Resphera Biosciences

Abstract Introduction: The overall response rates of immune-checkpoint inhibitors (ICI) in head and neck squamous cell carcinoma (HNSCC) range from only 13%-18%. The potential immune-related toxicities of ICI and their high cost have urged the search for prospectively validated predictive biomarkers of response to ICI in HNSCC patients. We and others have previously shown that the oral microbiome, including Fusobacterium nucleatum, Lactobacillus gasseri/johnsonii, and Lactobacillus vaginalis in saliva from Human Papilloma Virus (HPV) positive and HPV negative patients, may be crucial for tumor development and progression, treatment-related toxicity, and disease recurrence. Here, we explore the links between the tissue microbiome, PD-1 expression, and HPV status in HNSCC patients. Methods: We used high-resolution microbiome profiling to investigate microbial species present in samples from PD-1 positive and PD-1 negative HNSCC patients. We performed 16S rRNA sequencing in tissue samples from a discovery cohort of 23 HNSCC patients and 8 normal epithelium controls. We subsequently used the Axiom Microbiome Array (ThermoFisher), which harbors a total of 1.38 million probes, to assess microbial composition in a validation cohort of nine (9) PD-1 positive HNSCC, nine (9) PD-1 negative HNSSC, two (2) PD-1 positive normal oral epithelium, and two (2) PD-1 negative normal oral epithelium tissue samples. Results: We found a statistically significant difference in total microbial richness when stratifying cancer and normal tissue by PD-1 expression and HPV status (P< 0.0001), while strata of cancer versus normal tissue displayed no statistical difference in observed microbial richness. We found no difference in species abundance when comparing normal and tumor epithelium. We observed a reduction in overall microbiome composition among PD-1 positive patients when compared to PD-1 negative patients, and an increase in abundance of Bacteroidetes, Firmicutes and Fusobacteriacommunities in PD-1 positive patients, including Fusobacteria nucleatum and Lactobacteria gasseri-johnsonii. Conclusion: Microbiome modulation of PD-1/PD-L1 pathway induction and downstream maintenance of immune tolerance may directly impact the response of HPV positive and HPV negative patients to ICI. The overall biota reduction paired with an increase in specific bacterial communities we found in PD-1 positive patients suggest an important immunomodulatory role, which warrant further study to determine if the tissue microbiome plays a role in ICI treatment response among HNSCC patients. Citation Format: Laura Palmieri, Kelvin Navarro, Anne Jedlicka, Amanda Dziedzic, James White, Mariana Brait, Wayne Koch, Filipa Godoy-Vitorino, David Sidransky, Rafael E. Guerrero-Preston. Fusobacteria nucleatum and Lactobacteria gasseri-johnsonii associated to PD-1 expression in head and neck cancer patients [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2025; Part 2 (Late-Breaking, Clinical Trial, and Invited Abstracts); 2025 Apr 25-30; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2025;85(8_Suppl_2):Abstract nr LB133.

BackgroundFeatures of the gut microbiome (GM), such as increased alpha diversity and abundance of certain genera, have been associated with improved outcomes in patients with advanced non-small cell lung cancer (NSCLC) treated with immune checkpoint inhibitors (ICI).1–3 However, the association between baseline GM characteristics, or longitudinal GM changes, and outcome in patients receiving neoadjuvant ICI for earlier stage NSCLC is poorly understood. We studied the relationship between GM features, at baseline and on serial sampling, and clinical outcomes in patients receiving neoadjuvant ICI for resectable NSCLC.MethodsPatients treated with neoadjuvant ICI for NSCLC on two prospective clinical trials (J1772, J1414) had fecal samples collected at baseline and serially for GM analysis, together with prospective annotation of clinical outcomes. Fecal samples were analyzed using 16S rRNA amplicon sequencing of DNA at baseline and longitudinally thereafter. Baseline GM features were correlated with major pathologic response (MPR) (≤10% viable tumor in resected tumor specimens), a validated surrogate endpoint for overall survival. Using a bootstrap analysis from treatment start until 180 days of therapy, longitudinal GM features were used to create a ‘self-similarity’ score, incorporating the top 10 bacterial shared species in individual patient samples, and correlated with outcomes. Functional analysis using DESeq was performed to elucidate metabolic pathways implicated in these GM changes.ResultsWe identified 23 patients (n=86 samples, n=1–13 samples per patient), with 22 (96%) having at least one fecal sample within 42 days of commencing treatment and classified as ‘baseline’. Neoadjuvant regimens included nivolumab (n=10), nivolumab+ipilumumab (n=6), durvalumab+radiation (n=5) and nivolumab+platinum doublet chemotherapy (n=2). Twenty (87%) underwent surgical resection, and in 3 patients disease was inoperable. Five patients (22%) achieved MPR and 18 (78%) did not have a recurrence at 24 months. Baseline alpha diversity, measured by Simpson’s reciprocal index, was significantly lower among those with MPR (11.1, range 2.8 - 15.5) vs without (19.5, range 4.6 – 25.5) (p=0.02). Patients who achieved MPR had greater self-similarity at 180 days vs. no MPR (96.1 vs 92.1%, p=0.001). Metabolic pathways involved in short-chain fatty acid (SCFA) production were found to be more abundant in those with MPR vs without.ConclusionsIn a cohort of prospectively assessed GM samples from patients in neoadjuvant ICI clinical trials for earlier stage NSCLC, our findings suggest that decreased baseline alpha diversity and greater longitudinal stability in the GM are associated with MPR. A microbiome conducive to SCFA production is also associated with MPR.Trial RegistrationJ1772 had registration NCT03237377 J1414 had registration NCT02259621.ReferencesRouty B, Le Chatelier E, Derosa L, Duong CPM, Alou MT, Daillere R, et al. Gut microbiome influences efficacy of PD-1-based immunotherapy against epithelial tumors. Science. 2018;359(6371):91–7.Jin Y, Dong H, Xia L, Yang Y, Zhu Y, Shen Y, et al. The diversity of gut microbiome is associated with favorable responses to anti-programmed death 1 immunotherapy in chinese patients with NSCLC. J Thorac Oncol. 2019;14(8):1378–89.Hakozaki T, Richard C, Elkrief A, Hosomi Y, Benlaifaoui M, Mimpen I, et al. The gut microbiome associates with immune checkpoint inhibition outcomes in patients with advanced non-small cell lung cancer. Cancer Immunol Res. 2020;8(10):1243–50.Ethics ApprovalJ1772 received approval from the Johns Hopkins IRB, ID number IRB00127418, participants gave informed consent before taking part. J1414 received approval from the Johns Hopkins IRB, ID number NA_00092076, participants gave informed consent before taking part.

