Cancer Biomarker Discovery Research Articles

Bifunctional Nanomaterial Enabled High-specific Isolation of Urinary Exosomes for Cervical Cancer Metabolomics Analysis and Biomarker Discovery

Cervical cancer (CC) remains a critical public health issue, highlighting the importance of early detection. However, current methods such as cytological and HPV testing face challenges of invasiveness and low patient compliance. Exosomes, emerging as crucial in cancer diagnosis, offer promise due to their noninvasive, highly specificity, and abundant biomarkers. However, isolating exosomes efficiently remains challenging. In this study, we designed and synthesized a bifunctional affinity nanomaterial Fe3O4 @CD63-CLIKKPF, based on the synergistic interaction between its modified aptamer CD63 and peptide CLIKKPF, and CD63 protein and PS of exosomes which can achieve high specificity and high yield separation of urinary exosomes. Notably, the co-modified aptamer CD63 and peptide CLIKKPF not only enable efficient exosome isolation by leveraging dual-affinity mechanisms through a synergistic "AND" logic analysis, but also could be achieved on the Fe3O4 in one-step reaction at room temperature via Fe-S bonding. Combined with LC-MS/MS, we conducted exosome metabolomics analysis in healthy individuals and CC patients across various stages, and machine learning models demonstrated accurate classification (accuracy > 0.822) and prediction capabilities for CC. Furthermore, six key metabolites indicative of CC progression were identified and validated in additional patient samples, highlighting their potential as biomarkers. Overall, this study establishes a novel method for exosome metabolomics in CC, offering insights for non-invasive early diagnosis and progression prediction on a large scale.

Abstract B061: Assessing saliva for cancer biomarker discovery: A liquid biopsy approach

Abstract Introduction: Liquid biopsy has been gaining traction as a replacement for invasive tumor biopsies in oncology. Several blood-based in vitro diagnostics have been approved by the FDA and other regulatory agencies, and efforts are being made to validate urine, saliva, and buccal swabs as alternative matrices. The primary aim of this study was to investigate the relative mRNA expression in paired blood and saliva samples from cancer patients and healthy individuals. Methods: We collected matched saliva and blood samples from 35 subjects, comprising 26 patients with prostate, breast, colon, lung, melanoma, and pancreatic cancers, and 9 healthy individuals. Samples were collected using Wren's proprietary stabilization buffer kit, and total RNA extracted by TRIzol-based Qiagen reagents. RNA sequencing (RNA-Seq) was performed on a Novaseq 6000 platform, and data was analyzed using high-performance computing (HPC) with PartekFlow software, employing the BWA aligner (hg38). We focused on comparing gene expression in cancers (as a group) compared to controls. Results: The total number of genes detected in saliva was 3,985, compared to 14,459 genes detected in blood samples. A substantial overlap was observed, with more than 92.8% of saliva-detected genes also found in blood. We also examined housekeeping gene detection in both matrices. In this set of 1,003 housekeeping genes, 30% (n=305) were expressed in both saliva and blood. When we narrowed the scope to only include the 1,154 cancer-related genes as listed by the OncoKB database, 88.3% were detected in saliva and/or blood, with 37.1% of these genes (n=429) were expressed in both saliva and blood. Lastly, we examined a set of 241 cancer-associated biomarkers utilized in RT-PCR-based diagnostic blood assays clinically validated by our laboratory. This subset analysis revealed that 23.2% of these biomarkers were expressed in both saliva and blood. Conclusion: Our data indicate that the majority of mRNA detected in saliva was also detected in blood, with differential expression observed in cancer patients compared to normal controls. Given the relatively high limit of detection (low sensitivity) using RNA-Seq technology compared to PCR and the relatively small number of subjects, the data may suggest that saliva could function as a surrogate biomarker matrix for blood. Citation Format: Srinivas V Koduru, Mark Kidd, Abdel B Halim. Assessing saliva for cancer biomarker discovery: A liquid biopsy approach [abstract]. In: Proceedings of the AACR Special Conference: Liquid Biopsy: From Discovery to Clinical Implementation; 2024 Nov 13-16; San Diego, CA. Philadelphia (PA): AACR; Clin Cancer Res 2024;30(21_Suppl):Abstract nr B061.

Causality Between Immune Cells, Metabolites and Breast Cancer: Mendelian Randomization and Mediation Analysis.

