Showing papers by "James R. White published in 2022"

Comparative Analysis of Colon Cancer-Derived Fusobacterium nucleatum Subspecies: Inflammation and Colon Tumorigenesis in Murine Models

Abstract: Colon cancer is a leading cause of cancer morbidity and mortality, and it is hypothesized that dysbiosis in the gut microbiota contributes to colon tumorigenesis. Fusobacterium nucleatum, a member of the oropharyngeal microbiome, is enriched in a subset of human colon tumors. However, it is unclear whether this genetically varied species directly promotes tumor formation, modulates mucosal immune responses, or merely colonizes the tumor microenvironment. ABSTRACT Fusobacteria are commonly associated with human colorectal cancer (CRC), but investigations are hampered by the absence of a stably colonized murine model. Further, Fusobacterium nucleatum subspecies isolated from human CRC have not been investigated. While F. nucleatum subspecies are commonly associated with CRC, their ability to induce tumorigenesis and contributions to human CRC pathogenesis are uncertain. We sought to establish a stably colonized murine model and to understand the inflammatory potential and virulence genes of human CRC F. nucleatum, representing the 4 subspecies, animalis, nucleatum, polymorphum, and vincentii. Five human CRC-derived and two non-CRC derived F. nucleatum strains were tested for colonization, tumorigenesis, and cytokine induction in specific-pathogen-free (SPF) and/or germfree (GF) wild-type and ApcMin/+ mice, as well as in vitro assays and whole-genome sequencing (WGS). SPF wild-type and ApcMin/+ mice did not achieve stable colonization with F. nucleatum, whereas certain subspecies stably colonized some GF mice but without inducing colon tumorigenesis. F. nucleatum subspecies did not form in vivo biofilms or associate with the mucosa in mice. In vivo inflammation was inconsistent across subspecies, whereas F. nucleatum induced greater cytokine responses in a human colorectal cell line, HCT116. While F. nucleatum subspecies displayed genomic variability, no distinct virulence genes associated with human CRC strains were identified that could reliably distinguish these strains from non-CRC clinical isolates. We hypothesize that the lack of F. nucleatum-induced tumorigenesis in our model reflects differences in human and murine biology and/or a synergistic role for F. nucleatum in concert with other bacteria to promote carcinogenesis. IMPORTANCE Colon cancer is a leading cause of cancer morbidity and mortality, and it is hypothesized that dysbiosis in the gut microbiota contributes to colon tumorigenesis. Fusobacterium nucleatum, a member of the oropharyngeal microbiome, is enriched in a subset of human colon tumors. However, it is unclear whether this genetically varied species directly promotes tumor formation, modulates mucosal immune responses, or merely colonizes the tumor microenvironment. Mechanistic studies to address these questions have been stymied by the lack of an animal model that does not rely on daily orogastric gavage. Using multiple murine models, in vitro assays with a human colon cancer cell line, and whole-genome sequencing analysis, we investigated the proinflammatory and tumorigenic potential of several F. nucleatum clinical isolates. The significance of this research is development of a stable colonization model of F. nucleatum that does not require daily oral gavages in which we demonstrate that a diverse library of clinical isolates do not promote tumorigenesis.

Human Colon Cancer–Derived Clostridioides difficile Strains Drive Colonic Tumorigenesis in Mice

Abstract: The patient-derived Clostridioides difficile bacterial strain was found to drive colon tumorigenesis through its toxin TcdB, which induced Wnt signaling and a protumorigenic mucosal immune response.

Peripheral blood immune cell dynamics reflect antitumor immune responses and predict clinical response to immunotherapy