Background: Preclinical and clinical data indicate that chemoradiotherapy (CRT) in combination with checkpoint inhibitors may prime an anti-tumor immunological response in esophageal cancer. However, responses to neoadjuvant therapy can vary widely and the key biomarkers to determine response remain poorly understood. The fecal microbiome is a novel and potentially modifiable biomarker of immunotherapy response, and both fecal and tumor microbes have been found to associate with outcomes in esophageal cancer. Methods: Fecal and tumor samples were collected from patients with stage II-III resectable esophageal or gastroesophageal junction carcinoma treated with neoadjuvant immune checkpoint inhibitors (ICIs) plus CRT prior to surgical resection. Microbiome profiles were analyzed by 16S rRNA amplicon sequencing and taxonomic data were integrated with fecal metabolite analysis to assess microbial function. Results: The fecal microbiome of patients with pathological complete response (PCR) grouped in distinct clusters compared to patients with residual viable tumor (RVT) by Bray-Curtis diversity metric. Integrated taxonomic and metabolomic analysis of fecal samples identified a sphingolipid and primary bile acid as enriched in the PCR, the levels of which correlated with several bacterial species: Roseburis inulinivorans, Ruminococcus callidus, and Fusicantenibacter saccharivorans. Analysis of the tumor microbiome profiles identified several bacterial genera previously associated with esophageal tumors, including Streptococcus and Veillonella. Conclusions: These results further characterize the fecal and tumor microbiome of patients with operable esophageal cancer and identify specific microbes and metabolites that may help elucidate how microbes contribute to tumor response with neoadjuvant CRT combined with ICI.

<div>Abstract<p>Colorectal cancer is multifaceted, with subtypes defined by genetic, histologic, and immunologic features that are potentially influenced by inflammation, mutagens, and/or microbiota. Colorectal cancers with activating mutations in <i>BRAF</i> are associated with distinct clinical characteristics, although the pathogenesis is not well understood. The Wnt-driven multiple intestinal neoplasia (Min<sup>ApcΔ716/+</sup>) enterotoxigenic <i>Bacteroides fragilis</i> (ETBF) murine model is characterized by IL17-dependent, distal colon adenomas. Herein, we report that the addition of the <i>BRAF</i><sup>V600E</sup> mutation to this model results in the emergence of a distinct locus of midcolon tumors. In ETBF-colonized <i>BRAF</i><sup>V600E</sup><i>Lgr5</i><sup>Cre</sup>Min (BLM) mice, tumors have similarities to human <i>BRAF</i><sup>V600E</sup> tumors, including histology, CpG island DNA hypermethylation, and immune signatures. In comparison to Min ETBF tumors, BLM ETBF tumors are infiltrated by CD8<sup>+</sup> T cells, express IFNγ signatures, and are sensitive to anti–PD-L1 treatment. These results provide direct evidence for critical roles of host genetic and microbiota interactions in colorectal cancer pathogenesis and sensitivity to immunotherapy.</p>Significance:<p>Colorectal cancers with <i>BRAF</i> mutations have distinct characteristics. We present evidence of specific colorectal cancer gene–microbial interactions in which colonization with toxigenic bacteria drives tumorigenesis in <i>BRAF</i><sup>V600E</sup><i>Lgr5</i><sup>Cre</sup>Min mice, wherein tumors phenocopy aspects of human <i>BRAF</i>-mutated tumors and have a distinct IFNγ-dominant immune microenvironment uniquely responsive to immune checkpoint blockade.</p><p><i>This article is highlighted in the In This Issue feature, p. 1601</i></p></div>