Previous studies have shown that immune cells and metabolites are associated with the development of breast cancer, but the causal relationship is unclear. We use Mendelian randomization (MR) to explore potential connections between them. Based on two sample MR studies, we evaluated the causal relationship between 731 immune cell traits, 1400 metabolites and breast cancer. In addition, we evaluated the mediating role of metabolites between immune cells and breast cancer using two-step MR Studies. Pooled GWAS data on 731 immune cell traits (n = 3757), 1400 metabolites (n = 8299) and breast cancer (ncase = 122,977, ncontrol = 105,974) were obtained from publicly available authoritative databases. We mainly used inverse variance weighted (IVW) method combined with Bayesian weighted MR (BWMR), MR-Egger, weighted median, simple mode, and weighted mode methods. The results of MR Studies showed that 3 immune cells and 15 metabolites were associated with an increased risk of breast cancer. 8 immune cells and 11 metabolites are associated with a reduced risk of breast cancer. Mediating MR Analysis showed that 6 metabolites, Tricosanoyl sphingomyelin (d18:1/23:0) levels(Mediated proportion:17.5%), N-palmitoyl-heptadecasphingosine (d17:1/16:0) levels(8.74%), Inosine 5'-monophosphate (IMP) to phosphate ratio(11.3%), Glutamine to asparagine ratio(5.43%), Oleoyl-linoleoyl-glycerol (18:1/18:2) [2] levels(8.48%), and Sphinganine-1-phosphate levels(8.68%), were found to mediate the relationship between immune cells and breast cancer. Our study reveals a potential causal relationship among multiple immune cells traits, metabolites, and breast cancer. It provides valuable clues and potential therapeutic targets for breast cancer biomarker discovery and breast cancer treatment.

Metabolomics of Papanicolaou Tests for the Discovery of Ovarian Cancer Biomarkers.

Background: Ovarian cancer (OC) remains one of the most lethal cancers among women due to most cases going undiagnosed until later stages. The early detection and treatment of this malignancy provides the best prognosis, but the lack of an accurate and sensitive screening tool combined with ambiguous symptoms hinders these diagnoses. In contrast, screening for cervical cancer via Papanicolaou (Pap) tests is a widespread practice that greatly reduces the cancer's mortality rates. Interestingly, previous studies show evidence of OC cells in Pap tests, suggesting that proteins, and potentially lipids, shed from ovarian tumors end up in the cervix. The goal of this study is to evaluate the practicality of using Pap tests as biospecimens for OC-screening-related metabolomics. Methods: To evaluate the effectiveness of using residual Pap test samples as biospecimens for potential metabolomics work, 29 Pap test samples, collected from women over the age of 50 with normal cytology and no visible blood contamination, were first obtained from the University of Minnesota, with IRB approval. These samples were centrifuged to recover the cell pellets from the supernatants. The cell pellets underwent a biphasic extraction, followed by an RP-LC-MS analysis, while the supernatants underwent two separate extractions and analyses, including RP-LC-MS and HILIC-LC-MS. Non-targeted features were detected in the range of 220-1000 m/z to determine the sensitivity and scope of the various extraction and analytical workflows, as well as evaluating residual Pap test samples as viable metabolomics biospecimens. Results: The biphasic extraction and subsequent RP-LC-MS analysis of the isolated cell pellets from all 29 samples yielded informative, exploratory data, highlighting the potential of using residual Pap test samples as biospecimens for metabolomics, specifically lipidomics, studies. Each sample was analyzed in both the positive and negative ion mode, yielding the detection of 7318 in the positive ion mode and 3733 in the negative ion mode. Using multiple reference libraries, 22.85% and 36.19% of these features were annotated in the positive and negative ion mode, respectively. Among these detected features, 453 unique lipids, representative of 20 different lipid subclasses, were annotated in all 29 samples. Of the various lipid subclasses represented from the detected lipids, ceramides, triacylglycerols, hexosylceramides, and phosphatidylcholines contributed to over half (53.3%) of the detected lipids at 16.2%, 13.0%, 12.8%, and 11.3%, respectively. Conclusions: The detection of these 453 common lipids across all patients establishes a relative lipidome baseline for women over the age of 50 with normal cervical cytology. This exploratory study is the first investigation to utilize residual Pap test samples as biospecimens in a metabolomics/lipidomics workflow.

Discovery of RNA Biomarkers for Prostate Cancer Using Cross-Platform Transcriptomics.