Abstract: Background Despite treatment advancements with immunotherapy, our understanding of response relies on tissue-based, static tumor features such as tumor mutation burden (TMB) and programmed death-ligand 1 (PD-L1) expression. These approaches are limited in capturing the plasticity of tumor–immune system interactions under selective pressure of immune checkpoint blockade and predicting therapeutic response and long-term outcomes. Here, we investigate the relationship between serial assessment of peripheral blood cell counts and tumor burden dynamics in the context of an evolving tumor ecosystem during immune checkpoint blockade. Methods Using machine learning, we integrated dynamics in peripheral blood immune cell subsets, including neutrophil-lymphocyte ratio (NLR), from 239 patients with metastatic non-small cell lung cancer (NSCLC) and predicted clinical outcome with immune checkpoint blockade. We then sought to interpret NLR dynamics in the context of transcriptomic and T cell repertoire trajectories for 26 patients with early stage NSCLC who received neoadjuvant immune checkpoint blockade. We further determined the relationship between NLR dynamics, pathologic response and circulating tumor DNA (ctDNA) clearance. Results Integrated dynamics of peripheral blood cell counts, predominantly NLR dynamics and changes in eosinophil levels, predicted clinical outcome, outperforming both TMB and PD-L1 expression. As early changes in NLR were a key predictor of response, we linked NLR dynamics with serial RNA sequencing deconvolution and T cell receptor sequencing to investigate differential tumor microenvironment reshaping during therapy for patients with reduction in peripheral NLR. Reductions in NLR were associated with induction of interferon-γ responses driving the expression of antigen presentation and proinflammatory gene sets coupled with reshaping of the intratumoral T cell repertoire. In addition, NLR dynamics reflected tumor regression assessed by pathological responses and complemented ctDNA kinetics in predicting long-term outcome. Elevated peripheral eosinophil levels during immune checkpoint blockade were correlated with therapeutic response in both metastatic and early stage cohorts. Conclusions Our findings suggest that early dynamics in peripheral blood immune cell subsets reflect changes in the tumor microenvironment and capture antitumor immune responses, ultimately reflecting clinical outcomes with immune checkpoint blockade.

Improved accuracy and precision with three‐parameter simultaneous myocardial T1 and T2 mapping using multiparametric SASHA

Abstract: To develop and validate a three‐parameter model for improved precision multiparametric SAturation‐recovery single‐SHot Acquisition (mSASHA) cardiac T1 and T2 mapping with high accuracy in a single breath‐hold.

Hemodynamic Assessment in Bicuspid Aortic Valve Disease and Aortic Dilation: New Insights From Voxel-By-Voxel Analysis of Reverse Flow, Stasis, and Energetics

Abstract: Objectives: Clinical management decisions surrounding ascending aorta (AAo) dilation in bicuspid aortic valve (BAV) disease benefit from personalized predictive tools. 4D-flow MRI may provide patient-specific markers reflective of BAV-associated aortopathy. This study aims to explore novel 4D-flow MRI parametric voxel-by-voxel forward flow, reverse flow, kinetic energy and stasis in BAV disease. We hypothesize that novel parametric voxel-by-voxel markers will be associated with aortic dilation and referral for surgery and can enhance our understanding of BAV hemodynamics beyond standard metrics. Methods: A total of 96 subjects (73 BAV patients, 23 healthy controls) underwent MRI scan. Healthy controls had no known cardiovascular disease. Patients were clinically referred for AAo dilation assessment. Indexed diameters were obtained by dividing the aortic diameter by the patient’s body surface area. Patients were followed for the occurrence of aortic surgery. 4D-flow analysis was performed by a single observer in five regions: left ventricular outflow tract (LVOT), AAo, arch, proximal descending aorta (PDAo), and distal descending aorta (DDAo). In each region peak velocity, kinetic energy (KE), forward flow (FF), reverse flow (RF), and stasis were measured on a voxel-by-voxel basis. T-tests (or non-parametric equivalent) compared flow parameters between cohorts. Univariate and multivariate analyses explored associations between diameter and parametric voxel-by-voxel parameters. Results: Compared to controls, BAV patients showed reduced stasis (p < 0.01) and increased RF and FF (p < 0.01) throughout the aorta, and KE remained similar. In the AAo, indexed diameter correlated with age (R = 0.326, p = 0.01), FF (R = −0.648, p < 0.001), RF (R = −0.441, p < 0.001), and stasis (R = −0.288, p < 0.05). In multivariate analysis, FF showed a significant inverse association with AAo indexed diameter, independent of age. During a median 179 ± 180 days of follow-up, 23 patients (32%) required aortic surgery. Compared to patients not requiring surgery, they showed increased KE and peak velocity in the proximal aorta (p < 0.01), accompanied by increased RF and reduced stasis throughout the entire aorta (p < 0.01). Conclusion: Novel voxel-by-voxel reverse flow and stasis were altered in BAV patients and are associated with aortic dilation and surgical treatment.

