The recurrence rate following complete surgical resection of primary non-small cell lung cancer is as high as 55%, yet no approach currently exists to evaluate the risk of local recurrence. The premetastatic paradigm is the recognition that metastasis is preceded by reprogramming naïve tissues to prime a microenvironment for tumor cell survival and subsequent reactivation. Identification of biomarkers of the pre-metastatic niche would allow us to evaluate a patient's risk of local relapse in the normal lung parenchyma surrounding the resected tumor. We designed a workflow incorporating in vivo modelling, radiology, and deep learning-guided three-dimensional (3D) imaging, spatial proteomics, and transcriptomics to identify previously unreported signals associated with the early transformation of the lung parenchyma announcing regional metastasis. We curated biorepository spanning timepoints before and after resection of primary Lewis Lung Carcinoma (LLC) tumors. Using radiology and cellular resolution 3D histology, we calculated the number and distribution of metastases in mouse lungs and developed an algorithm to guide placement of spatial proteomics and transcriptomics to regions containing early micro-metastases and the pre-metastatic microenvironment. Molecular and tissue features associated with presence, size, and location of metastases guided the identification of both myeloid (F4/80) and senescent (p16/p21) cell signatures in the premetastatic and metastatic environments. Finally, multiparametric flow cytometry of metastatic lungs in a senescence reporter GEMM (tdTomato-p16 INKA mice) resolved senescent cells including alveolar macrophages as the cellular phenotypes associated with these early premetastatic signatures. Altogether, this work highlights a novel AI-assisted approach for detection of biomarkers of tissue remodeling during lung cancer invasion.

BACKGROUND:Ustekinumab is a monoclonal antibody therapy targeting interleukin-12 and interleukin-23 for the treatment of inflammatory bowel diseases, including ulcerative colitis (UC). While these pathways remain quite attractive for UC therapy, response to ustekinumab can be variable. There is an urgent need to better understand the underlying mucosal immune alterations associated with treatment to guide therapy decisions. This study aims to examine the mucosal immune signatures in individuals with UC with variable treatment response to ustekinumab. METHODS:Sigmoid colon tissue from individuals treated with ustekinumab were analyzed using a multimodal approach. Single-cell RNA and T cell receptor sequencing was performed on mucosal biopsies. In a subset of these patients, multiparameter flow cytometry and spatial transcriptomics was also completed on matched pre- and post-treatment tissue samples. Key findings were also validated on a larger cohort using immunohistochemistry. RESULTS:Ustekinumab significantly altered the frequency and phenotype of mucosal regulatory T cells (Tregs). Nonresponders to ustekinumab had a higher frequency of Tregs that expressed OX40 and GITR, which is associated with decreased suppressive abilities. In contrast, responders had Tregs with elevated GPR15 and reduced expression of the kinase PIM2, which can alter Treg stability and function. Additionally, T helper 17 cells in nonresponders demonstrated an enhanced proinflammatory gene expression profile. CONCLUSION:Nonresponse to ustekinumab in UC is linked to a mucosal immune environment enriched with proinflammatory T cell phenotypes and impaired regulatory T cell function. These findings suggest that Tregs are both targets and potential biomarkers of ustekinumab response, with their phenotypic and transcriptional features providing insight into mechanisms of therapeutic resistance.