Microarray and Single-Molecule Molecular Inversion Probe (smMIP)-based targeted RNA sequencing are two RNA profiling platforms for identifying disease-associated biomarkers. The microarray uses a GeneChip array with oligonucleotide probes to measure expression levels across thousands of genes, while smMIPs capture and quantify RNA transcripts and transcript variants via next-generation sequencing. To evaluate the strengths and weaknesses of both platforms, a comparative gene expression profiling study was conducted using RNA samples from 52 prostate tissues (normal, benign prostatic hyperplasia (BPH) and various prostate cancer (PCa) grades). Of all genes covered by both platforms, only 35% of the expression levels aligned, with 45% showing discrepancies. Both platforms identified the same 17 genes as potential PCa biomarkers. Microarray analysis identified an additional 253 genes that were not covered or not identified by smMIP technology, while smMIP technology identified eight markers not covered or not identified in the microarray core gene analysis, including fusion genes and splice variants. For high-grade prostate cancer (HG-PCa), the smMIP-method identified 8 markers, and the microarray identified 17 markers, with FOLH1, FAP and CLDN3 being common across both platforms. The choice of RNA expression analysis technology depends on research objectives; microarray technology is useful for the evaluation of a wide range of genes but has low throughput. In contrast, smMIP-based RNA sequencing enables sensitive analysis with minimal RNA in a medium- to high-throughput setting.

Diagnosis and Biomarker Screening of Endometrial Cancer Enabled by a Versatile Exosome Metabolic Fingerprint Platform.

Exosomes have emerged as a revolutionary tool for liquid biopsy (LB), as they carry specific cargo from cells. Profiling the metabolites of exosomes is crucial for cancer diagnosis and biomarker discovery. Herein, we propose a versatile platform for exosomal metabolite assay of endometrial cancer (EC). The platform is based on a nanostructured composite material comprising gold nanoparticle-coated magnetic COF with aptamer modification (Fe3O4@COF@Au-Apt). The unique design and novel synthesis strategy of Fe3O4@COF@Au-Apt provide the material with a large specific surface area, enabling the efficient and specific isolation of exosomes. The exosomes captured Fe3O4@COF@Au-Apt can be directly used as the laser desorption/ionization mass spectrometry (LDI-MS) matrix for rapid exosomal metabolic patterns. By integrating these functionalities into a single platform, the analytical process is simplified, eliminating the need for additional elution steps and minimizing potential sample loss, resulting in large-scale exosomal metabolic fingerprints. Combining with machine learning algorithms on the metabolic patterns, accurate discrimination between endometrial patients (EGs) and benign controls (CGs) was achieved, and the area under the receiver operating characteristic curve of the blind test cohort was 0.924. Confusion matrix analysis of important metabolic fingerprint features further demonstrates the high accuracy of the proposed approach toward EC diagnosis, with an overall accuracy of 94.1%. Moreover, four metabolites, namely, hydroxychalcone, l-acetylcarnitine, elaidic acid, and glutathione, have been identified as potential biomarkers of EC. These results highlight the great value of the integrated exosome metabolic fingerprint platform in facilitating low-cost and high-throughput characterization of exosomal metabolites for cancer diagnosis and biomarker discovery.

Triple and quadruple optimization for feature selection in cancer biomarker discovery

The proliferation of omics data has advanced cancer biomarker discovery but often falls short in external validation, mainly due to a narrow focus on prediction accuracy that neglects clinical utility and validation feasibility. We introduce three- and four-objective optimization strategies based on genetic algorithms to identify clinically actionable biomarkers in omics studies, addressing classification tasks aimed at distinguishing hard-to-differentiate cancer subtypes beyond histological analysis alone. Our hypothesis is that by optimizing more than one characteristic of cancer biomarkers, we may identify biomarkers that will enhance their success in external validation. Our objectives are to: (i) assess the biomarker panel’s accuracy using a machine learning (ML) framework; (ii) ensure the biomarkers exhibit significant fold-changes across subtypes, thereby boosting the success rate of PCR or immunohistochemistry validations; (iii) select a concise set of biomarkers to simplify the validation process and reduce clinical costs; and (iv) identify biomarkers crucial for predicting overall survival, which plays a significant role in determining the prognostic value of cancer subtypes. We implemented and applied triple and quadruple optimization algorithms to renal carcinoma gene expression data from TCGA. The study targets kidney cancer subtypes that are difficult to distinguish through histopathology methods. Selected RNA-seq biomarkers were assessed against the gold standard method, which relies solely on clinical information, and in external microarray-based validation datasets. Notably, these biomarkers achieved over 0.8 of accuracy in external validations and added significant value to survival predictions, outperforming the use of clinical data alone with a superior c-index. The provided tool also helps explore the trade-off between objectives, offering multiple solutions for clinical evaluation before proceeding to costly validation or clinical trials.