Incorporation of Data From Multiple Hypervariable Regions when Analyzing Bacterial 16S rRNA Gene Sequencing Data

Abstract: Short read 16 S rRNA amplicon sequencing is a common technique used in microbiome research. However, inaccuracies in estimated bacterial community composition can occur due to amplification bias of the targeted hypervariable region. A potential solution is to sequence and assess multiple hypervariable regions in tandem, yet there is currently no consensus as to the appropriate method for analyzing this data. Additionally, there are many sequence analysis resources for data produced from the Illumina platform, but fewer open-source options available for data from the Ion Torrent platform. Herein, we present an analysis pipeline using open-source analysis platforms that integrates data from multiple hypervariable regions and is compatible with data produced from the Ion Torrent platform. We used the ThermoFisher Ion 16 S Metagenomics Kit and a mock community of twenty bacterial strains to assess taxonomic classification of six amplicons from separate hypervariable regions (V2, V3, V4, V6-7, V8, V9) using our analysis pipeline. We report that different amplicons have different specificities for taxonomic classification, which also has implications for global level analyses such as alpha and beta diversity. Finally, we utilize a generalized linear modeling approach to statistically integrate the results from multiple hypervariable regions and apply this methodology to data from a representative clinical cohort. We conclude that examining sequencing results across multiple hypervariable regions provides more taxonomic information than sequencing across a single region. The data across multiple hypervariable regions can be combined using generalized linear models to enhance the statistical evaluation of overall differences in community structure and relatedness among sample groups.

Automated next-generation profiling of genomic alterations in human cancers

Abstract: The lack of validated, distributed comprehensive genomic profiling assays for patients with cancer inhibits access to precision oncology treatment. To address this, we describe elio tissue complete, which has been FDA-cleared for examination of 505 cancer-related genes. Independent analyses of clinically and biologically relevant sequence changes across 170 clinical tumor samples using MSK-IMPACT, FoundationOne, and PCR-based methods reveals a positive percent agreement of >97%. We observe high concordance with whole-exome sequencing for evaluation of tumor mutational burden for 307 solid tumors (Pearson r = 0.95) and comparison of the elio tissue complete microsatellite instability detection approach with an independent PCR assay for 223 samples displays a positive percent agreement of 99%. Finally, evaluation of amplifications and translocations against DNA- and RNA-based approaches exhibits >98% negative percent agreement and positive percent agreement of 86% and 82%, respectively. These methods provide an approach for pan-solid tumor comprehensive genomic profiling with high analytical performance.

Murine fecal microbiota transfer models selectively colonize human microbes and reveal transcriptional programs associated with response to neoadjuvant checkpoint inhibitors

Abstract: Human gut microbial species found to associate with clinical responses to immune checkpoint inhibitors (ICIs) are often tested in mice using fecal microbiota transfer (FMT), wherein tumor responses in recipient mice may recapitulate human responses to ICI treatment. However, many FMT studies have reported only limited methodological description, details of murine cohorts, and statistical methods. To investigate the reproducibility and robustness of gut microbial species that impact ICI responses, we performed human to germ-free mouse FMT using fecal samples from patients with non-small cell lung cancer who had a pathological response or nonresponse after neoadjuvant ICI treatment. R-FMT mice yielded greater anti-tumor responses in combination with anti-PD-L1 treatment compared to NR-FMT, although the magnitude varied depending on mouse cell line, sex, and individual experiment. Detailed investigation of post-FMT mouse microbiota using 16S rRNA amplicon sequencing, with models to classify and correct for biological variables, revealed a shared presence of the most highly abundant taxa between the human inocula and mice, though low abundance human taxa colonized mice more variably after FMT. Multiple Clostridium species also correlated with tumor outcome in individual anti-PD-L1-treated R-FMT mice. RNAseq analysis revealed differential expression of T and NK cell-related pathways in responding tumors, irrespective of FMT source, with enrichment of these cell types confirmed by immunohistochemistry. This study identifies several human gut microbial species that may play a role in clinical responses to ICIs and suggests attention to biological variables is needed to improve reproducibility and limit variability across experimental murine cohorts.