Abstract Introduction: Mucosal-associated invariant T (MAIT) cells are innate-like T lymphocytes expressing a semi-invariant T-cell receptor (TCR) α-chain (TRAV1-2 joined to TRAJ33/12/20) paired with a restricted TCR β repertoire. Unlike conventional T cells that respond to peptide antigens, MAIT cells recognize and kill infected or transformed cells presenting non-polymorphic MHC class-I (MR1) molecule covalently loaded with microbial riboflavin (vitamin B2) biosynthetic intermediates. A key riboflavin precursor, 5-amino-6-D-ribitylaminouracil (5-A-RU), undergoes non-enzymatic condensation to form potent MR1 ligands such as 5-OP-RU, which can prime tumor cells to express MR1. These features make MAIT TCRs an attractive platform for adoptive T-cell therapy across cancer types, independent of HLA molecules or tumor genetics. Methods: Human CD3+ T cells isolated from healthy donor peripheral blood mononuclear cells (PBMC) were engineered to express MAIT TCRs (TRAV1-2/TRAJ12 or TRAV1-2/TRAJ33) using a CRISPR-Cas12-mediated homology-directed repair (HDR) strategy. TCR knockout (TRAC/TRBC double-knockout) T cells lacking any introduced TCR served as controls. Antitumor activity was assessed in vitro and in vivo. Cytotoxicity against the human B-cell acute lymphoblastic leukemia line NALM6 was measured by flow cytometry. For in vivo evaluation, luciferase-expressing NALM6 cells were injected intravenously into NOD/SCID/ Il2rg−/− (NSG) mice, followed by infusion of engineered T cells. A 5-A-RU prodrug was administered intraperitoneally. Tumor burden was monitored by bioluminescent imaging. Results: In vitro, MAIT TCR-engineered human T cells exhibited distinct baseline cytotoxicity against NALM6 cells (approximately 60-70% for J12 and ∼10% for J33 in the absence of ligand). Addition of 5-A-RU resulted in near-complete tumor cell elimination for both MAIT TCR variants. In xenograft models, neither J12 nor J33 MAIT TCR T cells showed significant antitumor activity without exogenous 5-A-RU. However, intraperitoneal delivery of the 5-A-RU prodrug had no effect in TCR-knockout controls but markedly enhanced the antitumor activity of both MAIT TCR-engineered T-cell products. Mice receiving MAIT TCR T cells plus 5-A-RU showed delayed tumor progression and significantly improved survival compared with controls. Conclusion: MAIT TCR engineering confers potent antitumor activity to human T cells, which is strongly augmented by riboflavin pathway metabolites such as 5-A-RU. These findings support a model in which metabolite supplementation prime tumor cells to express MR1 and redirect MAIT TCR-engineered T cells cytotoxic activity towards tumor. Further studies are underway to define the therapeutic window, evaluate potential toxicities, and identify the breadth of tumor types responsive to MAIT TCR-based adoptive cell therapy. Citation Format: Ying Zheng, Pakhi Birla, Wanting Shan, Samuel Huang, Franck Housseau, Drew Pardoll. Riboflavin biosynthesis byproducts increase the vulnerability of cancer to adoptive MAIT TCR-engineered T-cell transfer [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2026; Part 1 (Regular Abstracts); 2026 Apr 17-22; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2026;86(7 Suppl):Abstract nr 127.

Abstract The foreign body response (FBR) is an immune mediated event that occurs with every material implant. The extent of the fibrosis is dependent on many factors including the biomaterial design, tissue location, and host factors such as age, sex, ancestry, diet. There are known clinical outcomes of implants dependent on age and sex, including increased fibrosis and implant failure in aged and female patients. As the population ages, there is a growing need to understand how aging affects the FBR, and how preclinical models can capture this to guide biomaterial design. Here, we investigated how chronic fibrosis in a murine model of the FBR is altered by two biological factors: age and sex. We investigated changes in fibrosis using a volumetric muscle loss (VML) injury model coupled with polycaprolactone (PCL) or polyethylene (PE) microparticle implants. Fibrosis was quantified through gene expression, microscopic analysis of histologic sections, and the corresponding immune response measured via gene expression and flow cytometry data. We found gene expression differences with immune pathways enriched in female mice, and microscopy revealed collagen birefringence area increased in young male mice. Both the innate and adaptive immune response were altered by age and sex via T cell and macrophage phenotype, and the effects of aging differed between sexes. These results reveal both variables contribute to discrepant outcomes in both fibrosis and the local immune response to synthetic material implants. This demonstrates a clear need to understand and account for the influence of biological factors in biomaterial design.

Mucosal-associated invariant T (MAIT) cells are innate-like T cells capable of MR1-dependent immune surveillance, but how intratumoral bacteria modulate MR1 expression in human lung tumors remains unclear. We studied intratumoral MAIT cells from paired single-cell RNA and TCR sequencing datasets of tumor-infiltrating CD3 T cells isolated from non-small cell lung cancer tumors in patients receiving neoadjuvant PD-1 blockade therapy. MAIT cells were subclustered to identify conventional MAIT-associated TCR clonotypes, which were then used to examine how bacterial exposure impacts cell-surface MR1 expression and downstream MAIT TCR activation. We found that select intratumoral Enterococcus species (spp.) did not directly activate MAIT cells but enhanced MR1-dependent MAIT activation in the presence of exogenous 5-OP-RU by increasing cell-surface MR1 expression on antigen-presenting cells including dendritic cells, B cells, and mononuclear phagocytes. This increase in MR1 cell surface expression is modulated through a posttranscriptional mechanism consistent with altered intracellular processing and trafficking of MR1. These findings reveal a role for tumor-associated bacteria in modulating MR1-dependent innate-like T cell activation and provide a basis for future studies examining whether this process influences response to immune checkpoint blockade.