324 (PB312): Comprehensive Omics and Phenotypic Data Integration for Enhanced Drug Target and Biomarker Discovery in Cancer

324 (PB312): Comprehensive Omics and Phenotypic Data Integration for Enhanced Drug Target and Biomarker Discovery in Cancer

Potential of quantum machine learning for solving the real-world problem of cancer classification

Quantum machine learning (QML) algorithms have demonstrated the power of quantum computing for solving complex problems and big data in certain tasks. In this study, we explore the capabilities of QML for the classification of real-world biological large datasets including ten different cancer types based on gene expression values. By comparing the classification results obtained from the quantum algorithm with those from classical approaches, we disclose that the QML algorithm overall achieves comparable and reliable results. Moreover, we identify novel biomarkers that can contribute to the understanding of cancer biology. Some of these biomarkers are consistent with DNA promoter methylation. Our findings highlight the potential of QML in cancer classification and biomarker discovery, paving the way for future advancements in other disease research and clinical applications.

PUM1-TRAF3 fusion protein activates non-canonical NF-κB signaling via rescued NIK in biliary tract cancer

Discovery and verification of diagnostic or therapeutic biomarkers for biliary tract cancer (BTC) is challenging owing to the low prevalence of the disease. Here, we identified and investigated the clinical impact of a fusion gene, Pumilio1-tumor necrosis factor receptor-associated factor 3 (PUM1-TRAF3), caused by 1;14 chromosomal translocation in BTC. PUM1-TRAF3 was initially identified in the RNA-sequencing of five BTC surgical tissues and confirmed by fluorescence in situ hybridization. Expression of the fusion gene was validated in an expanded cohort (5/55, 9.1%). Establishment and molecular assessment of PUM1-TRAF3 expressing BTC cells revealed that PUM1-TRAF3 activates non-canonical NF-κB signaling via NF-κB-inducing kinase (NIK). Abnormal TRAF3 activity, driven by competitive binding of PUM1-TRAF3 and TRAF3 to NIK, led to NIK rescue followed by P52/RelB nuclear translocation, all of which were reverted by an NIK inhibitor. The elevated expression of NIK and activated NF-κB signaling was observed in the PUM1-TRAF3-expressing regions of patient tissues. Expression of the PUM1-TRAF3 fusion was significantly correlated with strong NIK expression, which is associated with a poorer prognosis for patients with BTC. Overall, our study identifies a new fusion gene, PUM1-TRAF3, that activates NIK and non-canonical NF-κB signaling, which may be beneficial for developing precise treatment strategies for BTC.

MOTAI: A Novel Method for the Study of O-GalNAcylation and Complex O-Glycosylation in Cancer.

The Tn antigen, an immature truncated O-glycosylation, is a promising biomarker for cancer detection and diagnosis. However, reliable methods for analyzing O-GalNAcylation and complex O-glycosylation are lacking. Here, we develop a novel method, MOTAI, for the sequential analysis of O-glycosylation using different O-glycoproteases. MOTAI conjugates glycopeptides on a solid support and releases different types of O-glycosylation through sequential enzymatic digestion by O-glycoproteases, including OpeRATOR and IMPa. Because OpeRATOR has less activity on O-GalNAcylation, MOTAI enriches O-GalNAcylation for subsequent analysis. We demonstrate the effectiveness of MOTAI by analyzing fetuin O-glycosylation and Jurkat cell lines. We then apply MOTAI to analyze colorectal cancer and benign colorectal polyps. We identify 32 Tn/sTn-glycoproteins and 43 T/sT-glycoproteins that are significantly increased in tumor tissues. Gene Ontology analysis reveals that most of these proteins are ECM proteins involved in the adhesion process of the intercellular matrix. Additionally, the protein disulfide isomerase CRELD2 has a significant difference in Tn expression, and the abnormally glycosylated T345 and S349 O-glycosylation sites in cancer group samples may promote the secretion of CRELD2 and ultimately tumorigenesis through ECM reshaping. In summary, MOTAI provides a powerful new tool for the in-depth analysis of O-GalNAcylation and complex O-glycosylation. It also reveals the upregulation of Tn/sTn-glycoproteins in colorectal cancer, which may provide new insights into cancer biology and biomarker discovery.

Optimising Extracellular Vesicle Metabolomic Methodology for Prostate Cancer Biomarker Discovery.