Mid-wall striae fibrosis predicts heart failure admission, composite heart failure events, and life-threatening arrhythmias in dilated cardiomyopathy

Abstract: Heart failure (HF) admission is a dominant contributor to morbidity and healthcare costs in dilated cardiomyopathy (DCM). Mid-wall striae (MWS) fibrosis by late gadolinium enhancement (LGE) imaging has been associated with elevated arrhythmia risk. However, its capacity to predict HF-specific outcomes is poorly defined. We investigated its role to predict HF admission and relevant secondary outcomes in a large cohort of DCM patients. 719 patients referred for LGE MRI assessment of DCM were enrolled and followed for clinical events. Standardized image analyses and interpretations were conducted inclusive of coding the presence and patterns of fibrosis observed by LGE imaging. The primary clinical outcome was hospital admission for decompensated HF. Secondary heart failure and arrhythmic composite endpoints were also studied. Median age was 57 (IQR 47-65) years and median LVEF 40% (IQR 29-47%). Any fibrosis was observed in 228 patients (32%) with MWS fibrosis pattern present in 178 (25%). At a median follow up of 1044 days, 104 (15%) patients experienced the primary outcome, and 127 (18%) the secondary outcome. MWS was associated with a 2.14-fold risk of the primary outcome, 2.15-fold risk of the secondary HF outcome, and 2.23-fold risk of the secondary arrhythmic outcome. Multivariable analysis adjusting for all relevant covariates, inclusive of LVEF, showed patients with MWS fibrosis to experience a 1.65-fold increased risk (95% CI 1.11-2.47) of HF admission and 1-year event rate of 12% versus 7% without this phenotypic marker. Similar findings were observed for the secondary outcomes. Patients with LVEF > 35% plus MWS fibrosis experienced similar event rates to those with LVEF ≤ 35%. MWS fibrosis is a powerful and independent predictor of clinical outcomes in patients with DCM, identifying patients with LVEF > 35% who experience similar event rates to those with LVEF below this conventionally employed high-risk phenotype threshold.

Expanding the genomic encyclopedia of Actinobacteria with 824 isolate reference genomes

Abstract: The phylum Actinobacteria includes important human pathogens like Mycobacterium tuberculosis and Corynebacterium diphtheriae and renowned producers of secondary metabolites of commercial interest, yet only a small part of its diversity is represented by sequenced genomes. Here, we present 824 actinobacterial isolate genomes in the context of a phylum-wide analysis of 6,700 genomes including public isolates and metagenome-assembled genomes (MAGs). We estimate that only 30%-50% of projected actinobacterial phylogenetic diversity possesses genomic representation via isolates and MAGs. A comparison of gene functions reveals novel determinants of host-microbe interaction as well as environment-specific adaptations such as potential antimicrobial peptides. We identify plasmids and prophages across isolates and uncover extensive prophage diversity structured mainly by host taxonomy. Analysis of >80,000 biosynthetic gene clusters reveals that horizontal gene transfer and gene loss shape secondary metabolite repertoire across taxa. Our observations illustrate the essential role of and need for high-quality isolate genome sequences.

3-Dimensional Strain Analysis of Hypertrophic Cardiomyopathy: Insights From the NHLBI International HCM Registry.

Abstract: Abnormal global longitudinal strain (GLS) has been independently associated with adverse cardiac outcomes in both obstructive and nonobstructive hypertrophic cardiomyopathy.The goal of this study was to understand predictors of abnormal GLS from baseline data from the National Heart, Lung, and Blood Institute (NHLBI) Hypertrophic Cardiomyopathy Registry (HCMR).The study evaluated comprehensive 3-dimensional left ventricular myocardial strain from cine cardiac magnetic resonance in 2,311 patients from HCMR using in-house validated feature-tracking software. These data were correlated with other imaging markers, serum biomarkers, and demographic variables.Abnormal median GLS (> -11.0%) was associated with higher left ventricular (LV) mass index (93.8 ± 29.2 g/m2 vs 75.1 ± 19.7 g/m2; P < 0.0001) and maximal wall thickness (21.7 ± 5.2 mm vs 19.3 ± 4.1 mm; P < 0.0001), lower left (62% ± 9% vs 66% ± 7%; P < 0.0001) and right (68% ± 11% vs 69% ± 10%; P < 0.01) ventricular ejection fractions, lower left atrial emptying functions (P < 0.0001 for all), and higher presence and myocardial extent of late gadolinium enhancement (6 SD and visual quantification; P < 0.0001 for both). Elastic net regression showed that adjusted predictors of GLS included female sex, Black race, history of syncope, presence of systolic anterior motion of the mitral valve, reverse curvature and apical morphologies, LV ejection fraction, LV mass index, and both presence/extent of late gadolinium enhancement and baseline N-terminal pro-B-type natriuretic peptide and troponin levels.Abnormal strain in hypertrophic cardiomyopathy is associated with other imaging and serum biomarkers of increased risk. Further follow-up of the HCMR cohort is needed to understand the independent relationship between LV strain and adverse cardiac outcomes in hypertrophic cardiomyopathy.

