Metabolomics in oncology

A Comparative Study of Network-Based Machine Learning Approaches for Binary Classification in Metabolomics

Background/Objectives: Metabolomics has recently emerged as a key tool in the biological sciences, offering insights into metabolic pathways and processes. Over the last decade, network-based machine learning approaches have gained significant popularity and application across various fields. While several studies have utilized metabolomics profiles for sample classification, many network-based machine learning approaches remain unexplored for metabolomic-based classification tasks. This study aims to compare the performance of various network-based machine learning approaches, including recently developed methods, in metabolomics-based classification. Methods: A standard data preprocessing procedure was applied to 17 metabolomic datasets, and Bayesian neural network (BNN), convolutional neural network (CNN), feedforward neural network (FNN), Kolmogorov-Arnold network (KAN), and spiking neural network (SNN) were evaluated on each dataset. The datasets varied widely in size, mass spectrometry method, and response variable. Results: With respect to AUC on test data, BNN, CNN, FNN, KAN, and SNN were the top-performing models in 4, 1, 5, 3, and 4 of the 17 datasets, respectively. Regarding F1-score, the top-performing models were BNN (3 datasets), CNN (3 datasets), FNN (4 datasets), KAN (4 datasets), and SNN (3 datasets). For accuracy, BNN, CNN, FNN, KAN, and SNN performed best in 4, 1, 4, 4, and 4 datasets, respectively. Conclusions: No network-based modeling approach consistently outperformed others across the metrics of AUC, F1-score, or accuracy. Our results indicate that while no single network-based modeling approach is superior for metabolomics-based classification tasks, BNN, KAN, and SNN may be underappreciated and underutilized relative to the more commonly used CNN and FNN.

Exploring Metabolic Shifts in Kidney Cancer and Non-Cancer Cells Under Pro- and Anti-Apoptotic Treatments Using NMR Metabolomics

This study investigates the metabolic responses of cancerous (RCC) and non-cancerous (HK2) kidney cells to treatment with Staurosporine (STAU), which has a pro-apoptotic effect, and Bongkrekic acid (BKA), which has an anti-apoptotic effect, individually and in combination, using 1H NMR metabolomics to identify metabolite markers linked to mitochondrial apoptotic pathways. BKA had minimal metabolic effects in RCC cells, suggesting its role in preserving mitochondrial function without significantly altering metabolic pathways. In contrast, STAU induced substantial metabolic reprogramming in RCC cells, disrupting energy production, redox balance, and biosynthesis, thereby triggering apoptotic pathways. The combined treatment of BKA and STAU primarily mirrored the effects of STAU alone, with BKA showing little capacity to counteract the pro-apoptotic effects. In non-cancerous HK2 cells, the metabolic alterations were far less pronounced, highlighting key differences in the metabolic responses of cancerous and non-cancerous cells. RCC cells displayed greater metabolic flexibility, while HK2 cells maintained a more regulated metabolic state. These findings emphasize the potential for targeting cancer-specific metabolic vulnerabilities while sparing non-cancerous cells, underscoring the value of metabolomics in understanding apoptotic and anti-apoptotic mechanisms. Future studies should validate these results in vivo and explore their potential for personalized treatment strategies.

Global metabolomic profiling of tumor tissue and paired serum samples to identify biomarkers for response to neoadjuvant FOLFIRINOX treatment of human pancreatic cancer

Neoadjuvant chemotherapy (NAT) is increasingly used for the treatment of non-metastatic pancreatic ductal adenocarcinoma (PDAC) and is established as a standard of care for borderline resectable and locally advanced PDAC. However, full exploitation of its clinical benefits is limited by the lack of biomarkers that assess treatment response. To address this unmet need, global metabolomic profiling was performed on tumor tissue and paired serum samples from patients with treatment-naïve (TN; n = 18) and neoadjuvant leucovorin calcium (folinic acid), fluorouracil, irinotecan hydrochloride and oxaliplatin (FOLFIRINOX)-treated (NAT; n = 17) PDAC using liquid chromatography mass spectrometry. Differentially abundant metabolites (DAMs) in TN versus NAT groups were identified and their correlation with various clinical parameters was assessed. Metabolomics profiling identified 40 tissue and five serum DAMs in TN versus NAT PDAC. In general, DAMs associated with amino acid and nucleotide metabolism were lower in NAT compared to TN. Four DAMs-3-hydroxybutyric acid (BHB), 3-carboxy-4-methyl-5-propyl-2-furanpropanoic acid (CMPF), glycochenodeoxycholate and citrulline-were common to both tissue and serum and showed a similar pattern of differential abundance in both groups. A strong positive correlation was observed between serum carbohydrate 19-9 antigen (CA 19-9) and tissue carnitines (C12, C18, C18:2) and N8-acetylspermidine. The reduction in CA 19-9 following NAT correlated negatively with serum deoxycholate levels, and the latter correlated positively with survival. This study revealed neoadjuvant-chemotherapy-induced changes in metabolic pathways in PDAC, mainly amino acid and nucleotide metabolism, and these correlated with reduced CA 19-9 following neoadjuvant FOLFIRINOX treatment.

