In a development that could reshape how clinicians assess one of the most feared hospital-associated infections, researchers at the University of Virginia have identified a fecal protein that appears to signal which patients with Clostridioides difficile infection are likely to develop severe disease. The study, published in the journal Gut Pathogens, examined a panel of intestinal inflammation biomarkers in stool samples from 149 patients diagnosed with C. difficile infection, or CDI, and found that one protein in particular—S100 calcium-binding protein B, known simply as S100B—stood out from the pack as a measurable correlate of disease severity. The finding matters because CDI remains classified as an urgent threat in the United States, and clinicians currently lack reliable tools for predicting, at the time of diagnosis, which patients will slide into severe illness and which will recover without complication.

C. difficile is a spore-forming bacterium that thrives when the gut microbiome has been disrupted, most commonly by antibiotics. It produces toxins that damage the colonic epithelium, triggering an inflammatory cascade that ranges from mild diarrhea to life-threatening pseudomembranous colitis, toxic megacolon, and death. The severity of the disease, the researchers note, is strongly correlated with the degree of intestinal inflammation, which makes the gut lining a logical place to hunt for predictive signals. Because fecal specimens sit in direct contact with the inflamed mucosa, they can carry molecular fingerprints of the local immune battle—enzymes, cytokines, and damage-associated proteins shed into the intestinal lumen. The Virginia team set out to determine whether any of these fingerprints could do what blood-based markers have struggled to do: stratify CDI patients by likely outcome early in the course of infection.

The study design was straightforward but rigorous. Between April 2021 and February 2022, the investigators collected fecal samples from patients diagnosed with CDI and measured the concentrations of six candidate biomarkers using enzyme-linked immunosorbent assays, or ELISA. The panel included S100B, myeloperoxidase (MPO), lipocalin-2 (Lcn2), calprotectin (S100A8/A9), interleukin-6 (IL-6), and brain-derived neurotrophic factor (BDNF). Each represents a different facet of intestinal inflammation: MPO is an enzyme released by neutrophils, the first responders of the innate immune system; Lcn2 is an antimicrobial protein that sequesters iron from invading bacteria; calprotectin is a widely used clinical marker of intestinal inflammation; IL-6 is a pro-inflammatory cytokine central to systemic immune signaling; and BDNF, better known for its role in the nervous system, has emerged as a marker of mucosal stress and neuroimmune interaction. In parallel, the team cultured C. difficile isolates from matched specimens and used quantitative PCR to test for the binary toxin genes cdtA and cdtB, which are associated with hypervirulent strains.

The patient cohort painted a sobering picture of CDI’s clinical burden. Most patients were female—59.06 percent, or 88 of 149—with a mean age of 61.4 years, ranging from 21 to 96. More than four in ten patients, 42.28 percent, exhibited severe disease. Nearly three in ten, 29.53 percent, experienced recurrent CDI, the frustrating pattern in which the infection returns after apparent cure. Fifty-nine percent of patients were infected with a binary toxin-positive strain, and 19.46 percent—nearly one in five—died following their infection. These figures underscore why better prognostic markers are so urgently needed: a substantial fraction of patients arrive at the clinic looking deceptively well before deteriorating.

Against this clinical backdrop, the biomarker results delivered a surprise. Of the six proteins tested, only S100B showed a statistically significant increase in fecal levels among patients with severe CDI. The other five—IL-6, MPO, Lcn2, calprotectin, and BDNF—did not individually distinguish severe from non-severe disease in this cohort. When the researchers analyzed S100B’s diagnostic performance using receiver operating characteristic analysis, they found an area under the curve of 0.6253 with a p-value of 0.0091, indicating a statistically significant, though modest, association with severe disease. An AUC of 0.6253 suggests the marker performs better than chance but is not, on its own, a definitive diagnostic instrument. Importantly, fecal S100B was not associated with post-CDI mortality or with recurrent disease status, and patients with severe CDI were simply more likely to show elevated fecal levels of the protein.