Whole-Heart Assessment of Turbulent Kinetic Energy in the Repaired Tetralogy of Fallot

Abstract: Approximately 10% of congenital heart diseases (CHDs) include Tetralogy of Fallot (TOF). Fortunately, due to advanced surgical techniques, most patients survive until adulthood. However, these patients require frequent monitoring for postoperative complications leading to heart hemodynamic alterations. Turbulent kinetic energy (TKE), as derived from 4D-flow magnetic resonance imaging (4D-flow MRI), has been used to characterize abnormal heart hemodynamics in CHD. Hence, this study aimed to assess the difference in TKE between patients with repaired TOF (rTOF) and healthy volunteers. A total of 35 subjects, 17 rTOF patients and 18 controls, underwent standard-of-care cardiac MRI and research 4D-flow MRI using a clinical 3T scanner. Heart chambers and great vessels were segmented using 3D angiograms derived from 4D-flow MRI. The TKE was quantified within segmented volumes. TKE was compared to standard cardiac MRI metrics. Controls demonstrated higher TKE in the left atria and left ventricle. However, patients demonstrated higher TKE in the right atria, right ventricle (p < 0.05), and pulmonary artery. Lastly, no correlation was observed between TKE and standard clinical measurements. TKE can be a key indicator of the abnormal hemodynamics present in patients with rTOF and can assist future interventions and help monitor long-term outcomes.

Genome-wide mutational signatures in low-coverage whole genome sequencing of cell-free DNA

Abstract: Mutational signatures accumulate in somatic cells as an admixture of endogenous and exogenous processes that occur during an individual's lifetime. Since dividing cells release cell-free DNA (cfDNA) fragments into the circulation, we hypothesize that plasma cfDNA might reflect mutational signatures. Point mutations in plasma whole genome sequencing (WGS) are challenging to identify through conventional mutation calling due to low sequencing coverage and low mutant allele fractions. In this proof of concept study of plasma WGS at 0.3-1.5x coverage from 215 patients and 227 healthy individuals, we show that both pathological and physiological mutational signatures may be identified in plasma. By applying machine learning to mutation profiles, patients with stage I-IV cancer can be distinguished from healthy individuals with an Area Under the Curve of 0.96. Interrogating mutational processes in plasma may enable earlier cancer detection, and might enable the assessment of cancer risk and etiology.

Intra-cardiac pressure drop and flow distribution of bicuspid aortic valve disease in preserved ejection fraction

Abstract: Background Bicuspid aortic valve (BAV) is more than a congenital defect since it is accompanied by several secondary complications that intensify induced impairments. Hence, BAV patients need lifelong evaluations to prevent severe clinical sequelae. We applied 4D-flow magnetic resonance imaging (MRI) for in detail visualization and quantification of in vivo blood flow to verify the reliability of the left ventricular (LV) flow components and pressure drops in the silent BAV subjects with mild regurgitation and preserved ejection fraction (pEF). Materials and methods A total of 51 BAV patients with mild regurgitation and 24 healthy controls were recruited to undergo routine cardiac MRI followed by 4D-flow MRI using 3T MRI scanners. A dedicated 4D-flow module was utilized to pre-process and then analyze the LV flow components (direct flow, retained inflow, delayed ejection, and residual volume) and left-sided [left atrium (LA) and LV] local pressure drop. To elucidate significant diastolic dysfunction in our population, transmitral early and late diastolic 4D flow peak velocity (E-wave and A-wave, respectively), as well as E/A ratio variable, were acquired. Results The significant means differences of each LV flow component (global measurement) were not observed between the two groups (p > 0.05). In terms of pressure analysis (local measurement), maximum and mean as well as pressure at E-wave and A-wave timepoints at the mitral valve (MV) plane were significantly different between BAV and control groups (p: 0.005, p: 0.02, and p: 0.04 and p: <0.001; respectively). Furthermore, maximum pressure and pressure difference at the A-wave timepoint at left ventricle mid and left ventricle apex planes were significant. Although we could not find any correlation between LV diastolic function and flow components, Low but statistically significant correlations were observed with local pressure at LA mid, MV and LV apex planes at E-wave timepoint (R: −0.324, p: 0.005, R: −0.327, p: 0.004, and R: −0.306, p: 0.008, respectively). Conclusion In BAV patients with pEF, flow components analysis is not sensitive to differentiate BAV patients with mild regurgitation and healthy control because flow components and EF are global parameters. Inversely, pressure (local measurement) can be a more reliable biomarker to reveal the early stage of diastolic dysfunction.