Novel clinical trial designs emerging from the molecular reclassification of cancer

Next-generation sequencing has revealed the disruptive reality that advanced/metastatic cancers have complex and individually distinct genomic landscapes, necessitating a rethinking of treatment strategies and clinical trial designs. Indeed, the molecular reclassification of cancer suggests that it is the molecular underpinnings of the disease, rather than the tissue of origin, that mostly drives outcomes. Consequently, oncology clinical trials have evolved from standard phase 1, 2, and 3 tissue-specific studies; to tissue-specific, biomarker-driven trials; to tissue-agnostic trials untethered from histology (all drug-centered designs); and, ultimately, to patient-centered, N-of-1 precision medicine studies in which each patient receives a personalized, biomarker-matched therapy/combination of drugs. Innovative technologies beyond genomics, including those that address transcriptomics, immunomics, proteomics, functional impact, epigenetic changes, and metabolomics, are enabling further refinement and customization of therapy. Decentralized studies have the potential to improve access to trials and precision medicine approaches for underserved minorities. Evaluation of real-world data, assessment of patient-reported outcomes, use of registry protocols, interrogation of exceptional responders, and exploitation of synthetic arms have all contributed to personalized therapeutic approaches. With greater than 1 × 1012 potential patterns of genomic alterations and greater than 4.5 million possible three-drug combinations, the deployment of artificial intelligence/machine learning may be necessary for the optimization of individual therapy and, in the near future, also may permit the discovery of new treatments in real time.

The human volatilome meets cancer diagnostics: past, present, and future of noninvasive applications

BACKGROUND

Cancer is a significant public health problem, causing dozens of millions of deaths annually. New cancer screening programs are urgently needed for early cancer detection, as this approach can improve treatment outcomes and increase patient survival. The search for affordable, noninvasive, and highly accurate cancer detection methods revealed a valuable source of tumor-derived metabolites in the human metabolome through the exploration of volatile organic compounds (VOCs) in noninvasive biofluids.

AIM OF REVIEW

This review discusses volatilomics-based approaches for cancer detection using noninvasive biomatrices (breath, saliva, skin secretions, urine, feces, and earwax). We presented the historical background, the latest approaches, and the required stages for clinical validation of volatilomics-based methods, which are still lacking in terms of making noninvasive methods available and widespread to the population. Furthermore, insights into the usefulness and challenges of volatilomics in clinical implementation steps for each biofluid are highlighted.

KEY SCIENTIFIC CONCEPTS OF REVIEW

We outline the methodologies for using noninvasive biomatrices with up-and-coming clinical applications in cancer diagnostics. Several challenges and advantages associated with the use of each biomatrix are discussed, aiming at encouraging the scientific community to strengthen efforts toward the necessary steps to speed up the clinical translation of volatile-based cancer detection methods, as well as discussing in favor of (i) hybrid applications (i.e., using more than one biomatrix) to describe metabolite modulations that can be "cancer volatile fingerprints" and (ii) in multi-omics approaches integrating genomics, transcriptomics, and proteomics into the volatilomic data, which might be a breakthrough for diagnostic purposes, onco-pathway assessment, and biomarker validations.

Cancer pharmacoinformatics: Databases and analytical tools

Cancer is a subject of extensive investigation, and the utilization of omics technology has resulted in the generation of substantial volumes of big data in cancer research. Numerous databases are being developed to manage and organize this data effectively. These databases encompass various domains such as genomics, transcriptomics, proteomics, metabolomics, immunology, and drug discovery. The application of computational tools into various core components of pharmaceutical sciences constitutes "Pharmacoinformatics", an emerging paradigm in rational drug discovery. The three major features of pharmacoinformatics include (i) Structure modelling of putative drugs and targets, (ii) Compilation of databases and analysis using statistical approaches, and (iii) Employing artificial intelligence/machine learning algorithms for the discovery of novel therapeutic molecules. The development, updating, and analysis of databases using statistical approaches play a pivotal role in pharmacoinformatics. Multiple software tools are associated with oncoinformatics research. This review catalogs the databases and computational tools related to cancer drug discovery and highlights their potential implications in the pharmacoinformatics of cancer.