Conventional diagnostic tools for prostate cancer (PCa), such as prostate-specific antigen (PSA), transrectal ultrasound (TRUS), digital rectal examination (DRE), and tissue biopsy face, limitations in individual risk stratification due to invasiveness or reliability issues. Liquid biopsy is a less invasive and more accurate alternative. Metabolomic analysis of extracellular vesicles (EVs) holds a promise for detecting non-genetic alterations and biomarkers in PCa diagnosis and risk assessment. The current research gap in PCa lies in the lack of accurate biomarkers for early diagnosis and real-time monitoring of cancer progression or metastasis. Establishing a suitable approach for observing dynamic EV metabolic alterations that often occur earlier than being detectable by other omics technologies makes metabolomics valuable for early diagnosis and monitoring of PCa. Using four distinct metabolite extraction approaches, the metabolite cargo of PC3-derived large extracellular vesicles (lEVs) was evaluated using a combination of methanol, cell shearing using microbeads, and size exclusion filtration, as well as two fractionation chemistries (pHILIC and C18 chromatography) that are also examined. The unfiltered methanol-microbeads approach (MB-UF), followed by pHILIC LC-MS/MS for EV metabolite extraction and analysis, is effective. Identified metabolites such as L-glutamic acid, pyruvic acid, lactic acid, and methylmalonic acid have important links to PCa and are discussed. Our study, for the first time, has comprehensively evaluated the extraction and separation methods with a view to downstream sample integrity across omics platforms, and it presents an optimised protocol for EV metabolomics in PCa biomarker discovery.

NNBGWO-BRCA marker: Neural Network and binary grey wolf optimization based Breast cancer biomarker discovery framework using multi-omics dataset

Background and ObjectiveBreast cancer is a multifaceted condition characterized by diverse features and a substantial mortality rate, underscoring the imperative for timely detection and intervention. The utilization of multi-omics data has gained significant traction in recent years to identify biomarkers and classify subtypes in breast cancer. This kind of research idea from part to whole will also be an inevitable trend in future life science research. Deep learning can integrate and analyze multi-omics data to predict cancer subtypes, which can further drive targeted therapies. However, there are few articles leveraging the nature of deep learning for feature selection. Therefore, this paper proposes a Neural Network and Binary grey Wolf Optimization based BReast CAncer bioMarker (NNBGWO-BRCAMarker) discovery framework using multi-omics data to obtain a series of biomarkers for precise classification of breast cancer subtypes. MethodsNNBGWO-BRCAMarker consists of two phases: in the first phase, relevant genes are selected using the weights obtained from a trained feedforward neural network; in the second phase, the binary grey wolf optimization algorithm is leveraged to further screen the selected genes, resulting in a set of potential breast cancer biomarkers. ResultsThe SVM classifier with RBF kernel achieved a classification accuracy of 0.9242 ± 0.03 when trained using the 80 biomarkers identified by NNBGWO-BRCAMarker, as evidenced by the experimental results. We conducted a comprehensive gene set analysis, prognostic analysis, and druggability analysis, unveiling 25 druggable genes, 16 enriched pathways strongly linked to specific subtypes of breast cancer, and 8 genes linked to prognostic outcomes. ConclusionsThe proposed framework successfully identified 80 biomarkers from the multi-omics data, enabling accurate classification of breast cancer subtypes. This discovery may offer novel insights for clinicians to pursue in further studies.

An Updated Review on Molecular Biomarkers in Diagnosis and Therapy of Colorectal Cancer.

Colorectal cancer is one of the most common cancer types worldwide. Since colorectal cancer takes time to develop, its incidence and mortality can be treated effectively if it is detected in its early stages. As a result, non-invasive or invasive biomarkers play an essential role in the early diagnosis of colorectal cancer. Many experimental studies have been carried out to assess genetic, epigenetic, or protein markers in feces, serum, and tissue. It may be possible to find biomarkers that will help with the diagnosis of colorectal cancer by identifying the genes, RNAs, and/or proteins indicative of cancer growth. Recent advancements in the molecular subtypes of colorectal cancer, DNA methylation, microRNAs, long noncoding RNAs, exosomes, and their involvement in colorectal cancer have led to the discovery of novel biomarkers. In small-scale investigations, most biomarkers appear promising. However, large-scale clinical trials are required to validate their effectiveness before routine clinical implementation. Hence, this review focuses on small-scale investigations and results of big data analysis that may provide an overview of the biomarkers for the diagnosis, therapy, and prognosis of colorectal cancer.

A Tutorial Review of Labeling Methods in Mass Spectrometry-Based Quantitative Proteomics.