BACKGROUND:Colorectal cancer pathogenesis involves complex interactions between multiple risk factors including somatic mutations and microbial dysbiosis. A number of individual microbiota members have been implicated in colorectal cancer, including enterotoxigenic strains of Bacteroides fragilis (ETBF). Enterotoxigenic Bacteroides fragilis promotes inflammation in mouse models, which has been mechanistically linked to colon tumorigenesis. We hypothesized that ETBF would promote distinct patterns of colonic damage and inflammation in mice expressing different oncogenic mutations. METHODS:Mice expressing mutations in the Apc tumor suppressor gene or the BRAF or Kras oncogenes were colonized with ETBF to induce acute colitis. Seven days after colonization, tissues and stools were collected to assess for colonization, epithelial damage, and local and systemic immune responses. RESULTS:Despite uniform colonization of ETBF across all genotypes and some common features of colitis across groups, Apc, BRAF, and Kras mutations were associated with distinct patterns of colonic epithelial cell injury and goblet cell loss in response to ETBF. RNA sequencing analysis revealed varied transcriptional profiles based on mouse genotype and colon region. Flow cytometry of intra-epithelial leukocytes revealed differential recruitment of myeloid cells based on oncogenic mutation. In particular, mutant BRAF expression was uniquely associated with more systemic inflammation, resistance to goblet cell loss, an interferon gamma gene signature, and recruitment of a macrophage-like polymorphonuclear myeloid-derived suppressor cell (PMN-MDSC) population in the midproximal colon. CONCLUSIONS:Enterotoxigenic Bacteroides fragilis promotes acute colitis in mice expressing different oncogenic mutations, but with distinct patterns of colonic epithelial cell damage and inflammation dependent on host oncogene context.

The gut microbiota influences systemic immunity and the function of distal tissues, including the brain, liver, skin, lung, and muscle. However, the role of the gut microbiota in the foreign body response and fibrosis is largely unexplored. To investigate this connection, we perturbed the homeostasis of the murine gut microbiota via infection with the pathogenic bacterial species enterotoxigenic Bacteroides fragilis (ETBF) and implanted particulate material (mean particle size <600 μm) of the synthetic polymer polycaprolactone (PCL) into a distal muscle injury. ETBF infection in mice led to increased neutrophil and γδ T cell infiltration into the PCL implant site. ETBF infection alone promoted systemic inflammation, increased levels of neutrophils in lymphoid tissues, and altered skeletal muscle gene expression. At the PCL implant site, we found significant changes in the transcriptome of sorted stromal cells between infected and control mice, including differences related to ECM components such as proteoglycans and glycosaminoglycans. However, we did not observe ETBF-induced differences in fibrosis levels. These results demonstrate the ability of the gut microbiota to mediate long-distance effects such as immune and stromal responses to a distal biomaterial implant.