Surface-Enhanced Raman Scattering Monitoring of Tryptophan Dynamics in 3D Pancreatic Tumor Models

In the intricate landscape of the tumor microenvironment, both cancer and stromal cells undergo rapid metabolic adaptations to support their growth. Given the relevant role of the metabolic secretome in fueling tumor progression, its unique metabolic characteristics have gained prominence as potential biomarkers and therapeutic targets. As a result, rapid and accurate tools have been developed to track metabolic changes in the tumor microenvironment with high sensitivity and resolution. Surface-enhanced Raman scattering (SERS) is a highly sensitive analytical technique and has been proven efficient toward the detection of metabolites in biological media. However, profiling secreted metabolites in complex cellular environments such as those in tumor-stroma 3D in vitro models remains challenging. To address this limitation, we employed a SERS-based strategy to investigate the metabolic secretome of pancreatic tumor models within 3D cultures. We aimed to monitor the immunosuppressive potential of stratified pancreatic cancer-stroma spheroids as compared to 3D cultures of either pancreatic cancer cells or cancer-associated fibroblasts, focusing on the metabolic conversion of tryptophan into kynurenine by the IDO-1 enzyme. We additionally sought to elucidate the dynamics of tryptophan consumption in correlation with the size, temporal evolution, and composition of the spheroids, as well as assessing the effects of different drugs targeting the IDO-1 machinery. As a result, we confirm that SERS can be a valuable tool toward the optimization of cancer spheroids, in connection with their tryptophan metabolizing capacity, potentially allowing high-throughput spheroid analysis.

Recent Advancements and Innovations in Pediatric Precision Oncology

Precision oncology incorporates comprehensive genomic profiling into the individualized clinical care of pediatric cancer patients. In recent years, comprehensive pan-cancer analyses have led to the successful implementation of genomics-based pediatric trials and accelerated approval of novel targeted agents. In addition, disease-specific studies have resulted in molecular subclassification of myriad cancer types with subsequent tailoring of treatment intensity based on the patient's prognostic factors. This review discusses the progress of the field and highlights developments that are leading to more personalized cancer care and improved patient outcomes. Increased understanding of the evolution of precision oncology over recent decades emphasizes the tremendous impact of improved genomic applications. New technologies and improved diagnostic modalities offer further promise for future advancements within the field.

Metabolic Profiling of Aromatic Compounds

Metabolic profiling is a powerful modern tool in searching for novel biomarkers and indicators of normal or pathological processes in the body [...].

Unveiling Disrupted Lipid Metabolism in Benign Prostate Hyperplasia, Prostate Cancer, and Metastatic Patients: Insights from a Colombian Nested Case-Control Study

Prostate cancer is a significant global health concern, and its prevalence is increasing worldwide. Despite extensive research efforts, the complexity of the disease remains challenging with respect to fully understanding it. Metabolomics has emerged as a powerful approach to understanding prostate cancer by assessing comprehensive metabolite profiles in biological samples. In this study, metabolic profiles of patients with benign prostatic hyperplasia (BPH), prostate cancer (PCa), and metastatic prostate cancer (Met) were characterized using an untargeted approach that included metabolomics and lipidomics via liquid chromatography and gas chromatography coupled with high-resolution mass spectrometry. Comparative analysis among these groups revealed distinct metabolic profiles, primarily associated with lipid biosynthetic pathways, such as biosynthesis of unsaturated fatty acids, fatty acid degradation and elongation, and sphingolipid and linoleic acid metabolism. PCa patients showed lower levels of amino acids, glycerolipids, glycerophospholipids, sphingolipids, and carnitines compared to BPH patients. Compared to Met patients, PCa patients had reduced metabolites in the glycerolipid, glycerophospholipid, and sphingolipid groups, along with increased amino acids and carbohydrates. These altered metabolic profiles provide insights into the underlying pathways of prostate cancer's progression, potentially aiding the development of new diagnostic, and therapeutic strategies.