Recent advancements in mass spectrometry (MS) have revolutionized quantitative proteomics, with multiplex isotope labeling emerging as a key strategy for enhancing accuracy, precision, and throughput. This tutorial review offers a comprehensive overview of multiplex isotope labeling techniques, including precursor-based, mass defect-based, reporter ion-based, and hybrid labeling methods. It details their fundamental principles, advantages, and inherent limitations along with strategies to mitigate the limitation of ratio-distortion. This review will also cover the applications and latest progress in these labeling techniques across various domains, including cancer biomarker discovery, neuroproteomics, post-translational modification analysis, cross-linking MS, and single-cell proteomics. This Review aims to provide guidance for researchers on selecting appropriate methods for their specific goals while also highlighting the potential future directions in this rapidly evolving field.

DeepKEGG: a multi-omics data integration framework with biological insights for cancer recurrence prediction and biomarker discovery.

Deep learning-based multi-omics data integration methods have the capability to reveal the mechanisms of cancer development, discover cancer biomarkers and identify pathogenic targets. However, current methods ignore the potential correlations between samples in integrating multi-omics data. In addition, providing accurate biological explanations still poses significant challenges due to the complexity of deep learning models. Therefore, there is an urgent need for a deep learning-based multi-omics integration method to explore the potential correlations between samples and provide model interpretability. Herein, we propose a novel interpretable multi-omics data integration method (DeepKEGG) for cancer recurrence prediction and biomarker discovery. In DeepKEGG, a biological hierarchical module is designed for local connections of neuron nodes and model interpretability based on the biological relationship between genes/miRNAs and pathways. In addition, a pathway self-attention module is constructed to explore the correlation between different samples and generate the potential pathway feature representation for enhancing the prediction performance of the model. Lastly, an attribution-based feature importance calculation method is utilized to discover biomarkers related to cancer recurrence and provide a biological interpretation of the model. Experimental results demonstrate that DeepKEGG outperforms other state-of-the-art methods in 5-fold cross validation. Furthermore, case studies also indicate that DeepKEGG serves as an effective tool for biomarker discovery. The code is available at https://github.com/lanbiolab/DeepKEGG.

Abstract 7092: Cancer biomarker discovery with scalable biofluid proteomics

Abstract Biofluid proteomics opens opportunities for the discovery of early detection, diagnostic, and prognostic cancer biomarkers through non-invasive or minimally invasive screening. Plasma and serum are accessible sample types, however, the high dynamic range of proteins creates analytical challenges for the sensitive detection of potential biomarkers. In this work, we compare two proteomic approaches for large-scale cancer biomarker discovery: the Olink Proximity Extension Assay and the Seer Proteograph Product Suite. Olink Explore 3072 is an immunoassay that gains specificity by using dual antibody recognition to measure approx. 3,000 proteins in each sample. Antibody pairs are tagged with oligonucleotides, which hybridize and are amplified before measurement by next generation sequencing. Seer Proteograph is an enrichment strategy that uses a proprietary mix of functionalized magnetic nanoparticles with affinity for certain amino acids and functional groups in native state proteins. The resulting peptides are measured using untargeted mass spectrometry to identify proteins at the peptide level, paving the way for the identification of proteoforms and peptide variants. Aliquots of human plasma from 78 samples representing a range of cancer indications were obtained through the Discovery Life Sciences biobank and used for analysis by both the Olink and Proteograph workflows. Olink analysis achieved Pearson correlation scores of greater than or equal to 0.8 when correlating each of seven process controls to each other for all approx. 3000 proteins. On average, greater than or equal to 75% of the proteins were detected across all panels in all sample groups except for Normal and Benign groups (64% and 74% respectively). After stringent FDR filtering to ensure only high-quality data was analyzed, the Seer Proteograph workflow quantified 4,323 proteins across all 78 samples. Pearson correlation scores of greater than or equal to 0.96 were achieved when comparing digestion controls, process controls, and sample cleanup controls. Both platforms achieved detection of proteins across the dynamic range of plasma, including low abundance proteins. The Olink Explore 3072 and Seer Proteograph platforms were found to be complementary, with approx. 15% overlap in measured proteins. Cancer biomarker discovery at the proteomic level is bolstered by recent advances in analytical capabilities, as revealed by the Olink Explore and Seer Proteograph platforms. Importantly, complementary implementation of these two workflows will enable comprehensive biomarker analysis. Citation Format: Danielle B. Gutierrez, Jessica L. Moore, Jaison Arivalagan, Tiffany Louie, Leslie Wilkinson, Ray Clemens. Cancer biomarker discovery with scalable biofluid proteomics [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 7092.

