen to examine the changes in their gut microbiota. Our results showed that Chao1 indexes decreased significantly as the treatment progressed, indicating a decline in community richness of gut microbiota. It is worth noting that the majority of patients in our study (42/49) received chemotherapy concurrent with radiotherapy. Previous studies found that chemotherapy-induced gut microbiota dysbiosis can impair mucosal homeostasis (van Vliet et al., 2010; Wei et al., 2021). However, we found no significant difference in the alpha diversity indexes (Shannon and Chao1) of gut microbiota in the mid-treatment samples of patients who received chemoradiotherapy relative to those who received radiotherapy alone. Fisher’s exact test also confirmed that the incidence of severe ARIE was not statistically different between these two groups, although only 7 patients were in the group that received radiotherapy alone. It is generally believed that a stable and diverse gut microbiota is essential to the normal physiological processes and mucosal immune function of the host. Normal microbiota can prevent colonization by foreign pathogens and the overgrowth of local pathogens, and an altered composition of commensal microbes may cause immune imbalance and increase the risk of multiple diseases (Kayama et al., 2020).

Twenty of the 49 patients (40.82%) in our study developed severe ARIE, comparable to the incidence of severe ARIE in EC patients who received chemoradiotherapy in previous studies (Crosby et al., 2013; Yu et al., 2021). We hypothesized that the distribution of gut microbiota prior to radiotherapy determines a patient’s susceptibility to mucosal radiation injury. Our analysis of fecal samples of all patients prior to radiotherapy indicated that the group which subsequently developed the severe ARIE had a lower initial richness of gut microbiota. Furthermore, compared to the mild ARIE subgroup, fecal samples from the severe ARIE subgroup had significantly higher abundances of Fusobacterium and Burkholderiaceae, both of which are associated with harmful infections in humans. In particular, Fusobacterium has the ability to produce a variety of virulence factors and Fusobacterium nucleatum, one of the most prevalent species isolated from human infections, has been implicated in atherosclerosis, adverse pregnancy outcomes, rheumatoid arthritis, and organ abscesses and infections (Han, 2015; Brennan and Garrett, 2019). Mechanistic studies of F. nucleatum showed it promoted nuclear factor kappa-light-chain-enhancer of activated B cell (NF-κB) activation, and indirectly led to a proinflammatory genetic signature that included increased expression of interleukin-1β (IL-1β), interleukin-6 (IL-6), interleukin-8 (IL-8), and tumor necrosis factor-α (TNFα) (Kostic et al., 2013). A previous study demonstrated that F. nucleatum secreted outer membrane vesicles (OMVs) (Liu et al., 2019), “buds” of the outer membrane that typically contain antigenic components that can activate Toll-like receptors (TLRs) on epithelial cells or immune cells. TLR activation is also linked to the activation of downstream targets, such as extracellular signal-regulated kinase (ERK), cAMP-response element binding protein (CREB), and NF-κB. Fusobacteria also encode Fusobacterium adhesin A (FadA), which likely provides it with the ability to invade epithelial and endothelial cells, and increase the permeability of the endothelial layer (Han, 2015). The Burkholderiaceae is a well-known family of heterotrophic bacteria that can colonize the guts of immunosuppressed patients, and this family includes species associated with several animal and human diseases of varying severity (Yang et al., 2016; Ye et al., 2021). Further research is needed to identify the exact pathogenic species of Fusobacterium and Burkholderiaceae that are responsible for the effects observed here.