PURPOSE:Combined inhibition of lymphocyte-activation gene 3 (LAG-3) and PD-1 improves outcomes in patients with melanoma. Increased LAG-3 expression in colorectal cancer correlates with reduced survival. Higher mucin and PD-L1 expression in the mismatch repair-proficient (pMMR) colorectal cancer tumor microenvironment was associated with increased LAG-3 and retrospectively with prolonged progression-free survival upon PD-1 blockade. This led to the hypothesis that LAG-3/PD-1 inhibition would improve clinical outcomes in this pMMR colorectal cancer subset. PATIENTS AND METHODS:NCT03642067 was a phase II study evaluating the combination of relatlimab (LAG-3 inhibitor) and nivolumab (PD-1 inhibitor) in patients with previously treated metastatic pMMR colorectal cancer. Patients were enrolled into one of three cohorts: A, mucin/PD-L1-high; B, mucin/PD-L1-low; or C, mucin/PD-L1 unselected. The primary endpoint for each cohort was the objective response rate. RESULTS:We enrolled 59 evaluable patients; best treatment responses were partial response (3), stable disease (6), and progressive disease (50). Response rates did not differ significantly between cohorts. Subgroup analyses demonstrated that two of five patients with lung-only metastases had a partial response. Comparison of liver and lung metastases identified higher baseline dendritic cell densities in lung lesions. Nivolumab/relatlimab resulted in increased intratumoral cytotoxic T cells. Lower baseline intratumoral regulatory T cells and ADAM10+ cancer cells correlated with clinical response. CONCLUSIONS:This investigation did not reach its primary endpoint for any of the three treatment cohorts but does provide critical insight into the effects of combining nivolumab/relatlimab on the colorectal cancer tumor microenvironment and identifies subgroups that may derive greater benefit from this combination.

Supplementary Table 7. Summary of All Adverse Events Attributed to Nivolumab and/or Relatlimab.

Supplementary Table 2. Antibodies Used for IMC Staining.

The role of mucosal invariant T cells (MAITs), in the lung tumor microenvironment re-mains poorly understood, especially in the setting of immune checkpoint inhibitors. We identified intratumoral MAIT cells from paired single cell RNA and TCR sequencing datasets of tumor infil-trating CD3 T cells isolated from non-small cell lung cancer tumors in patients receiving neoadju-vant PD-1 blockade therapy. MAIT cells were subclustered to identify conventional MAIT-associ-ated TCR clonotypes predicted to recognize intratumoral bacteria, which we then tested for func-tional recognition using a MAIT TCR capture functional assay. Strikingly, although not directly recognized by MAIT cells, previously identified probiotic Enterococcus spp and detected in the intratumoral microbiome of lung cancer patients, selectively synergized with exogenous riboflavin biosynthesis-derived metabolites to induce expression of MR1 by antigen presenting cells, includ-ing dendritic cells, B cells and mononuclear phagocytes. Boosting of MR1 cell surface expression resulted from perturbation of endo-lysosomal vacuolar pathway by Enterococcus and recycling of early endosomal MR1 to the cytoplasmic membrane. Riboflavin auxotrophic Enterococcus spp may therefore exercise their beneficial immunomodulatory functions upon immune checkpoint blockade treatment, at least in part, by promoting intratumoral MR1 expression and innate like T cell activation. Our results indicate that composition of the intratumoral microbiome during im-mune checkpoint inhibitor treatment has the potential to impact the function of human intratumoral MAIT cells. ### Competing Interest Statement The authors have declared no competing interest. Commonwealth foundation Willowcroft foundation

Supplementary Figure 3. CyTOF analysis of circulating PBMCs.

Supplementary Table 1. Antibodies Used for CyTOF Staining.

Supplementary Table 6. Summary of All Grade 3 or Higher Adverse Events.

Background:Ulcerative Colitis (UC) is characterized by chronic, relapsing and remitting inflammation in the colon and rectum. Pathogenic T cell activity is thought to play a major role in this process. T cell effector function is determined by the T cell receptor (TCR) and the antigen it recognizes. Examining the TCR repertoire can provide key insights into the adaptive immune response. Objective:To characterize the longitudinal TCR repertoire of patients with UC across disease activity to determine if recurrent antigen(s) are responsible for active inflammation. Design:Bulk TCR Vβ sequencing was done on colon tissue of 20 patients with UC across multiple time points of disease. Corresponding clinical metadata was also obtained over the same time period for each patient to map their clinical disease course. The top ten most highly abundant clones from each time point were longitudinally tracked and correlated with disease phenotype. Results:Seventy-five percent of patients did not have overlapping abundant TCR clones across multiple time points of disease. The remaining 25% of patients had one to five TCR clones present in high abundance in their tissue during every time point analyzed. Conclusion:These results demonstrate that most patients with UC do not share a similar TCR repertoire over time, indicating that times of inflammation are associated with unique antigen exposures. A smaller group of patients have persistent, private TCR clones with high abundance, 60% of whom had more unremitting, active disease.