Abstract 936: Targeted proteomics and next-generation sequencing (NGS) for biomarker discovery in metastatic colorectal cancer (mCRC)

Abstract Background: Multi-omics analysis is widely used in cancer diagnosis, prognosis and prediction of treatment response. Here we introduce the use of targeted proteomics coupled with NGS as tools to assess oncogenic alterations along with quantitative protein expression in patients with mCRC. Our proteomics panel included proteins that are antibody drug conjugates (ADCs) targets (ERBB 1-4, TROP2, CEACAM5 etc.) as well as proteins involved in sensitivity to ADC payload (SLFN11, drug efflux pump, TOP1 etc.). Methods: The HORIZON III trial (NCT00384176) evaluated FOLFOX + bevacizumab or cediranib as first-line treatment in patients with mCRC (Schmoll J Clin Oncol 2012). In the current analysis, a subset of samples from the HORIZON III trial were assessed for molecular alterations using NGS and proteomics. Pre-existing HER2 IHC scores were used as previously referenced (Cecchi J Clin Oncol 2023). Samples were processed (n=135) with an automated extraction and library preparation method, then sequenced with both a Whole Exome and targeted-custom gene panel. NGS analysis had a success rate of 116/135 samples (87.3%). Absolute quantitation of protein targets was performed using parallel reaction monitoring (PRM) mass spectrometry after collecting tumor epithelial cells by laser microdissection and punch biopsies (n=60) followed by trypsin digestion. Results: The NGS analysis showed that patients had mutated RAS (50%), BRAF (10%), PIK3CA (16%), and TMB-H (18%) and MSI-H (5%) status. A distinguished pattern of mutual exclusivity among KRAS, NRAS, and BRAF mutations is evident. The HER2-positive patients were wildtype RAS, BRAF, and PIK3CA, MSS and TMB-L. Samples that were IHC HER2-positive matched those with NGS ERBB2 amplification. Using PRM, nearly all the evaluable samples had detectable HER2. HER2 IHC 3+ sample showed the highest protein expression at 30,000 amol/ug (equivalent to 1M receptors/cell). Within the HER2-negative patients, HER2 protein expression showed broad dynamic range. CEACAM5 and HER2 are the only ADC targets that reached the level of 1M receptors/cell. Highest TROP2 protein expression is 5 times lower than the highest HER2 value. Conclusions: Our preliminary findings for this cohort indicate the benefit of the use of multi-modal methodologies to inform better strategies for patient stratification and precision medicine, though further assessment of proteogenomic characteristics is needed. This data may inform future clinical development and guide patient eligibility to clinical trials and possible treatment sequencing for ADCs. Citation Format: William J. Kelly, Barrett Nuttall, Marlon Rebelatto, Marcus Schick, Armin Meier, Steve Sweet, Beom-Jun Kim, David Chain, Yeoun Jin Kim, Zhongwu Lai, Ying Wang, Todd H. Creasy, Ramya Gurrapu, Sakshi Gulati, Alice Barkell, Brian Dougherty, Danielle Carroll, Fabiola Cecchi. Targeted proteomics and next-generation sequencing (NGS) for biomarker discovery in metastatic colorectal cancer (mCRC) [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 936.

Abstract 4930: Identification of CBX7 and PCDHB18 as novel prognostic biomarkers of cervical cancer: RNA-sequencing and machine learning analysis