Automated myocardial segmentation of extra-cellular volume mapping cardiac magnetic resonance images using fully convolutional neural networks

Abstract: Extra-cellular volume (ECV) mapping cardiac magnetic resonance (CMR) imaging allows for the characterization of expanded myocardial extracellular space, a common feature of myocardial fibrosis (MF). Quantification of MF is feasible using ECV mapping techniques; however, prior manual delineation of the endocardial and epicardial borders is required. In this study, we propose a method for automated myocardial delineation of ECV maps using convolutional neural networks (CNNs). We compare two methods based on the standard U-Net and the U-Net++ architectures using a five-fold cross validation on basal, mid, and apical short-axis ECV maps of the left ventricle (LV) in 73 patients with ischemic (n=38) or dilated (n=35) cardiomyopathies. The standard U-Net and U-Net++ -based architectures yielded DSC metrics of 87.61% and 87.89%, respectively, against manual contours derived by an expert. Precision and recall were reported >85% and relative error <12% for both CNNs. The U-Net++ architecture outperformed the standard U-Net on the order of 1-2% for all metrics. An inter-operator variability analysis was performed on a subset of myocardial contours derived by three operators. The inter-operator analysis demonstrated significant differences in the distribution of myocardial ECV values among three operators as per the Kruskal-Wallis H-test (average pair-wise P-value = 0.040), but operator differences failed to show significance against U-Net++ or standard U-Net (average pair-wise P-value 0.055 and 0.060, respectively). Correlation of global ECV improved for operators against U-Net++ (𝜌=0.88) and against standard U-Net (𝜌=0.877) compared to correlation of global ECV values between all operators (ρ=0.828).

Natural Language Processing Approaches for Retrieval of Clinically Relevant Genomic Information in Cancer

Abstract: The accelerating impact of genomic data in clinical decision-making has generated a paradigm shift from treatment based on the anatomic origin of the tumor to the incorporation of key genomic features to guide therapy. Assessing the clinical validity and utility of the genomic background of a patient’s cancer represents one of the emerging challenges in oncology practice, demanding the development of automated platforms for extracting clinically relevant genomic information from medical texts. We developed PubMiner, a natural language processing tool to extract and interpret cancer type, therapy, and genomic information from biomedical abstracts. Our initial focus has been the retrieval of gene names, variants, and negations, where PubMiner performed highly in terms of total recall (91.7%) with a precision of 79.7%. Our next steps include developing a web-based interface to promote personalized treatment based on each tumor’s unique genomic fingerprints.

Abstract 1973: Patients with operable esophageal cancer and improved responses to combined chemoradiotherapy and immunotherapy display distinct microbiome profiles enriched in multiple Bacteroides species