C. difficile induces the activation of myeloid cells and IL17-producing lymphoid subsets. Multiparameter flow cytometry analysis of colonic lamina propria innate (myeloid and ILCs) and adaptive immune cell populations of LPL isolated from colons collected at 2 weeks p.i. from GF ApcMin/+ mice gavaged with the 3728T isolates with (n = 5) or without (n = 6) C. difficile strain CIm_3728T. Results from two independent experiments. A, Opt-SNE projection of myeloid and ILC populations. B, Dot plots of myeloid cell type abundance as a percentage of the CD45 population. C, Representative gating of DCs, PMN, and IL17 expression by ILC3 cells. D, Opt-SNE projection of adaptive populations (B and T lymphocytes). E, Dot plots of changes in abundance of lymphoid populations with C. difficile. F, Representative gating of lymphoid subsets. Opt-SNE projections are used for visualization purposes only. Statistical analyses in B and E represent Mann–Whitney t tests based on the gating in C and F, respectively. DC, dendritic cells; EILP, early innate lymphoid progenitors; ILC, innate lymphoid cells; IMC, innate myeloid cells; LPL, lamina propria leukocytes; MAC, macrophages; MO, monocytic cells; NK, natural killer; PMN, polymorphonuclear cells; Treg, regulatory T cells.

Supplementary Table 3. Patient Characteristics by Metastasis Status.

Supplementary Figure 2. Comparison of the tumor microenvironment (TME) of liver metastases across patients.

C. difficile and the 3728T slurry induce rapid tumorigenesis in GF ApcMin/+ mice despite representing only a fraction of the luminal and mucosal microbiota. A and B, Twenty-five GF ApcMin/+ mice were gavaged with the 3728T slurry; five mice were sacrificed per time point. A, Colonic microadenoma and adenoma counts. B, Hematoxylin and eosin (H&E) staining of a representative microadenoma is shown from a mouse at 2 weeks p.i. (top), whereas a large adenoma coated by a dense BF is shown from a mouse at 4 weeks p.i. (bottom). Confocal imaging of Carnoy's-fixed mouse distal colons stained with the EUB338 universal probe (magenta), Cd198 C. difficile probe (yellow), or DAPI counterstain (blue). White arrows highlight individual C. difficile rods. C and D, Microadenomas (C) and FISH staining (D) of 2 weeks p.i. distal colons from GF ApcMin/+ mice gavaged with the 3728T isolates ± C. difficile CIm_3728T. Crosses represent mouse deaths. n = 6 mice per group, from two independent experiments. FISH images represent three separate mice per inoculum. White arrowheads denote areas with occasional mucus-invasive BFs directly interacting with epithelium in mice gavaged with the 3728T isolates without C. difficile (3728T isos – Cd). White arrows (bottom) represent areas with crypt-invasive BFs in the mice gavaged with the 3728T isolates + C. difficile. Statistical significance in A and C was calculated via Kruskal–Wallis followed by Mann–Whitney tests. Mann–Whitney P values are shown. E, Relative abundance (RA) of C. difficile derived from 16S rRNA amplicon sequencing of the original human slurry (3728T) or mouse stool obtained at sacrifice of mice in the time course experiment following gavage with the 3728T slurry. dpi, days post-inoculation. F, FISH of the original patient tumor 3728T. CRC, colorectal cancer. G–M, scRNA-seq and gene set enrichment analysis of 3728T slurry (n = 3)– versus 3979T slurry (n = 3)–gavaged GF ApcMin/+ mice at 2 weeks p.i. Heat maps depict genes enriched in canonical Wnt (G) or ROS (H) signaling pathways, with Wilcoxon rank-sum test P values above each gene. I, Inflammatory signaling pathways enriched in specific cell types from 3728T slurry– versus 3979T slurry–gavaged mice at 2 weeks p.i. KEGG, Kyoto Encyclopedia of Genes and Genomes. J, Inhibition and activation of the most significant upstream regulators in either progenitor or differentiated colonocytes. Expression log ratio is an effect size statistic. K and L, Network analysis of transcription factors from IPA upstream regulators enriched in the progenitor cell gene list revealed activation of Wnt/Myc signaling and proliferation, respectively. Lines connect transcription factors with shared target genes in the gene list. Darker lines indicate a greater number of shared target genes. M, Similar network analysis for differentiated colonocytes showed activation of the innate immune response.