Abstract Background- Cervical cancer, ranking fourth in prevalence among women worldwide, is highly preventable when diagnosed early. The advent of innovative technologies including bioinformatics and machine learning, is revolutionizing the discovery and development of novel biomarkers for cancer. Our study uniquely integrates artificial neural networks with RNA-sequencing data of cervical cancer patients. This combination enhances the accuracy and reliability of biomarker predictions, providing a better understanding of the potential clinical utility of the identified biomarkers. Methods- RNA-sequencing data and clinicopathological details of 304 Cervical Squamous Cell Carcinoma and Endocervical Adenocarcinoma (CESC) were obtained from the GDAC database (https://gdac.broadinstitute.org/). Differentially expressed genes (DEGs) were identified (P<0.05, |log2 fold change (FC)| > 1.5, false discovery rate (FDR) < 0.05). Pathway enrichment analysis was performed and protein-protein interactions of DEGs were constructed using the STRING database. Prognostic biomarkers were identified using Kaplan-Meier and Cox proportional hazard methods adjusted for potential confounders such as age, disease stage, and comorbidities (HR<1, p<0.05). Deep learning algorithms were applied for predictive marker identification, utilizing Weight by Correlation feature selection. The model used AUC (Area Under the Curve), accuracy, MSE (Mean Squared Error), and R2 (R-squared) as evaluation metrics in a 70-30 training-test split. A combined ROC curve assessed diagnostic biomarkers, validated externally with a GEO dataset on cervical cancer patients. Results- In our study, 4153 DEGs were identified. Pathway analysis revealed that key dysregulated genes play a pivotal role in extracellular matrix organization. The survival analysis identified that six upregulated genes (CT62, SLC7A5P1, C10orf110, FDPSL2A, TNFSF15, and TTLL13) and seven downregulated genes (BAI3, CBX7, DKFZp566F0947, PCDHB18, GRAPL, PCDHB19P, and C4orf38) decreased the overall survival in patients. The machine learning model demonstrated high predictive accuracy (AUC=1, accuracy=99.02%, R2=0.99), identifying twenty genes with a positive correlation to cervical cancer risk. Notably, CBX7 emerged as a prognostic and diagnostic biomarker (AUC=0.99, sensitivity=0.93, specificity=1.00). Conclusion- Our study uncovers the prognostic significance of two novel genes in cervical cancer: CBX7, a vital regulator of tumor suppression, and PCDHB18, a member of the protocadherin beta gene cluster functioning as a cell adhesion molecule. Downregulation of these genes is associated with decreased overall survival. Further functional analyses and validation of these candidate biomarkers are crucial to fully assess their potential clinical value in cervical cancer. Citation Format: Ghazaleh Pourali, Mohsen Zeinali, Mahshid Arastonejad, Nima Khalili-Tanha, Elham Nazari, Ghazaleh Khalili-Tanha, Adetunji T. Toriola. Identification of CBX7 and PCDHB18 as novel prognostic biomarkers of cervical cancer: RNA-sequencing and machine learning analysis [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 4930.

Abstract 1074: Plasma EV profiling facilitates low abundance biomarker discovery in pancreatic cancer

Abstract The lack of reliable biomarkers for early detection of pancreatic ductal adenocarcinoma (PDAC) results in poor prognosis due to advanced disease at diagnosis. Biopsies for pancreatic cancer are known to be particularly difficult to perform because of the location of the pancreas and low volume of primary tumors that metastasize aggressively. Currently, there is no blood test or other screening modality in use, for stratification of individuals with higher-than-average risk (5-8-fold) of developing PDAC in their lifetime. Given this existing void in the US and worldwide, an FDA-compliant, minimally invasive, quantitative biomarker test for the early detection of PDAC would be critical for screening subpopulations with, a) inherent genetic or familial predisposition; history of chronic pancreatitis; c) precursor lesions of pancreas; and d) older patients (>50 years) with new onset diabetes or sudden weight loss. Extracellular vesicles (EVs) are shed by normal and cancer cells and carry a rich molecular cargo that can be leveraged for biomarker discovery. Whereas molecular profiling of bio-fluids poses intrinsic limitations for detection of low abundant, disease-specific biomarkers due to matrix effects; EVs, on the other hand, offer promise as a robust and minimally invasive biomarker resource. Mass spectrometry data were developed in our laboratory, based on multi-omics (proteomics, lipidomics and metabolomics) analyses of plasma EVs derived from patients diagnosed with early-stage PC (N=60), pancreatitis (N=39), precursor lesions of pancreas (N=45) as well as normal controls (N=50). These clinically annotated samples were made available by the Georgetown University Medical Center (GUMC) repository. Feature selection using a training-validation design allowed the development of a multiplexed biomarker panel comprising of 12-analytes (5-lipids, 2-metabolites and 5-proteins) that can robustly stratify early-stage PC from normal controls with high accuracy (AUC>95%). Availability of a FDA compliant, multi-analyte-based classification algorithm that can stratify patients with precursor lesions of the pancreas that are at potential risk of progression to PC with > 90% specificity and sensitivity is a critical clinical need. Refinement of panel to improve positive predictive value as well as identification of markers predictive of progression of precursor lesions to malignant transformation is ongoing. Citation Format: Shivani Bansal, Shu Wang, Yaoxiang Li, Sunil Bansal, Jill Smith, John B. Tyburski, Keith Unger, Amrita Cheema. Plasma EV profiling facilitates low abundance biomarker discovery in pancreatic cancer [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 1074.