Abstract: Background: Preclinical and clinical data indicate that neoadjuvant chemoradiotherapy (CRT) may prime an anti-tumor immunological response in esophageal cancer driven by intratumoral CD8+ T cells and PD-L1 expression. LAG-3 is also highly expressed in esophagogastric cancers. The microbiome, a novel and potentially modifiable, biomarker of IO response, has not yet been examined in the neoadjuvant setting in esophageal cancer and is the goal of our study. Methods: Fecal samples were collected from patients with stage II/III esophageal or gastroesophageal junction carcinoma eligible for curative resection treated with the standard of care regimen of carboplatin paclitaxel (50mg/m2), radiation 50.4 Gy in 28 fractions and an Ivor-Lewis esophagectomy 6-10 weeks after last CRT and immunotherapy (IO) dose. Patients on arm A (n=11) received 2 cycles of induction with nivolumab plus 3 additional cycles on week 1, 3 and 5 of CRT. Patients on arm B (n=8) received nivolumab plus relatlimab on the same schedule (Clinical trial: NCT03044613). We examined longitudinal fecal samples from n=19 patients across both arms (n=90 samples) using 16S rRNA amplicon sequencing. Patients were classified based on pathological response: complete response (CR) and grades 1, 2, and 3 (G1, G2, G3) with increasing residual tumor visible in the resected specimen. Sequencing data was trimmed and filtered for contaminants, followed by high-resolution taxonomic assignment and normalization of reads across all samples. Analysis was performed using multiple metrics for alpha diversity and beta-diversity, with principal coordinates analysis/PERMANOVA, and pathway analysis using Phylogenetic Investigation of Communities by Reconstruction of Unobserved States (PICRUSt). Results: Patients with improved response in the neoadjuvant setting (CR/G1 vs G2/G3) grouped in distinct clusters using Bray-Curtis (p < 0.001). Patients with CR had higher alpha diversity, using both measures of richness and evenness, compared to patients with a G3 responses (p < 0.03). Specifically, family Bacteroidaceae and genus Bacteroides were enriched in patients with CR vs G3 (p < 0.02). At the species level, B. finegoldii, B. ovatus, and B. uniformis were enriched in patients with CR vs G3 (p < 0.02). In contrast, genus Klebsiella and Clostridium termitidis were enriched in patients with a poor response, G3 (p <0.001, both). Pathway analysis found two metabolic pathways enriched in patients with CR: secondary bile acid biosynthesis (p=0.005) and lysine biosynthesis (p=0.02). Conclusions: Patients with operable esophageal cancer and improved responses to combined CRT and IO had distinct microbiome profiles enriched in multiple Bacteroides species. Further analyses and validation efforts are underway to confirm metabolomic pathways. Citation Format: Fyza Y. Shaikh, James R. White, Ronan J. Kelly, Ali H. Zaidi, Jenna V. Canzoniero, Josephine L. Feliciano, Russell K. Hales, K Ranh Voong, Richard J. Battafarano, Blair A. Jobe, Stephen C. Yang, Stephen Broderick, Jinny Ha, Kellie N. Smith, Elizabeth Thompson, Eun J. Shin, Ali I. Amjad, Patrizia Guerrieri, Benny Weksler, Chen Hu, Valsamo Anagnostou, Vincent K. Lam, Cynthia L. Sears. Patients with operable esophageal cancer and improved responses to combined chemoradiotherapy and immunotherapy display distinct microbiome profiles enriched in multiple Bacteroides species [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2022; 2022 Apr 8-13. Philadelphia (PA): AACR; Cancer Res 2022;82(12_Suppl):Abstract nr 1973.

Machine learning prediction of atrial fibrillation in cardiovascular patients using cardiac magnetic resonance and electronic health information

Abstract: Background Atrial fibrillation (AF) is a commonly encountered cardiac arrhythmia associated with morbidity and substantial healthcare costs. While patients with cardiovascular disease experience the greatest risk of new-onset AF, no risk model has been developed to predict AF occurrence in this population. We hypothesized that a patient-specific model could be delivered using cardiovascular magnetic resonance (CMR) disease phenotyping, contextual patient health information, and machine learning. Methods Nine thousand four hundred forty-eight patients referred for CMR imaging were enrolled and followed over a 5-year period. Seven thousand, six hundred thirty-nine had no prior history of AF and were eligible to train and validate machine learning algorithms. Random survival forests (RSFs) were used to predict new-onset AF and compared to Cox proportional-hazard (CPH) models. The best performing features were identified from 115 variables sourced from three data domains: (i) CMR-based disease phenotype, (ii) patient health questionnaire, and (iii) electronic health records. We evaluated discriminative performance of optimized models using C-index and time-dependent AUC (tAUC). Results A RSF-based model of 20 variables (CIROC-AF-20) delivered an overall C-index of 0.78 for the prediction of new-onset AF with respective tAUCs of 0.80, 0.79, and 0.78 at 1-, 2- and 3-years. This outperformed a novel CPH-based model and historic AF risk scores. At 1-year of follow-up, validation cohort patients classified as high-risk of future AF by CIROC-AF-20 went on to experience a 17.3% incidence of new-onset AF, being 24.7-fold higher risk than low risk patients. Conclusions Using phenotypic data available at time of CMR imaging we developed and validated the first described risk model for the prediction of new-onset AF in patients with cardiovascular disease. Complementary value was provided by variables from patient-reported measures of health and the electronic health record, illustrating the value of multi-domain phenotypic data for the prediction of AF.