S100B is an intriguing molecule in this context. It belongs to the S100 family of calcium-binding proteins, which participate in calcium signal transduction and regulate a wide range of cellular processes including inflammation, cell proliferation, and tissue repair. S100B itself is produced predominantly by astrocytes in the central nervous system, where elevated serum levels serve as a marker of brain injury, but it is also expressed by adipocytes, chondrocytes, and cells of the immune system. Its presence in feces during colonic inflammation suggests it may be released by activated glia-like enteric cells or infiltrating immune populations in the gut, though the precise cellular source during CDI remains an open question that the authors of the study and other investigators will likely pursue. The fact that it outperformed established inflammation markers like calprotectin and MPO in this specific setting hints that S100B may capture a dimension of CDI-related tissue damage that conventional markers miss.

One of the most practically significant findings, however, concerned the biomarker panel as a whole. When the researchers combined all six proteins—S100B, IL-6, MPO, Lcn2, S100A8/A9, and BDNF—the combined signature was associated not only with severe disease but also with recurrence and mortality after CDI. This multiplex approach reflects a growing consensus in biomarker research that single molecules rarely capture the full complexity of infectious disease outcomes. CDI severity emerges from an interplay between bacterial virulence factors, the host immune response, the integrity of the epithelial barrier, and the composition of the residual microbiome. A composite fecal inflammatory signature, measured cheaply by ELISA from a routine stool specimen, could in principle give clinicians a multidimensional risk profile at the bedside without invasive procedures.

The detection of binary toxin genes by qPCR added a microbiological dimension to the analysis. Binary toxin, encoded by the cdtA and cdtB genes, is a hallmark of hypervirulent strains such as the epidemic NAP1/BI/ribotype 027 lineage, and it acts by ADP-ribosylating actin, disrupting the cytoskeleton of host cells and enhancing toxin adherence. With 59 percent of the cohort infected by binary toxin-positive strains, the study population carried a substantial burden of potentially hypervirulent organisms, making the search for host-side prognostic markers all the more relevant. Understanding how bacterial virulence and host inflammation markers interact to produce severe outcomes remains a central challenge in CDI research, and this study provides one piece of that puzzle.

The implications for clinical practice, while promising, come with caveats that the researchers themselves acknowledge through the structure of their findings. An AUC of 0.6253 means S100B alone cannot yet serve as a stand-alone decision-making tool; it is best understood as a component of a broader prognostic framework. The study was also conducted at a single center over roughly ten months of enrollment, and validation in larger, multi-center, and more diverse cohorts will be essential before fecal S100B testing migrates from the research bench to the clinical laboratory. Furthermore, the study’s cross-sectional design links elevated S100B with severity at the time of diagnosis but does not fully resolve whether the protein plays a mechanistic role in disease progression or is simply a byproduct of tissue injury. Still, the observation that a single fecal protein outperformed five established inflammatory markers in predicting severity is striking, and it opens a clear path for follow-up studies.

For patients, the stakes are considerable. CDI affects hundreds of thousands of Americans each year, imposes billions of dollars in healthcare costs, and carries a one-year mortality that can approach 20 percent in elderly populations—figures consistent with the 19.46 percent post-infection death rate observed in this cohort. Current severity scoring relies on clinical parameters such as white blood cell counts, serum creatinine, hypotension, and shock, supplemented by imaging when complications are suspected. A stool-based molecular test could be performed on the very specimen already collected for C. difficile diagnostic testing, adding prognostic information at essentially no additional burden to the patient. If subsequent research confirms these findings and refines the composite biomarker panel, clinicians could one day use a routine stool sample to identify high-risk patients on day one, escalating treatment, considering early monoclonal antibody therapy such as bezlotoxumab, or intensifying monitoring before complications arise.

The University of Virginia team—led by first authors Laasya Doppalapudi and Matthew Z. Behm, with senior authors Cirle A. Warren and Deiziane V. S. Costa, alongside statisticians and clinician-investigators including Xin-Qun Wang, Mônica J. Rodrigues-Jesus, Ayesha E. Tora, and Jae Hyun Shin—published their work as open access, making the data freely available to the research community. As the search for fecal biomarkers of intestinal inflammation accelerates across gastroenterology, from inflammatory bowel disease to irritable bowel syndrome, this study adds a new candidate to the list and demonstrates the value of systematically comparing multiple markers within a single well-characterized patient cohort. Whether S100B ultimately earns a place in clinical guidelines will depend on replication, but the message of the study is clear: the answers to some of CDI’s hardest prognostic questions may be sitting in the sample the laboratory already has in hand.