Defining the complex role of the microbiome in colorectal cancer (CRC) and the discovery of novel, pro-tumorigenic microbes are areas of active investigation. In the present study, culturing and reassociation experiments revealed that toxigenic strains of Clostridioides difficile drove the tumorigenic phenotype of a subset of CRC patient-derived mucosal slurries in germ-free ApcMin/+ mice. Tumorigenesis was dependent on the C. difficile toxin TcdB and was associated with induction of Wnt signaling, reactive oxygen species, and pro-tumorigenic mucosal immune responses marked by infiltration of activated myeloid cells and interleukin-17 (IL-17)-producing lymphoid and innate lymphoid cell subsets. These findings suggest that chronic colonization with toxigenic C. difficile is a potential driver of CRC in patients.

Abstract Batch effects in microbiome data arise from differential processing of specimens and can lead to spurious findings and obscure true signals. Most existing strategies for mitigating batch effects rely on approaches designed for genomic analysis, failing to address the zero-inflated and over-dispersed microbiome data. Strategies tailored for microbiome data are restricted to association testing, failing to allow other analytic goals such as visualization. We develop the Conditional Quantile Regression (ConQuR) approach to remove microbiome batch effects using a two-part quantile regression model. It is a fundamental advancement in the field because it is the first comprehensive method that accommodates the complex distributions of microbial read counts, and it generates batch-removed zero-inflated read counts that can be used in and benefit all usual subsequent analyses. We apply ConQuR to real microbiome data sets and demonstrate its state-of-the-art performance in removing batch effects while preserving or even amplifying the signals of interest.

Abstract There is an urgent need to identify therapies to prevent and treat SARS-CoV-2 infection. We performed a statistical evaluation of in vitro gene expression profiles reflecting exposure to 1,835 drugs, and found topoisomerase inhibitors and PI3K/mTOR pathway inhibitors among the strongest candidates for reduced expression of ACE2, a host gene associated with SARS-CoV-2 infection. Retrospective clinical data suggest that patients on these agents may be less likely to test positive for SARS-CoV-2.

Longitudinal analysis of Crohn's disease (CD) incidence has identified an inverse correlation with helminth infestation and recent studies have revealed that intestinal tuft cell hyperplasia is critical for helminth response. Tuft cell frequency was decreased in the inflamed ilea of CD patients and a mouse model of TNFα-induced Crohn's-like ileitis (TNFΔARE). Single-cell RNA sequencing paired with unbiased differential trajectory analysis of the tuft cell lineage in a genetic model of tuft cell hyperplasia (AtohKO) demonstrated that the tuft cell lineage had increased tricarboxylic acid (TCA) cycle gene signatures. Commensal microbiome-derived succinate was detected in the ileal lumen of these animals while microbiome depletion suppressed tuft cell hyperplasia. Therapeutic succinate treatment in TNFΔARE animals reduced pathology in correlation with induced tuft cell specification. We provide evidence implicating the modulatory role of intestinal tuft cells in chronic intestinal inflammation, which could facilitate leveraging this rare and elusive cell type for CD treatment.

Mucus-invasive bacterial biofilms are identified on the colon mucosa of approximately 50% of colorectal cancer (CRC) patients and approximately 13% of healthy subjects. Here, we test the hypothesis that human colon biofilms comprise microbial communities that are carcinogenic in CRC mouse models. Homogenates of human biofilm-positive colon mucosa were prepared from tumor patients (tumor and paired normal tissues from surgical resections) or biofilm-positive biopsies from healthy individuals undergoing screening colonoscopy; homogenates of biofilm-negative colon biopsies from healthy individuals undergoing screening colonoscopy served as controls. After 12 weeks, biofilm-positive, but not biofilm-negative, human colon mucosal homogenates induced colon tumor formation in 3 mouse colon tumor models (germ-free ApcMinΔ850/+;Il10-/- or ApcMinΔ850/+ and specific pathogen-free ApcMinΔ716/+ mice). Remarkably, biofilm-positive communities from healthy colonoscopy biopsies induced colon inflammation and tumors similarly to biofilm-positive tumor tissues. By 1 week, biofilm-positive human tumor homogenates, but not healthy biopsies, displayed consistent bacterial mucus invasion and biofilm formation in mouse colons. 16S rRNA gene sequencing and RNA-Seq analyses identified compositional and functional microbiota differences between mice colonized with biofilm-positive and biofilm-negative communities. These results suggest human colon mucosal biofilms, whether from tumor hosts or healthy individuals undergoing screening colonoscopy, are carcinogenic in murine models of CRC.