Abstract 10325: Feasibility and Predictive Utility of Three-Dimensional Myocardial Deformation Analysis From Multi-Phase Computed Tomography Angiography in Transcatheter Aortic Valve Replacement

Abstract: Introduction : Multi-phase computed tomography angiography (CTA) for the pre-procedural planning of TAVR presents a unique opportunity to assess 3D myocardial biomechanics. Using a novel approach, we assessed the feasibility and predictive utility of 3D myocardial deformation analysis (3D-MDA) to deliver principal strain (PS) based markers of left ventricular (LV) health for the prediction of time to heart failure hospitalization or death following TAVR. Methods: 205 patients undergoing a pre-TAVR CTA and followed clinically for >6 months were identified. Whole-heart segmentation with 3D-mesh modelling was followed by 3D-MDA of the LV. 3D global LV minimum PS (minPS) was calculated for endocardial, epicardial and transmural layers. Cox regression models were performed to evaluate associations between 3D minPS and the composite outcome of all-cause mortality or heart failure hospitalization. Results: Of 205 patients, 196 (96%) had analyzable CTA data for 3D-MDA. Median (IQR) age was 85 (79.5-88) years (55% male) with median STS-PROM score 3.10 (2.10-4.55)% and median echocardiography LVEF 60 (55.9-65.0)%. Over 25 (11-36) months 55 patients (28%) experienced all-cause death or HF hospitalization. Patients with lower 3D minPS, below -23.7%, experienced a 3-fold increased risk of the primary outcome (p<0.001). Following adjustment for baseline characteristics, inclusive of STS and LVEF, 3D minPS remained independently associated with the primary outcome: endocardial 3D minPS providing highest prognostic value [HR (95% CI) of 1.09 per 1% change (1.04-1.15), p<0.001]. Conclusions: 3D-MDA of pre-TAVR multi-phase CTA is feasible and delivers principal-strain based markers strongly predictive of future clinical outcomes. The potential of this approach to optimize patient selection and post-procedural management requires future evaluation in a multi-centre setting.

Machine Learning Patient-Specific Prediction of Heart Failure Hospitalization Using Cardiac MRI-Based Phenotype and Electronic Health Information

Abstract: Background Heart failure (HF) hospitalization is a dominant contributor of morbidity and healthcare expenditures in patients with systolic HF. Cardiovascular magnetic resonance (CMR) imaging is increasingly employed for the evaluation of HF given capacity to provide highly reproducible phenotypic markers of disease. The combined value of CMR phenotypic markers and patient health information to deliver predictions of future HF events has not been explored. We sought to develop and validate a novel risk model for the patient-specific prediction of time to HF hospitalization using routinely reported CMR variables, patient-reported health status, and electronic health information. Methods Standardized data capture was performed for 1,775 consecutive patients with chronic systolic HF referred for CMR imaging. Patient demographics, symptoms, Health-related Quality of Life, pharmacy, and routinely reported CMR features were provided to both machine learning (ML) and competing risk Fine-Gray-based models (FGM) for the prediction of time to HF hospitalization. Results The mean age was 59 years with a mean LVEF of 36 ± 11%. The population was evenly distributed between ischemic (52%) and idiopathic non-ischemic cardiomyopathy (48%). Over a median follow-up of 2.79 years (IQR: 1.59–4.04) 333 patients (19%) experienced HF related hospitalization. Both ML and competing risk FGM based models achieved robust performance for the prediction of time to HF hospitalization. Respective 90-day, 1 and 2-year AUC values were 0.87, 0.83, and 0.80 for the ML model, and 0.89, 0.84, and 0.80 for the competing risk FGM-based model in a holdout validation cohort. Patients classified as high-risk by the ML model experienced a 34-fold higher occurrence of HF hospitalization at 90 days vs. the low-risk group. Conclusion In this study we demonstrated capacity for routinely reported CMR phenotypic markers and patient health information to be combined for the delivery of patient-specific predictions of time to HF hospitalization. This work supports an evolving migration toward multi-domain data collection for the delivery of personalized risk prediction at time of diagnostic imaging.