We found that patients in the mild ARIE subgroup had significantly higher abundances of Roseburia, Veillonella, Prevotella_9, Megasphaera, and Ruminococcus_2, genera associated with the production of short-chain fatty acids (SCFAs) (de la Cuesta-Zuluaga et al., 2017; Louis and Flint, 2017; D’Amico et al., 2021), including butyrate, propionate, and acetate. Many studies have examined butyrate due to its anti-inflammatory and anti-cancer effects. In particular, butyrate is a main energy source for normal epithelial cells, but not for cancer cells (Vatanen et al., 2016a). Interestingly, previous research reported that the presence of Fusobacterium spp. in the gut microbiota was associated with reduced butyrate production (Hertel et al., 2021). Researchers have proposed two main mechanisms for the anti-inflammatory and immunomodulatory effects of butyrate. First, butyrate can inhibit the phosphorylation of the NF-κB p65 signaling pathway in host immune cells by binding to G-protein-coupled receptors (GPR43, GPR41, and GPR109A), thereby blocking inflammatory responses and suppressing the release of TNF-α and IL6 (Lewis et al., 2010). Second, butyrate can inhibit histone deacetylase (HDAC), thus increasing Forkhead box protein P3 (FOXP3) expression and inducing Treg cell polarization; these cells are important producers of IL-10, a potent anti-inflammatory cytokine (Furusawa et al., 2013). Butyrate also assists in maintaining the gut barrier function by regulating the expression of mucin 2 and tight-junction proteins, such as claudin-1 and Zonula Occludens-1, which prevent bacterial and lipopolysaccharide (LPS) translocation and systemic inflammation (Tajik et al., 2020; Zhang et al., 2021). It is also important that SCFAs can mediate a range of extra-intestinal effects and impact development of immune and inflammatory responses, both locally and at distant sites, after entering the circulation via active transport mediated by monocarboxylate transporters (Stumpff, 2018). The therapeutic potential of butyrate and other SCFAs for the management of ARIE should therefore be considered in further research.

Notably, we found that Klebsiella was also enriched in the fecal samples of patients in the mild ARIE subgroup. The LPS of Klebsiella is generally considered to be a toxin that promotes inflammation via TLR4-mediated NF-κB activation and the production of various inflammatory factors, such as TNF-α, IL-6, and IL-1β. However, Vatanen et al. (2016b) suggested that the lipid A domain of LPS, which has six acyl chains (produced by Enterobacteriaceae, including Klebsiella), may be a beneficial innate immune activator. Initial exposure to LPS from Enterobacteriaceae prevented TNF-α production after restimulation, referred to as induction of endotoxin tolerance. In contrast, the lipid A domain of LPS, which has four or five acyl chains (produced by Bacteroides), may inhibit innate immune activation and endotoxin tolerance (Vatanen et al., 2016a; D’Hennezel et al., 2017). Based on microbial signatures, we constructed ROC curves to examine genera that have potential for predicting the severity of ARIE. Notably, models that examined a single genus or all seven genera provided reliable predictions of severe ARIE (RTOG 2–3).

Previous research showed that secretion of IL-6 and TNF-α amplified leukocyte activation, and led to increased levels of WBCs and NEs (Fujishima and Aikawa, 1995). The NLR and LY level are also good indicators of bacteremia (de Jager et al., 2010). Recent research proposed use of the SII (platelets × NLR) as an integrated and novel inflammatory biomarker (Mazza et al., 2020). Researchers initially used the SII to assess the prognosis of patients with solid cancers and coronary heart disease, and now also use it as a measure of inflammation status. We found that Roseburia and Ruminococcus_2 had significant correlations with these indicators of clinical immune-inflammation. This supports the presence of potential relationships among the characteristics of gut microbiota, ARIE, and the systemic inflammatory response.

Our study identified an association between the characteristics of the gut microbial community of EC patients and the severity of ARIE, as well as changes of the gut microbial structure during chemoradiotherapy. We suggest that the pre-existing gut microbial community in these patients may be useful in predicting which patients are more likely to progress to severe ARIE. This study represents, as far as we know, the largest longitudinal cohort study of the association of human gut microbiota with ARIE, but future studies with larger sample sizes are needed to verify our results. Furthermore, we suggest a shotgun microbiota metagenomics approach to better understand the structure and function of the gut microbial community at the species level. Germ-free and gnotobiotic animals are also invaluable experimental tools for examination of host-microbe interactions.

In summary, our results showed that the characteristics of the gut microbial community of EC patients prior to radiotherapy correlated with subsequent ARIE severity. Microbiota-based strategies may be useful for the early prediction and prevention of severe ARIE of patients with EC.

The datasets presented in this study can be found in online repositories. The names of the repository/repositories and accession number(s) can be found below: FigShare, https://figshare.com/articles/dataset/allsample_txt/19235931; doi:10.6084/m9.figshare.19235931.

The current study was approved by the Ethics Committee of Fujian Medical University Cancer Hospital, Fuzhou, China (K2022-008-01). The patients/participants provided their written informed consent to participate in this study.

J-CL, Q-WY, and M-QL conceived the idea. M-QL, Y-HW, and H-CL performed the experiments. JY, L-YL, and Y-LY analyzed the data. M-QL wrote the manuscript. J-CL, Q-WY, and Y-HW revised the manuscript. All authors helped to perform the research.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.